SPRINGER BRIEFS IN COMPUTER SCIENCE

Shaun Foster
David Halbstein

Integrating
3D Modeling,
Photogrammetry
and Design
SpringerBriefs in Computer Science

Series editors
Stan Zdonik
Peng Ning
Shashi Shekhar
Jonathan Katz
Xindong Wu
Lakhmi C. Jain
David Padua
Xuemin Shen
Borko Furht
V. S. Subrahmanian
Martial Hebert
Katsushi Ikeuchi
Bruno Siciliano

For further volumes:

[Link]
Shaun Foster David Halbstein
•

Integrating 3D Modeling,
Photogrammetry and Design

123
Shaun Foster David Halbstein
Rochester Institute of Technology Rochester Institute of Technology
Rochester, NY Rochester, NY
USA USA

ISSN 2191-5768 ISSN 2191-5776 (electronic)

ISBN 978-1-4471-6328-2 ISBN 978-1-4471-6329-9 (eBook)
DOI 10.1007/978-1-4471-6329-9
Springer London Heidelberg New York Dordrecht

Library of Congress Control Number: 2013958143

 The Author(s) 2014

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed. Exempted from this legal reservation are brief
excerpts in connection with reviews or scholarly analysis or material supplied specifically for the
purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the
work. Duplication of this publication or parts thereof is permitted only under the provisions of
the Copyright Law of the Publisher’s location, in its current version, and permission for use must
always be obtained from Springer. Permissions for use may be obtained through RightsLink at the
Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for
any errors or omissions that may be made. The publisher makes no warranty, express or implied, with
respect to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media ([Link])

Preface

This book was inspired by observations, research over the last several years, and
subsequent talk on the subject at SIGGRAPH 2012. The field of 3D Digital Design
has added several photogrammetry and retopology-based technologies resulting in
many new and accelerated workflows, which have also affected the 3D design
education process.
It has been a journey not only in the writing, but in the research and approach.
What we describe in these pages is an ongoing case study that started many years
ago, and will continue for as long as we continue to teach and to learn. As ‘‘early
adopters’’ in the field of technology influenced design-based education we natu-
rally push for advanced technology, but as part of a larger institution we must
accept what the constraints of budget and adherence to tradition dictate. We have
found that these opposing forces create a natural balance; that in some cases
constraints lead to greater creativity than freedom ever can—but in other cases the
opposite is true.
The book looks at the convergent nature of technology and its relationship to
the field of photogrammetry, and 3D design. This is a facet of a broader discussion
of the nature of technology itself and the relationship of technology to art, as well
as an examination of the educational process itself. We have addressed the
question of how we are adjusting and will continue to adapt to further disruptive
technologies.
Over the years, we have been faced with this question and more; in the
development of a post-secondary curriculum we are always questioning—are we
teaching art and design, or are we training technology? Are we teaching our
students to think, to ideate, to create, or are we teaching them how to use tools?
Clearly, the two concepts are inexorably intertwined, but in any educational
institution or college curriculum it is important to make the distinction, and to
understand the areas that are clearly defined as well as those that are not.
Each year we are presented with new technology. Software and hardware tools
are developing at an extremely rapid rate and with each new change, each new
outgrowth, we must make important decisions about whether or not these new
tools are important enough to incorporate into our existing curriculum. Do we keep
pace for the sake of keeping pace, or will our adaptation to change foster creativity
and new thought? Given the existing constraints of time and budget, what are we
willing to sacrifice in order to embrace these changes and fold them into our

v
vi Preface

curriculum? Which tools, methods, approaches, or classes can give way to this
change? Clearly these questions seek answers, this book offers insights for ways to
integrate some of these new technologies into the field of design, and from a
broader standpoint it also looks ahead, raising further questions and looking to the
near future as to what additional technologies might cause further disruptions to
3D design as well as wonderful creative opportunities.
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . ......... . . . . . . . . . . 1
1.1 Introduction . . . . . . . . . . . . . . . . . . ......... . . . . . . . . . . 1
1.2 An Introduction to Photogrammetry. . ......... . . . . . . . . . . 2
1.3 Computer Vision . . . . . . . . . . . . . . . ......... . . . . . . . . . . 3
1.4 Surface Topology and Retopologizing Workflows . . . . . . . . . . . 3
1.5 Design Process . . . . . . . . . . . . . . . . ......... . . . . . . . . . . 4

2 Technology Leaders and Adopters . . . . . . . . . . . . . . . . . . . . . . . . 5

2.1 The Nature of Technological Spread . . . . . . . . . . . . . . . . . . . . 5
2.2 A Further Introduction and History of Photogrammetry . . . . . . . 7
2.2.1 Photography and Plane Tables . . . . . . . . . . . . . . . . . . . 8
2.2.2 Stereoplotters and Airplanes . . . . . . . . . . . . . . . . . . . . . 9
2.2.3 Computers and Mathematical Models . . . . . . . . . . . . . . 9
2.2.4 Digital and Computer Vision . . . . . . . . . . . . . . . . . . . . 9
2.2.5 Looking Ahead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.6 The Cycle Between Art and Science . . . . . . . . . . . . . . . 10
2.3 Photogrammetry Resources Connected to the Entertainment
Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.4 Microsoft Kinect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.5 Microsoft (CV) Kinect Fusion. . . . . . . . . . . . . . . . . . . . . . . . . 11
2.6 Photosynth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.7 Weta Digital: Photospace . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.8 Autodesk 123D Catch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.9 Challenges and Current Limitations . . . . . . . . . . . . . . . . . . . . . 14
2.9.1 Occlusions and Number of Photographs Necessary . . . . . 14
2.9.2 Photographs Need Features . . . . . . . . . . . . . . . . . . . . . 14
2.9.3 No Transparent, Reflective, or Glossy Subjects . . . . . . . 15
2.9.4 Subjects Cannot Move During the Image
Capture Process . . . . . . . . . . . . . . . . . . . . . . . . . . ... 15
2.9.5 Consistent Lighting . . . . . . . . . . . . . . . . . . . . . . . . ... 15
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 15

vii
viii Contents

3 Conceptualizing Basic 3D Modeling Workflows. . . . . . . . . . . . . . . 17

3.1 Compartmentalizing the Learning Curve . . . . . . . . . . . . . . . . . 17
3.2 3D Modeling: A General Conceptual Approach . . . . . . . . . . . . 20
3.3 NURBS Patch Modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.3.1 Methodologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.3.2 Attach/Detach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.4 Box Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.4.1 Langer’s Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.5 Normal Mapping, Displacement Mapping, and UV Layout. . . . . 37
3.6 Subtractive Modeling Workflow . . . . . . . . . . . . . . . . . . . . . . . 42
3.6.1 Topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.6.2 Retopology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.7 Digital Sculpture: Case Studies Using Pixologic ZBrush . . . . . . 51
3.7.1 Little Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.7.2 Batman . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

4 Photogrammetry and Design Workflows . . . . . . . . . . . . . . . . . . . . 59

4.1 Photogrammetry Workflows . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.2 Case Study: Roberley Bell . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.3 Case Study: Sucker Punch Studio’s . . . . . . . . . . . . . . . . . . . . . 66
4.3.1 3D Scanning of Faces and Clothing for
‘‘Infamous: Second Son’’ Game . . . . . . ............ 66
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............ 67

5 3D Design and Photogrammetry . . . . . . . . . . . . . . . . . . . . . . . . . . 69

5.1 Elements of Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.1.1 Color . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
5.1.2 Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.1.3 Shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.1.4 Texture/Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.1.5 Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
5.1.6 Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.2 Principles of Design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.2.1 Camera: Lens, Film, and Exposure . . . . . . . . . . . . . . . . 86
5.2.2 Cinematography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.2.3 Unity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.2.4 Gestalt Grouping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.2.5 Emphasis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5.2.6 Balance: Symmetry and Asymmetry . . . . . . . . . . . . . . . 91
5.2.7 Repetition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.2.8 Proximity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5.2.9 Dynamics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Contents ix

6 Conclusions and Looking Forward . . . . . . . . . . . . . . . . . . . . . . . . 97

Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

7 Resources and Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

7.1 Photogrammetry Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
7.2 Retopologizing Resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
7.3 Color Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Chapter 1
Introduction

Abstract The introductory chapter opens the book by giving a brief overview of
the main areas of discussion: An introduction to photogrammetry, computer
vision, 3D modeling, surface topology, and retopology and finishes with a brief
discussion of the design process.

 
Keywords Introduction Photogrammetry Computervision  design process 
  
Retapology 3d design Education Entertainment field

1.1 Introduction

The focus of this book is to identify and discuss the rapid changes to 3D computer
modeling workflows, as represented by the integration of photogrammetry, high-
polygon density-interactive modeling, and retopology software capabilities. We
recognize that these approaches have been part of the entertainment and broader
‘‘visualization/remote sensing industries’’ for many years; however, our focus is on
the accessibility in the field of design and an examination of the impact these
technologies will have on the field of design as well as the pedagogical nature of
3D design. As these advanced capabilities become less expensive, more refined,
and therefore more available to secondary and post-secondary educational insti-
tutions, it is incumbent upon those institutions to address their integration into a
changing curriculum and recognize the changes these approaches represent to a
pedagogic paradigm. In addition, we will also give a historical context for the tools
for the fields these technologies originated in using this as a basis to briefly discuss
technological convergence as a trend between multiple fields, and to the rela-
tionship it has to elements and principles of design in addition to considering what
further aesthetic and technical convergences might be on the horizon.
Therefore, within the pages of this book, we will discuss modeling methodologies
that are currently being taught, explore their advantages and disadvantages, and offer
conjecture as to how this field is changing from the perspective of education and

S. Foster and D. Halbstein, Integrating 3D Modeling, Photogrammetry and Design, 1

SpringerBriefs in Computer Science, DOI: 10.1007/978-1-4471-6329-9_1,
 The Author(s) 2014
2 1 Introduction

curriculum development. The tools, workflows, and art of generating 3D computer

models are expanding and rapidly changing. Over the past 20 years, most 3D objects
have been created by highly trained, technical 3D artists through a labor-intensive
process. This usually starts with simple object, adding complexity until the desired
form is reached. A small minority of 3D objects were created using scanning 3D
technology which has not gained broad use due to high cost and also difficulty of
using the final product generated by laser scan. The 3D pieces generated by these
scans were ‘‘heavy’’ on geometry, starting in the hundreds of thousands to millions
of polygons which make them inefficient to modify in some way, for example a high-
polygon object will take much more processing power to bend, twist, attach to other
objects, and also there is higher computer costs for rendering, the transformation
process of turning 3D objects into 2D still images.
Several maturing technological components are combining which open a very
different workflow, reducing the costs requirements of human labor and output of
heavy geometry. The first component is photogrammetry technology, which is the
process of using photography to determine geometric measurements of objects.
The second are several technologies from the field of computer vision. This field is
applying computer algorithms that automate the understanding and output from
photogrammetry. Finally, the third component is ‘‘retopologizing.’’ New tools in
restructuring geometry into clean and efficient forms have the potential to instantly
create 3D objects and optimize their geometry.
As in any emerging technology-dependent medium, the tools used to create
visual images have a great impact on the final output. From an aesthetic point of
view this can be both unintentional and undesirable. Rather than allow the limi-
tations of technology to drive our artistic goals, as artists we seek the reverse, we
aim to enable our artistic vision to define the tools we use. Within this context, we
will survey the main components of design and investigate how going from an
additive process of constructing 3D objects (which is currently a major component
of university curricula in this discipline) to a subtractive process (as represented by
the new technology discussed in this book) potentially changes what people
design, by freeing them from the constraints of technical limitations. We also
discuss where we see new opening areas for creative process as well as potential
pitfalls for digital artists.
This book concludes with a list of online resources, many of which are free, and
outlines ways of delivering structured content for those interested in learning and
participating in building 3D objects, manipulating them, and using them in 3D
visualizations or even outputting them to 3D printers.

1.2 An Introduction to Photogrammetry

Photogrammetry is the science of extracting reliable measurements from two-

dimensional (2D) images, usually photographic. This field of study has a history
which connects it to many other fields and disciplines. Some of the disciplines
1.2 An Introduction to Photogrammetry 3

include: optics, projective geometry, remote sensing, and more recently computer
vision. From a technical standpoint, we are mainly focused on the output of three-
dimensional (3D) meshes generated from a workflow utilizing a more sophisti-
cated technique of photogrammetry known as stereophotogrammetry. Stereopho-
togrammetry involves estimating 3D coordinates of an object by comparing
multiple photographic images taken from different positions. From these multiple
images, a ray (or line) can be calculated to a 3D point(s) on an object.
More about the history and basics of photogrammetry field will be discussed in
Chap. 2.

1.3 Computer Vision

This field is concerned with the process of acquiring, processing, and analyzing
images with an overall goal being to create ‘‘useful’’ information. Computer
understanding of images in a way that can produce multiple, usable, different types
of information can then be applied in many ways. Some applications are similar to
vision-related tasks which humans currently perform, such as inspecting products
on an assembly line; other tasks transcend human tasks by ‘‘seeing’’ beyond the
light spectrum visible to humans, X-rays, and infrared. Similar to photogrammetry,
this field is at the heart of multiple other emerging technologies; artificial intel-
ligence, robotics, machine learning, psychology, and many more traditional fields
such as mathematics and physics. The discussion in this book of the computer
vision field will be constrained to its connection to photogrammetry. However,
some discussion will be given to speculation of ways where emerging technologies
being developed by computer vision might affect 3D design in the future.

1.4 Surface Topology and Retopologizing Workflows

3D computer graphics for film and video games use surface modeling techniques
to generate 3D ‘‘meshes.’’ The 3D model is comprised of a topology of connected
points that connect to form the geometry which describes the surface of the object
(Fig. 1.1). Surface modeling is mainly concerned with describing the 3D object for
use in image creation. Depending on the complexity of the surface, a high amount
of detailed geometry may be necessary. The usual process of surface 3D modeling
entails building from a simple form and adding complexity and details. In this
book, we will address several methods for building 3D objects as well as define
surface versus solid modeling. In order to find a balance between an over complex
geometric surface and one that does not provide enough detail, ‘‘retopologizing’’ is
a process of optimizing 3D mesh geometry by rebuilding it (Fig. 1.2). The number
of polygons is reduced over 200 %, while the overall form remains recognizable,
reducing resources needed by the computer.
4 1 Introduction

Fig. 1.1 3D Mesh. Autodesk screen shots reprinted with the permission of Autodesk, Inc

Fig. 1.2 Retopologization. Autodesk screen shot (Maya) reprinted with the permission of
Autodesk, Inc

1.5 Design Process

While many designers work to avoid the ‘‘medium’’ becoming the message, the
technical and creative process for planning and creating can be highly influenced
by the tools used. In this book, we will discuss many of the creative avenues and
design possibilities that are opened by combining new tools in ways that are just
beginning to be explored. Many of the elements and principles of design connect
in various ways to the process of 3D form as well as perception of objects.
Chapter 2
Technology Leaders and Adopters

Abstract This chapter gives an overview of the nature of technology and

investigates many of the other technologies which are combined for photogram-
metry to emerge in its current state. From general technology theory to a brief
history of photogrammetry, it investigates some of the technologies which
incorporate photogrammetry technologies and are also connected to the fields of
design and entertainment: Microsoft Kinect, photosynth, Weta Digital: Photo-
space, Autodesk 123D Catch. This chapter concludes by investigating some of the
current challenges and limitations.

Keywords Major technology 

Convergence 
Future studies Technology 
  
cycles Art and design Photogrammetry history Nature of technology 

Photogrammetry Microsoft Kinect  
Photosynth Kinect fusion What a 
  
digital Photos space Autodesk 123D catch Challenges Limitations 

2.1 The Nature of Technological Spread

Technology has evolved in alongside human societies for the last several thousand
years. While multiple definitions exist, a combination of its qualities shows it to be
a continuously evolving process of harnessing and amplifying sources of power
which were originally part of the natural world in the form of tools or processes
which serve to increase human abilities to fulfill our needs and desires [1]. Two
components of the nature of technology are being addressed here: the nature of
technology and accelerating change.
The nature of technology reflects similarities to the evolutionary nature of
biological organisms. There are evolutionary cycles in technology similar to
biological development; many new technologies are tested, and the most useful
combinations move ahead. The combinations often start off somewhat weak (birth)
but then (grow) strengthen and optimize (mature) over time. As the technologies
reach maturity, there is often a need to combine with additional technologies in
order to spawn a new cycle in order to avoid disuse (decline).

S. Foster and D. Halbstein, Integrating 3D Modeling, Photogrammetry and Design, 5

SpringerBriefs in Computer Science, DOI: 10.1007/978-1-4471-6329-9_2,
 The Author(s) 2014
6 2 Technology Leaders and Adopters

Technology Speed of Adoption

45
40
35
30
25
20
15
10
5
0

years to reach adoption 10 % penetration

years to reach adoption 10 to 40 % penetration

Fig. 2.1 Mass use of inventions [Technology Review, ITU, New York Times, WSJ, US Census
Bureau]

The field of future studies has described accelerating change as the perception of
increased numbers of technological changes over the last 2,000 years of history.
There has been a growing amount of statistical research showing a logarithmic trend
of major technological change, combined with a reduction in the time a society takes
to adopt the technology. Figure 2.1 presents several recent technologies and their
speed of adoption [2]. A simple but controversial view connects Moore’s Law, which
describes the logarithmic growth of transistors in microprocessors, to accelerating
technological progress [3]. While there is some controversy as to exact formula,
others work to present more accurate ones [4]. Some of the factors linked for new
technology adoption appear to be costly and technically difficult to use.
The result of the accelerating change has been perceived as ‘‘disruptions’’ to
many fields, which had been stable. The disruptive technologies have at times
completely destroyed (or radically changed) whole industries.
Over the last 20 years, 3D computer graphics has been a rapidly moving,
growing, and evolving field. Many of the technical challenges within the field have
been solved. The creation of 3D objects has been in high demand for use in
multiple fields, from visual effects, animated films, architecture, video games, etc.
Creating those objects has been done by the briefly described process of 3D
modeling. Over time, those tools have improved and an industry has developed
around building necessary 3D objects. While there have been incremental
improvements in the tools and processes used for creating 3D objects, the field of
3D modeling had a similarity to skilled craftsmen’s guild, due to the mix of high
technical skill mixed with artistic knowledge. Given the pace of technology, it
seemed inevitable for this field to have a disruptive change. This change is in the
2.1 The Nature of Technological Spread 7

Technology Approximate Dates

Photography & plane tables 1850-1900
Stereoplotters & Airplanes 1901-1950
Computers & Mathematical Models 1951-1971
Digital & Computer Vision 1972-Present

Fig. 2.2 Stages of photogrammetry development

integration of photogrammetry (using real images and video technology) and

procedural object creation (objects made by the use of pre-programmed algo-
rithms). In this book, we will be focusing on the emerging workflow that involves
photogrammetry and retopologization.

2.2 A Further Introduction and History

of Photogrammetry

Photogrammetry relies on gathering data from images that can be analyzed to

extract and yield further information. It has an extensive history, with origins
connected to mathematics, optics, and geometry, which can be traced as far back as
1492, with DiVinci describing the principles of optical perspectivity [5]. Over time,
its first broad application was to aid in the creation of highly detailed cartography.
As the technology progressed, the data that photogrammetry and remote sensing
equipment are capable of acquiring have expanded to physical, temporal, and
semantic types of information ([8], p. 4). The field, however, still has strong con-
nection to its origins. Photogrammetry falls under the field of ‘‘remote sensing.’’
This is because while it is doing advanced calculations and analytic measurement of
objects, there is no physical contact during analysis.
In the scientific fields, there are sometimes additional categorizations which
place the remote sensing aspect of photogrammetry as information only from
imagery which is received from satellites (images from above the earth). For the
sake of this book, we will mostly discuss photogrammetry as it is being applied to
the creation of 3D objects and will not be addressing the scientific and mathe-
matical components of remote sensing.
The history of photogrammetry can be separated into four close differing stages
based on improvements in technology. The stages, titles, and dates vary slightly by
different scholars ([6, 7], p. 282) and ( [8], p. 7); however, the general stages of
development can be separated into the following approximated time periods
(Fig. 2.2).
8 2 Technology Leaders and Adopters

Fig. 2.3 Perspective can be generated using Euclidean geometry

2.2.1 Photography and Plane Tables

The current state of photogrammetry required many technological advancements

over the last several hundred years. Mathematics and optics were critical initial
components, since photogrammetry relies on the scientific analysis of photo-
graphs. As previously mentioned, the history of photogrammetry in the West can
be traced back to as far as the fifteenth century [10] (Italian Renaissance). Rules of
linear perspective can be derived from the basic Euclidean geometry. To this day,
artists use underlying principles of Euclidean geometry within their drawings to
create realistic 3D representations of the world. The translation from real-world
3D image, can be done ‘‘by eye’’ seen from real life, or using more exact processes
such as camera obscura images projected onto a grid, or more recently images
taken with a digital camera which can then be traced. The goal of photogrammetry
is to do the opposite: to use mathematics to extract information from images.
Photography and plane table analysis was performed on images taken from high
rooftops, from hills, or from balloons or even using kites and pigeons. Once these
images were taken, linear perspective combined with simple math was able to
manually compare the known heights in order to derive additional highs of objects
within the image (Fig. 2.3).
2.2 A Further Introduction and History of Photogrammetry 9

2.2.2 Stereoplotters and Airplanes

Stereoplotters are stereoscopic images combined with technologies for automating

the measurements of objects from two overlapping and optically corrected images.
There were multiple inventions (floating mark, stereocomparator, stereoauto-
graphs, serial-photo aerial camera, etc.), which began to automate height analysis
of images in the early twentieth century. By the 1930s, advances and refinements
of these technologies as well as improvements in aviation resulted in the use of
aerial stereophotography becoming the main method of map making.

2.2.3 Computers and Mathematical Models

The further refinement of analog devices for stereoplotting and photography was
combined with the rapid calculations which are able to be performed by computers
allowed for even greater accuracy. Specifically, some of the more complex
‘‘algorithms for orientation and triangulation…’’ were developed, dramatically
accelerating results and accuracy.

2.2.4 Digital and Computer Vision

The first readable images of Earth sent from orbit were from Landsat in 1972. The
transition from partial to near-complete automation has taken another 40 years.
Digital photography, further improvements in algorithms, as well as the ability to
record in multiple formats allow for greater interpretations. ‘‘Pan-chromic imag-
ery, near-infrared, and color’’ could all be taken simultaneously using digital,
removing the need for chemical development of images and scanning. Also, many
of the physical plotting devices such as the stereoplotters have been replaced by
lIght and range detecting (LIDAR) remote sensing technology that uses lasers
instead of analglyphic stereoimagery to map depth and distance information.

2.2.5 Looking Ahead

When it started, the field was used for extremely large mapping applications, but in
the near future, a microscopy technique using stereophotogrammetry is being
planned for the analysis and creation of ultrasmall components [9]. The technology
is also being used to look and map areas farther away. Currently LIDAR, a remote
sensing technology, which combines laser and radar technology, is mapping the
moon. Many of the photogrammetry techniques are beginning to move into other
10 2 Technology Leaders and Adopters

research fields as well as become accessible to mainstream population for use in

creative, educational, or hobbyist application.

2.2.6 The Cycle Between Art and Science

A current list [6] of applications of photogrammetry is

• Mapping and acquisition of geoinformation
• Documentation
• Monument preservation and architecture
• Aerial, terrestrial, and underwater archaeology
• Monitoring earth surface and building deformations
• Civil engineering studies
• Automobile, aeronautical, and nautical industries
• Dental, orthopedic medicine and biomechanics
• Forensic applications.
This book specifically deals with some of the ways a new category is being
added to this list:
The fields of art and entertainment, specifically how new technologies integrate
into college 3D art and design education. Interestingly, in the past, artists adopted
mathematics and linear perspective to derive 2D images that look 3D, and now, the
current field of 3D computer graphics and digital design is once again connecting
with the fields of science (photogrammetry) and full cycle, by adopting photo-
grammetry technology to derive 3D objects into artwork. The field of 3D computer
graphics has begun using photogrammetry information to rapidly generate 3D
models, in order to make visual images based on real life. While there have been
various market leaders in industry who have sparsely applied these technologies,
the factors of low cost and mass accessibility are laying the groundwork for broad
adoption.
As the cycle continues of art, science, and technology, we will now introduce
several of the applications created by the entertainment industry for use in research
and application.

2.3 Photogrammetry Resources Connected

to the Entertainment Industry

Photogrammetry for deriving 3D assets (and analysis) has existed in many forms
and in various products within other fields. Within the design and technology fields,
the technology has arrived relatively recently. In the next few sections, we will focus
on a few technologies developed by Microsoft, Autodesk, and Weta Digital , due to
2.3 Photogrammetry Resources Connected to the Entertainment Industry 11

their connections to the media and design fields. However, we have also compiled a
list of additional photogrammetry resources in the appendix of this book.

2.4 Microsoft Kinect

Due to its industry-leading research laboratories in computer vision combined with

its Xbox and Kinect video game platform, Microsoft has been a market leader with
combining photogrammetry and computer vision technology. The Kinect, a
breakthrough device, was launched as a peripheral to the Xbox 360 game system
in November 2010. However, Microsoft (and others) launched additional software
development kit (SDKs) which allowed the device to be used in ways beyond its
intended use as a next-generation game controller.
The Kinect device combines a standard camera, infrared projector, and
microphone. By analyzing visual information generated by the camera and depth
information generated by the infrared sensor, it creates a large set of ways to
interact with computers without a mouse or keyboard. Gesture-based interactive
technology in the Kinect was acquired by Microsoft from a range camera devel-
oped by PrimeSense, an Israeli 3D sensing company. Beyond 3D sensing, the new
3D and gestural interface has had a large impact on the field of user interface
design as well (Fig. 2.4).
In May 2013, Xbox One was announced by Microsoft, and the upgraded
‘‘Kinect One’’ sensor is due out with the release of the gaming system late 2013.
With this product will come further improvements to the Kinect Fusion
technology.
The Kinect One hardware will feature a big step forward in technology. It
incorporates greater gesture recognition, up to six people tracking (via facial
recognition) and a full HD 1080P resolution (as compared to the original sensor
which offers only standard definition). It also promises to deliver a nearly latency-
free workflow.
Probably, the most significant feature listed above in respect to this book is the
nearly 4X increase in resolution and speed of capture. 3D computer models, which
will be generated by the Kinect One, will have greater detail for more accuracy.
This additional fidelity will make capturing 3D models much more accurate.

2.5 Microsoft (CV) Kinect Fusion

First presented at SIGGRAPH 2011, this technology is a series of software

libraries that connect to the Kinect hardware. Kinect Fusion was developed at the
Microsoft Computer Vision Laboratories. One component of the Kinect Fusion
technology allows for high-quality, 3D renderings of environments and people in
real time. Most current photogrammetry to 3D mesh workflows involves taking
12 2 Technology Leaders and Adopters

Fig. 2.4 Microsoft Kinect sensor

multiple photographs and then having comparatively analyzed in order to

extrapolate 3D point cloud volumes, from which 3D meshes are derived. Kinect
Fusion develops 3D meshes in real time. We anticipate faster and more accurate
detailed meshes that will rapidly emerge as the technology further evolves. Other
products have also incorporated 3D the Kinect as a platform for scanning tech-
nologies (for example, [Link]
In March 2013, Microsoft released a significant update to the Kinect SDK. This
update includes the Kinect Fusion technology in the SDK feature set (Microsoft
Kinect SDK) [11].
• Real-time, GPU-assisted 3D object and scene reconstruction by using the Kinect
for Windows sensor
• Ability to infer relative sensor position and orientation from a 3D scene for
augmented reality application
• Advanced algorithms that are powerful enough for large sensor movements and
scene changes during scanning
• Direct X11 compatible graphics cards supported
• AMD Radeon 7950 and NVidia GTX560 have been validated to run at inter-
active rates
• Kinect Fusion Studio and samples demonstrate 3D scanning capabilities
• Non-real time CPU mode for non-interactive rate scenarios.

2.6 Photosynth

This product allows for two different types of visualization based on photogram-
metry technologies. It was developed in collaboration between University of
Washington and Microsoft research laboratories. There are currently two functions
within Photosynth, Synths, and panoramas. The ‘‘synth’’ uses multiple images that
are analyzed in order to generate a three-dimensional image of the space. The
second is a panorama which allows the user to take multiple pictures in a three-
dimensional space. These pictures are then processed. The processing analyzes the
images and combines them in a process called stitching. Figure 2.5 demonstrates
2.6 Photosynth 13

Fig. 2.5 Stitched image

the output of a stitched image, warped in a way so that it can be mapped to a

spherical interactive QuickTime VR environment. This image was created using
Photosynth, a product developed by Microsoft computer vision research
laboratories.
There are a large number of Photosynth experiences is available on the
Microsoft Web site: [Link] In addition to giving a
description and showing the date created, many of them are geotagged (GPS info
embedded in the metadata) for easy connection to map information.

2.7 Weta Digital: Photospace

In their presentation and published paper [12], they discuss the integration of a
photogrammetry-based workflow for physical props for digitization and use in
visual effects sequences. In their paper, they discuss previous workflows which use
photographic reference, 3D modeling and 3D scanners. They describe the pho-
togrammetry (vision-based) approach as combining the best features of multiple
approaches. They described the workflow process as having three parts: (1) the
capture session, (2) photogrammetry processing session, and (3) reference gen-
eration session. Images are captured, they are processed, and then, the three-
dimensional models that are generated are handed to 3D artists for retopologizing
or remodeling in order to make them efficient for use in the rendering pipeline.
They also discussed many of the challenges and limitations connected to other
photogrammetry-based workflows (see below).
14 2 Technology Leaders and Adopters

2.8 Autodesk 123D Catch

123D Catch is a free photogrammetry software tool created by Autodesk Corpo-

ration. It has been produced for online use, in mobile device ‘‘app’’ form as well as
standard desktop. The simplicity and quality of output generated by this software
have made it very popular. Creating 3D meshes from photography is a simple
process. The user takes 20–40 images of an object (with a maximum of 70) and
feeds them to the program, which develops a 3D mesh from an analysis of the
images. The software finds and matches common features in order to construct a
3D mesh from the identified feature sets.

2.9 Challenges and Current Limitations

While there have been many improvements in both speed and accuracy over the
last 20 years, there are still hurdles to overcome. Many of the algorithms that are
used for translating computer vision resources into 3D mesh objects can be easily
broken, which generates a failed, problematic, or incomplete 3D mesh. These
limitations listed below are fairly standard among the previously listed and most
other current photogrammetry technologies. However, most softwares list these as
heuristics to overcome many of the issues with current programs, because knowing
the use guidelines can help overcome failed or problematic meshes.

2.9.1 Occlusions and Number of Photographs Necessary

Occlusions can cause problems. An occlusion is when and unwanted object comes
between the camera and the target photogrammetry object. In small scenes, this
can be things like imaging a human who wears glasses (also see below regarding
transparent objects). An example of a similar problem for a large exterior scene
would be a tree in front of a house. Most semicomplex objects have moderate self-
occlusion, especially if they have multiple folding or complex parts. However,
with enough images from the correct angle, many current photogrammetry algo-
rithms are robust enough to solve for the objects. For heavy occlusion, images
every 5–10, overlapping as much as 50 %, might be necessary. For object with
little or no occlusion, images ever 20 or more can be used for excellent results.

2.9.2 Photographs Need Features

The algorithm searches for parallax shifts between known features within multiple
images. Taking images of blank walls or large empty/non-focused areas will cause
2.9 Challenges and Current Limitations 15

the algorithm to fail. Patterns, strong lines, and differentiated features are what the
algorithm looks for when tracking. Sometimes adding features with tape, stickers,
or draped cloth can get better results. Similar to objects with no features (like the
previously discussed blank wall), repetition of features can cause a similar con-
fusion for the algorithm when trying to match feature sets.

2.9.3 No Transparent, Reflective, or Glossy Subjects

This creates certain difficulties for many objects. Even semitransparent objects can
pose problems. When applicable, spray-painting them with matte finish can
remove any transparent or shiny qualities. However, for large objects, for example
buildings with reflective and semitransparent windows, they can pose difficulties.

2.9.4 Subjects Cannot Move During the Image Capture

Process

In order for the algorithm to match the feature sets, object cannot significantly
move.

2.9.5 Consistent Lighting

This not only means not changing light sources, but the algorithm works best
without strong directional sources of light. Meaning outdoors, cloudy days
(ambient light) would create the best lighting conditions. Indoors might require
diffusers to create as ambient and consistent lighting conditions as possible.
Additionally, the use of a flash will cause problems due to the fact it will create a
unique (directional) lighting situation for each image.
For further reference listing the limitations of the 123D Catch, please follow
this link to the following video: [Link]
w#at=64.

References

1. Arthur, W.B.: The Nature of Technology: What it is and How it Evolves. Free Press, New
York (2009)
2. Degusta, M.: Are Smart Phones Spreading Faster than any Technology in Human History?
Technology Review: Published by MIT [Link]
(2012)
16 2 Technology Leaders and Adopters

3. [Link]
4. Hamasher, K.: Accelerating Changes in our Epoch and the Role of Time-Horizons,
pp. 205–215. Springer Netherlands, Heidelberg (2006)
5. Ghosh, S.K.: Fundamentals of Computational Photogrammetry. Concept Publishing
Company, New Delhi (2005)
6. Redweik, P.: Photogrammetry. In: Xu, G. (ed.) Sciences of Geodesy—II: Innovations and
Future Developments (Chap. 4), pp. 133–183. Springer, Heidelberg (2013). ISBN 3-642-
27999-6, 978-3-642-27999-7
7. Konecny, G.: Geoinformation: Remote Sensing Photogrammetry and Geographic
Information Systems, 2nd edn, p. 280. CRC Press, Boca Raton (2002)
8. Schenk, T.: Introduction to photogrammetry. Department of Civil and Environment
Engineering and Geodetic Science, The Ohio State University, Columbus (2005)
9. Boutin, C.: Microscopy Technique could Help Computer Industry Develop 3D Components.
NIST Tech Beat. [Link] June 2013
10. Gruner, H.: Photogrammetry: 1776–1976. Photogram. Eng. Remote Sens. 43(5), 569–574
(1977)
11. Microsoft Kinect SDK [Link]
12. Bhat, P., Burke, S.: Weta digital PhotoSpace: a vision based approach for digitizing props,
ACM SIGGRAPH 2011 Talks Article No. 1, New York (2011)
Chapter 3
Conceptualizing Basic 3D Modeling
Workflows

Abstract This chapter traces the history of the development of the methods and
practices used in 3D modeling. It also draws parallels between this development
and its relationship to sculpture and design. In its various sections, a detailed
discussion of the various approaches is provided; illustrating the changes in
method as hardware and software has become more sophisticated over the years. It
demonstrates how an artist uses existing and new technology to achieve aesthetic
goals and how technology has been influenced by this achievement.

Keywords Additive 
Subtractive Modeling  Topology 
Workflows  
  
NURBS Patch modeling Surface parameterization Polygons Box model- 
    
ing Edge flow Normal maps UV layout Langer Retopology Maya 3D  
  
Coat Mudbox Zbrush Digital sculpture

3.1 Compartmentalizing the Learning Curve

In teaching 3D modeling in general, and more specifically organic 3D modeling in

a post-secondary environment, we find that there are two distinct areas of ability
among our students, the aesthetic and the technical. Some of our students have
developed skills in one area over another—as a result of personal interest, program
availability in their high school, or other circumstantial or motivational reasons.
We can, of course, rate these skills using a very general scale of low–high in
either area, which creates the following metric:

S. Foster and D. Halbstein, Integrating 3D Modeling, Photogrammetry and Design, 17

SpringerBriefs in Computer Science, DOI: 10.1007/978-1-4471-6329-9_3,
 The Author(s) 2014
18 3 Conceptualizing Basic 3D Modeling Workflows

Low Technical, Low Technical,

Low Aesthetic High Aesthetic

High Technical, High Technical,

Low Aesthetic High Aesthetic

Of course, we want to build on whatever skills the student has in either area,
while creating a foundation in the complementary area. Our goal is to develop our
students from both sides, concurrently providing a highly developed aesthetic,
strong sense of design, understanding of advanced principles of 2D and 3D design,
and a well-cultivated understanding of anatomy on one side, and a mastery of the
technology from both a procedural and conceptual approach, as represented by the
following Venn diagram:

Strong Strength Strong

in
Technical Aesthetic
Both
Ability DIsciplines
Ability

Separation

In any kind of process that uses technology or technical methodologies in the

creation of art, the artist has to master the craft as a necessary means to an end.
Wood grain, metallurgy, chemistry—all are part of the technical training for
designers, sculptors, ceramicists. Without an understanding of the technology and
craftsmanship involved in the structural integrity of a piece, the best designs will
fall apart, or will be unusable. 3D modeling is no different.
3.1 Compartmentalizing the Learning Curve 19

In many disciplines, students are asked to design on paper first, without regard
to the details of material or structural integrity. The approach taken by many
instructors is ‘‘You design it first—then we will figure out how to put it together.’’
Because of the laws of physics, material science, structural engineering—this is
frequently the only viable approach.
As in any other discipline, it has been difficult if not outright impossible to
separate the technological processes from the aesthetic development while using
3D modeling software. Understanding of edge placement, ‘‘flow’’ of polygons or
patches, limitations on acceptable geometric forms (‘‘quads’’ vs. triangles, non-
manifold geometry, edge quality, etc.) is vital to the most basic kind of con-
struction and presents a serious impediment to the creation of 3D art.
However, these barriers have come down almost entirely, thanks to the parallel
development of three distinct types of software.
The first of these is digital sculpture software. The most commonly used digital
sculpture applications today are Pixologic ‘‘ZBrush,’’ Autodesk ‘‘Mudbox,’’ and
‘‘3D-Coat,’’ but there are at least a dozen more that are available as standalone
packages, plug-ins, or as integrated modules in well-known professional and
consumer while each package offers unique toolsets and focuses on different
workflows and capabilities, all have the capability of ‘‘voxel sculpting.’’ Voxel
sculpting, simply put, is the process of using specific toolsets to push, pull, smooth,
grab, pinch, or otherwise manipulate a digital object as though it were made of a
real-life substance such as clay.
Voxel sculpting represents a departure from traditional modeling methods in
that it can rely entirely on high-polygon meshes for detail. The old rules of clean
topology will always apply in the end, but the methodologies outlined in this
chapter will focus on the ‘‘brute force’’ capabilities of voxel sculpting without
regard to those rules. This brute force approach is used in order to maintain the
separation between purely interactive sculpture in a very traditional sense and the
attention to the topology details which will be addressed in a second pass.
The second type of software is called ‘‘photogrammetry’’ and would be used
instead of the voxel-sculpting described above. Photogrammetry also produces a
high-density mesh that achieves detail by brute force, increasing density where
additional detail is needed with no regard for model efficiency. Rather than use
interactive sculpting tools, however, the process of photogrammetry uses photo-
graphic data and various triangulation algorithms to determine the geometric
properties of the subject. For our purposes, photogrammetry software enables us to
take a series of photographs of a single object from various points of view and
derive a dense, but highly accurate, 3D mesh from those photographs.
The third type of software, and the one that makes the first two viable modeling
options in the first place, deals with retopology. Retopology is the process of
taking a high-density mesh and tracing it with new polygons in order to reconstruct
the original using less dense, more efficient topology. Using this process on a pre-
sculpted model allows the user to focus entirely on issues of edge placement, edge
flow, and efficient geometry, with no additional regard to the sculptural aesthetics
of the piece.
20 3 Conceptualizing Basic 3D Modeling Workflows

The workflow, therefore, would begin with the acquisition of an accurate,

though overly dense, 3D mesh via interactive sculpture or photogrammetric cap-
ture, followed by a retopology process to create an efficient, lower-density mesh
with accurate and systematically contrived edge placement.
Once the retopologized mesh has been created, it can be reacquired by the
voxel-sculpting software for additional sculptural details, proper UV layout, and
the subsequent generation of normal and displacement maps, which will be dis-
cussed later in this chapter.

3.2 3D Modeling: A General Conceptual Approach

Throughout the short history of 3D modeling, the level of complexity of 3D

models has been directly related to the graphics capability of workstation hard-
ware. This relationship has historically been a limitation, and it is this limitation
that directed the development of commonly accepted modeling methods that we
refer to as ‘‘Additive’’ workflows.
Whether using NURBS or polygonal modeling workflows, the additive methods
all start with a simple form which is built up stepwise line by line, edge by edge,
until the final detailed model is achieved. This method has usually been deemed
necessary and proper, needed in order to prevent a work-in-progress from
becoming unmanageably dense. To this end, particular attention is paid to edge
placement and edge flow, both of which act as character lines necessary to define
large shapes.
For the examples of modeling methods illustrated below, we have chosen the
human head as the subject. The human head requires a substantial amount of
attention to a relatively high degree of detail in order to accurately express the
underlying musculature, the general morphology and the character of the person
being modeled. For models intended for animation, the edge flow is doubly
important, as the placement of these edges will be the lines defining the folds,
wrinkles, and muscle masses as facial expressions change over time.
Regardless of the technological approach used, we believe it is vital to begin
our modeling instruction with a study of the shape and forms of the subject,
including the underlying structures—muscles, tendons, cartilage, and bones. The
layout of these structures, their character lines and flow-lines, will help us to
determine our edge placement when we begin modeling in our 3D software
(Fig. 3.1).
In Fig. 3.2, one student begins by comparing an image of facial anatomy from
an anatomy text to the head she wants to model. Using this as a reference, she then
identifies important ‘‘Landing Zones,’’ or skeletal ridges, volumes, edges, and
masses that contribute to the overall shape of the head. These skeletal structures
are key in defining the origin and insertion locations of various muscle groups and
help to begin the strategy of edge flow.
3.2 3D Modeling: A General Conceptual Approach 21

Fig. 3.1 Muscles of the face

and neck

Fig. 3.2 Underlying

structure of the head and neck

The next step in developing this understanding of facial structure is to continue

the draw-over process. In Figs. 3.3, 3.4, 3.5, and 3.6, another student begins fol-
lowing the various surface areas of the face while referencing medical illustration
imagery.
By going through the exercises and manually drawing in the lines, the student
begins to develop a feel for the shapes, the forms, and the outlines of the structures
that lie underneath the skin and contribute to the shapes, the lines, and the high-
lights on the surface that represent this structure.
Students are then asked to use these underlying forms to develop a strategy for
their edge flow. It is important to understand that at this point, the ultimate
modeling method, whether NURBS, polygons, or subdivision Surfaces, is unim-
portant; the placement of edges for the determination of shapes is the same
throughout.
22 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.3 Surface of the face,

head and neck, showing the
underlying structures

Fig. 3.4 Student overlay

illustration of underlying
structures

Fig. 3.5 Student overlay

illustration showing
underlying bone, muscle, and
structure

Using printed images and tracing paper or digital images and layers in an
imaging program like Photoshop, and with their previous drawings as reference,
students lay out their edges.
3.2 3D Modeling: A General Conceptual Approach 23

Fig. 3.6 Student illustration

of facial bone and muscle

At this point, it makes no difference if the resulting edge flow is targeting a final
output in NURBS, polygons, or subdivision surfaces, the goal is always to put the
edges where they will perform optimally in terms of defining and deforming the
surfaces in the most accurate and efficient way possible.

3.3 NURBS Patch Modeling

NURBS ‘‘patch modeling’’ is a highly technical process, with emphasis placed on

mathematical accuracy and respect for the formula-driven creation of curves and
surfaces. The ability to sculpt accurately with NURBS is severely hampered by
these technical constraints; students learning to model must develop a level of
mastery of the workflow before applying it to any creative endeavor.
NURBS surfaces are defined in one of the two ways, either ‘‘uniform param-
eterization’’ or ‘‘chord-length parameterization.’’ These terms refer to the under-
lying mathematics of the surfaces.
‘‘Uniform’’ parameterization refers to the numeric identification of each isop-
arm in the U or V direction of the surface. Beginning at the origin of the surface,
the isoparms in U and V are numbered sequentially, beginning with zero.
Uniform parameterization can either be expressed by the total number of
isoparms (which define the number of spans) using sequential integers (0, 1, 2, 3
…) or by using evenly spaced decimals within a 0–1 numeric system. In either
case, the position of any given span relative to the world-space measurement of the
surface is disregarded (Figs. 3.7, 3.8).
In contrast, chord-length parameterization does take into account the world-
space measurement of the entire surface when assigning numeric values to the
individual isoparms. The first isoparm is given a numeric value of 0, and the
numeric value of the last isoparm reflects the entire length of the surface in world
24 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.7 Numeric values

assigned to isoparms in a
uniform NURBS surface,
with values arranged as ‘‘0–
#Spans’’

Fig. 3.8 Numeric values in

the U and V directions are
also reflected as min/max
values in Maya’s Attribute
Editor

Fig. 3.9 The isoparms in a

chord-length surface are
assigned scalar values, which
are ratios that represent the
position of the isoparm
relative

Fig. 3.10 Just as is the case

with uniform
parameterization, numeric
values in the U and
V directions are also reflected
as min/max values in Maya’s
Attribute Editor

space. The value assigned to any given interior isoparm is a ratio of that position
relative to the entire surface length in the given direction (Figs. 3.9, 3.10).
Whether uniform or chord length, NURBS patches must be ‘‘square’’; in other
words, the number of patches at the beginning of the U or V direction must be the
same at the end; no holes or detachments are allowed. Single NURBS patches
cannot fork in two directions (Figs. 3.11, 3.12).
3.3 NURBS Patch Modeling 25

Fig. 3.11 This configuration

is impossible in a single
NURBS surface

Fig. 3.12 What you actually

see is two surfaces that are
touching

Fig. 3.13 A hole like this is

also impossible in a single
NURBS surface

It is important to note that when creating complex surfaces, holes may be

absolutely necessary, and the edge flow of models must change direction fre-
quently—which presents an inherent technical problem for the NURBS modeler.
In order to solve this problem, surfaces must be broken up into individual square
patches and stitched together, maintaining continuity between adjacent surfaces
(Figs. 3.13, 3.14).
‘‘Surface continuity’’ refers to the way in which one surface connects to an
adjacent surface. In the field of engineering and industrial design, multiple options
for surface continuity with NURBS models can be achieved. Note that the degrees
of surface continuity mentioned below are sequential; the set of rules defining any
level of continuity contains all the rules of the level beneath it, with additional
factors added.
G-0, or ‘‘positional continuity,’’ requires only that two surfaces share an edge,
that they are tangent to each other. No consideration is given to the quality of that
edge; it is generally assumed to be a hard corner.
26 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.14 What you actually

see is eight surfaces that are
touching

Fig. 3.15 Two positionally

continuous surfaces. Note
that their shared edge is
marked by a hard break in
illumination

G-1, or ‘‘tangent continuity,’’ means that two surfaces share not only an edge,
but a tangent line at the point of positional continuity, as measured at the surface
normal. If the normals are also both perpendicular to the same tangent line, they
are also coincident. Coincident end-normals that reflect positional continuity
produce a transition from one surface to the next with no visual breaks. These
breaks would generally be apparent when light shining on the surface is calculated,
represented by a hard line in a specular highlight or diffuse attribute of the sur-
face’s response to light. Deviation from absolute coincident can be acceptable
based on predetermined tolerance factors and is typically calculated in nonzero
angular measurement between the two normals in question.
For organic modelers, character modelers, or digital sculptors, tangent conti-
nuity is typically the ideal, which will be discussed in more detail later in this
chapter (Figs. 3.15, 3.16).
G-2, or ‘‘curvature continuity’’ is of more importance to engineers. In addition
to a shared tangent at the point of positional continuity, G-2 continuity requires
equivalent rates of curvature for both surfaces as additional spans with increasing
parametric distance from the point of transition are considered.
For the purpose of engineering, industrial design, and manufacturing complex
surfaces, additional levels of surface continuity are possible; however, for the
purpose of this chapter, we will stay focused on G-1, or tangent continuity.
From a practical, rather than theoretical standpoint, the test for tangent conti-
nuity is relatively simple, and this kind of continuity can be achieved in a number
of ways.
3.3 NURBS Patch Modeling 27

Fig. 3.16 Two tangent-

continuous surfaces. There is
no break in the illumination
where the two surfaces join

Fig. 3.17 Because the end

vertices of curve 1 and curve
2 are coincident, these two
curves are positionally
continuous. However, the
tangent line cannot intersect
the coincident vertices and
both adjacent vertices, so no
tangent continuity

Fig. 3.18 The adjacent

vertices have been moved so
they now intersect the same
tangent line that passes
through the coincident
vertices. These curves are
tangent continuous

For two curves to be tangent continuous, the end vertices (CVs) must be
coincidental (positional continuity). If a straight line can be drawn from the vertex
adjacent to the end on one curve to the vertex adjacent to the end on the other, and
this line intersects the coincidental end vertices, then the two curves are tangent
continuous (Figs. 3.17, 3.18).
By extension, the same test is given to surfaces. While it is not a mathematical
requirement that adjacent surfaces have the same number of spans, it is a practical
one. For purposes of determining and achieving tangent continuity between two
surfaces, the coincident vertices at the end of each span and their adjacent vertices
must pass the same test, above (Fig. 3.19).
28 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.19 Vertices adjacent

to the coincident vertices are
on the same tangent line (blue
lines); therefore, these two
surfaces are tangent
continuous

3.3.1 Methodologies

There are essentially three methods of achieving tangent continuity between two
surfaces: attach/detach, manual alignment, and automatic stitching.

3.3.2 Attach/Detach

Most NURBS modeling software packages have an ‘‘attach’’ function, where two
surfaces can be made into one. A common option within this function is ‘‘blend,’’
which creates one continuous, smooth surface. When this surface is then detached,
the continuity is inherited, and an inspection of the adjacent vertices along the
common edge reveals the placement of ‘‘tangent vertices,’’ with each set of four
passing the tangent test described above.
A prerequisite to this attach is that the surfaces have the same parameterization
type (either ‘‘uniform’’ or ‘‘chord length’’) and, if they are uniform, follow the
same isoparm numbering convention (‘‘0–1’’ or ‘‘0–#of spans’’). While any of
these will satisfy these prerequisites, it is generally accepted that ‘‘uniform, 0–1’’ is
the preferred convention. Therefore, if two adjacent surfaces are of different
parameter types, or follow different numerical conventions, they must be rebuilt
prior to the attaching.
Further, NURBS patches should be aligned in the same direction—with their
U and V directions parallel and their normals facing the same direction. The
beginning of one surface (Isoparm #0) should attach to the end of the other surface
(Isoparm #1). If an inspection of these surface parameters reveals non-alignment,
the surfaces must be reversed in either or both directions, and/or the directions
must be swapped.
Finally, when two surfaces are attached, the numerical values assigned to the
isoparms will change. When they are detached, these values do not revert, so the
surface must again be rebuilt.
As complex as it may appear here, the attach/detach method of surface conti-
nuity is the simplest process to follow when creating organic models using
NURBS. The workflow becomes rote: rebuild, reverse, attach, detach, rebuild,
and delete history (Figs. 3.20, 3.21).
3.3 NURBS Patch Modeling 29

Fig. 3.20 A complex

NURBS model must be
pieced together with square
patches that meet all of the
conditions described in this
chapter

Fig. 3.21 If any surface in

the model maintains tangent
continuity with respect to its
adjacent surfaces, the result
will be a smooth model, with
no breaks in surface
highlight, giving the
appearance of a single

Once a model has been pieced together using the above methods, after a fair
amount of cleanup and reworking, inconsistencies, pinch-points, and non-tangent-
continuous surfaces will be discovered. Fortunately, these areas are not too pre-
valent and can usually be fixed using manual methods of achieving continuity.
Usually, these methods involve drawing a straight line between two vertices,
extending that line and snapping the remaining vertices into place using a curve
snap or point snap function. Several users have written simple scripts to simplify
this process.
The other method of achieving tangency is automatic and software-specific.
Usually, this is a tool called ‘‘Align Surfaces’’ or ‘‘Stitch’’ or ‘‘Global Stitch.’’ The
tool is simply an automated way of performing all the manual tests outlined above
(positional continuity, parametric consistency and alignment, and normal-angle
coincidence) across multiple surfaces, making a series of corrections based on
30 3 Conceptualizing Basic 3D Modeling Workflows

preset tolerances and procedural algorithms. From a practical point of view, it is

generally a bad idea to depend too heavily on these automated methods initially;
they are best used in small areas or in areas where corrections are few and rela-
tively minor.
With all the technical requirements and complexities inherent in NURBS patch
modeling described above, it is important to note that we have not yet addressed
the most important aspects of organic modeling in general—which are edge flow
and sculptural aesthetics. These important considerations must also be taken into
account during the process of rebuild, reverse, attach, detach, rebuild, and delete
history. It is for this reason that NURBS modeling (as a final output) has fallen out
of favor among studios. A solid mix of technical ability and aesthetic expression is
vital to this style of modeling and is contrary to the compartmentalization model
that photogrammetry allows.
It is also difficult and computationally expensive to maintain tangent continuity
among surfaces that deform during animation. With each deformation (facial
expressions, lip-synching, muscle or other organic deformation), tangent conti-
nuity must be recalculated and adjusted—typically through adherence to some
form of construction history algorithm. This, too, is a major drawback to the
NURBS modeling pipeline.
This is not to say that there is not a place for NURBS modeling, either in the
process of teaching modeling or within a production pipeline. NURBS tools are
very robust and, sometimes, are the preferred method for achieving certain shapes
or volumes. From an educational perspective, the discipline required of a NURBS
work-flow helps to develop a naturally clean, efficient approach for students, and
builds an inherent tendency to seek accurate and efficient edge flow that is vital to
the aesthetics of a model, but not necessarily required for its technical success.

3.4 Box Modeling

Box modeling is the method that is most typically used and taught as the tradi-
tional method for polygonal organic modeling. As the name implies, box modeling
begins with a simple primitive shape, typically a cube, although in many cases, a
model might begin with a sphere or cylinder.
The goal with this kind of model is to define the size and proportion of the
figure while exercising restraint in the process of adding edges until they are
needed. For this reason, an understanding of the major shapes that make up the
human form is vital and plays a big part in the students’ decisions as to where to
introduce new edges.
In Fig. 3.22, the model was started with a cube that was scaled to the pro-
portions of a human torso. In Fig. 3.23, the first edge introduced is the waistline,
which helps to differentiate the shape of the upper body (chest, rib cage) from the
lower (pelvic girdle). The centerline is introduced in Fig. 3.24 with two adjacent
edges which will differentiate the groin from the legs, which will be extruded or
3.4 Box Modeling 31

Fig. 3.22 A simple cube is

the beginning of a human
torso

Fig. 3.23 Major character

lines are inserted first

Fig. 3.24 Additional lines

are inserted in order to begin
rough-in of general shape
32 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.25 Lines are added to

define hips, rib-cage,
shoulder area

Fig. 3.26 Lines are averaged

to achieve more roundness

attached later. Vertices are moved up along the outer pelvic area to define the dish
shape of the pelvis and to simulate the natural lines of the abdomen resting within.
With the placement of each edge, the model is evaluated as to its likeness to the
reference material, including the anatomical references and draw-overs previously
executed. In every case, the attempt is made to refine the shape as much as possible
without the addition of new edges.
As the model progresses, the existing edges are manipulated into positions
where they can define specific structures beneath the skin. In Fig. 3.25, the
waistline is moved up to define the lower part of the rib cage, while additional
edges are added to create the pelvic structure and the chest.
In Fig. 3.26, no additional edges have been added but the existing vertices have
been averaged, or softened, to relax the cube shape and make it appear more
cylindrical.
In Fig. 3.27, additional edges were added to the chest, which are then used as
extrusion points to create the deltoid and the upper arm.
In this way, shapes are visualized and edges are placed to achieve these shapes.
Attention is continually paid to the original draw-overs that identify the larger
shapes, than the smaller ones—from planes to masses to individual muscles and
bones (Figs. 3.28, 3.29, 3.30, 3.31).
3.4 Box Modeling 33

Fig. 3.27 A reasonably accurate shape can be attained with very few lines

Fig. 3.28 Student draw-over illustrations, showing major ‘‘landing zones’’ for polygon edge-
flow strategy

As the figure begins to take shape, the edges begin to create rows of individual
polygons, which can then be shaped into the forms observed and drawn in the
earlier exercises.

3.4.1 Langer’s Lines

Another consideration is the natural flow of tension lines along the skin, known as
‘‘Langer’s iines.’’ Karl Langer was a nineteenth-century surgeon who produced a
series of diagrams depicting lines of cleavage in the skin. Anatomically, these lines
correspond to the natural orientation of collagen fibers in the dermis and are an
important consideration for surgeons in deciding the placement and orientation of
34 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.29 Student draw-over

illustrations, showing major
‘‘landing zones’’ for polygon
edge-flow strategy

Fig. 3.30 Additional

strategy for polygon layout

surgical incisions. If an incision follows Langer’s lines, the scar will align with the
natural tension lines of the skin and will heal in a more camouflaged state.
Collagen is what dictates the shape and direction of wrinkles in the skin—those formed by
facial expressions, body position and aging. Langer, Karl (1861)

For artists, these lines are also important. When determining proper edge
placement for a model regardless of the method being used, if the edge flow
describes the underlying shapes and creates a matrix for realistic deformation,
more realism can be achieved (Figs. 3.32, 3.33).
3.4 Box Modeling 35

Fig. 3.31 Finished layout,

following the strategies
developed in the draw-overs

Fig. 3.32 Langer’s Lines of

the torso and upper
appendages

Fig. 3.33 Langer’s lines

revealed through the flow-
lines of the skin
36 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.34 Langer’s lines of

the face

Fig. 3.35 Facial wrinkles

showing Langer’s lines

Whether designing body or a face, the artist must take into consideration the
structure underneath the skin, and the structure of the skin itself when determining
the proper placement of edges, lines, and polygons. By faithful attention to these
anatomical details, the modeler can achieve a higher degree of realism (Figs. 3.34,
3.35).
All of these details present a very practical problem to the student who is first
learning how to create an organic model. For many, as described in previous
chapters, the technical and aesthetic challenges are often mutually exclusive. As
educators, we are frequently faced with the dilemma that ‘‘you have to know
everything before you can learn anything.’’ The process of organic modeling,
therefore, is frequently one of trial and error—discarding early versions and
starting over as new concepts are learned on both sides.
3.5 Normal Mapping, Displacement Mapping, and UV Layout 37

3.5 Normal Mapping, Displacement Mapping, and UV

Layout

In the world of 3D graphics, the notion of ‘‘location’’ can be variable and must be
clearly defined within the context of the specific tasks we are undertaking.
Picture, for the moment, the location of your home. Within a certain context,
you can regard this location as a constant—you always know where your home is,
you can find it on a map and in the world. It never moves. However, your home is
on the planet Earth, which revolves on its own axis, orbits the sun, and is part of a
solar system that traverses the universe. In a broader system of coordinates, those
that define the universe for example, the location of your home is constantly
changing.
This is an example of the difference between ‘‘world space’’ and ‘‘local space.’’
In a 3D modeling system, we typically define the Cartesian grid system as the
universe where everything takes place. Position in this universe is constant; the X,
Y, and Z axes define our world. We can keep track of the position of our objects
and the variations of their surfaces by defining the X, Y, and Z positions of the
individual vertices in our calibrated universe. This is ‘‘world space.’’
However, we also have a need to define the position of a given point relative to
a surface. If I want to place a beauty mark on the cheek of a model of Marilyn
Monroe, I want that beauty mark to remain in a constant position relative to the
model, regardless of the position of the model in world space. This different, but
equally important, location system is commonly known as UV space. ‘‘U’’ and
‘‘V’’ are names given to a set of surface-based coordinates by which we are able to
find our way around the surface of a model.
This kind of local space definition has been used for centuries, in the field of
cartography. In the image of the roadmap in (Fig. 3.36), we use the row of letters
at the top or bottom and the column of numbers at the sides to find the location of
the area of the map we want. In this example, the ‘‘Catedral’’ is in the box defined
by G-9. Of course, hyper-accurate location information is also available via GPS
data, which can provide latitude and longitude information accurate to within a
few feet.
In computer graphics, the layout of U and V space is only slightly more
complex than locating points of longitude and latitude. A 3D surface is ‘‘unfol-
ded,’’ or laid flat—much like a Mercator or other projection of the globe onto a 2D
space. A calibrated, 2D grid is then projected onto this unfolded surface, which is
then used to identify coordinate space relative to the surface itself rather than
world space.
In Figs. 3.37 and 3.38, the sculpted head at the left has been unfolded and laid
flat; the flattened image is shown in Maya’s UV texture editor. By using the XY
grid overlaid on top of the organic, irregular shape of the unfolded head model,
specific points on the surface can be identified for the purpose of applying color,
texture, and other specialty ‘‘maps’’ such as normal maps.
38 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.36 Roadmap

Fig. 3.37 Sculpted head

In 3D computer graphics, the term ‘‘normal’’ refers to an imaginary line that is

perpendicular to a line tangent to a surface at any given point. The direction of
these normal vectors are compared to the vector of the line of sight of a camera or
a light, and the resulting angle (or ‘‘facing ratio’’) can be used to calculate any
number of dynamic visual relationships. For example, in a Lambertian
3.5 Normal Mapping, Displacement Mapping, and UV Layout 39

Fig. 3.38 UV ‘‘Roadmap’’ unfolded

Fig. 3.39 Angle of the light

is calculated relative to the
surface normal vectors

illumination calculation (illustrated below), the amount of reflected light energy

returned by a given surface is equal to the incident energy multiplied by the cosine
of h, where h is the angle created by the incident light vector and the surface
normal, multiplied by a diffuse value determined by the given material. This is
known as ‘‘Lambert’s Cosine Law’’ and provides a realistic representation of light
falloff as the angle of the surface changes relative to the angle of the light.
In Fig. 3.39 we can see that when the angle formed by the light and the normal
vector is zero (the light is pointing directly at the surface), the reflectance value is
1 (cos 0 = 1); and when the angle formed is 90, the reflectance value is 0 (cos
90 = 0).
40 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.40 The angle of the

light relative to the direction
of the surface is represented
by RGB color information in
a ‘‘normal map’’

Fig. 3.41 The surface

differences between the high-
res sculpture and the low res
form are represented in
grayscale for use as a
displacement map

The process of normal mapping uses this Lambertian cosine model to translate
the detail of a high-polygon model into a low-polygon model, using the following
process:
First, a sculpted model must be topologized correctly, as shown in previous
sections of this book. The model is then unfolded, and a UV map is extracted,
providing accurate surface coordinates.
The model is then imported into a digital sculpture application such as Au-
todesk Mudbox or Pixologic ZBrush, where the resolution density can be increased
incrementally for more detailed sculpture. As shown earlier, these applications are
designed to handle densities of millions of polygons, which provide the capability
of sculpting very fine detail such as wrinkles or fine skin texture. The density of the
model is increased systematically—with each face divided by a factor of four—so
that each level of resolution can ‘‘inherit’’ the same UV texture space as the low-
resolution version.
An analysis of the normal vectors is then performed on the high-resolution
model; the normal angle of each of the millions of faces is determined and stored
as an RGB value in a bitmapped image (Fig. 3.40). This bitmapped image is then
applied to the low-resolution image and used as a guide for perturbing the normals
so that they are systematically ‘‘misread’’ by the camera. Since the normal vectors
are a key component for the Lambertian diffuse illumination model, the result is a
highly detailed embossing pattern that exactly replicates the look of the higher
resolution geometry.
At the same time, another kind of map is derived, called a ‘‘displacement map’’
(Fig. 3.41). This is a grayscale image that acts in a similar fashion to the normal
3.5 Normal Mapping, Displacement Mapping, and UV Layout 41

Fig. 3.42 The RGB

information contained in the
normal map is interpreted as
an embossing layout, to show
surface detail

Fig. 3.43 Low resolution

sculpture appears to be very
highly detailed with the
application of the normal
maps and the displacement
maps

map; however, with this kind of map, the alpha channel of the image will phys-
ically displace the low-resolution geometry to approximate the changes in surface
contours represented by the high-resolution image.
The combination of the displacement map, which changes the actual shape of
the geometry, and the normal map, which acts as an ‘‘embossing’’ filter to rep-
resent the finer details such as wrinkles, lines, and pockmarks, allows for an
extremely high level of detail on a relatively low-resolution model.
Figure 3.42 shows a screen capture of the low-resolution image with the normal
map applied as color, for display purposes only. Notice how the patterns of color
line up with the finer detail lines of the finished sculpture (Fig. 3.43).
The accepted workflow is thus ‘‘Sculpt’’—Retopologize—Derive UV Map—
‘‘Up-Res’’—Sculpt—Derive Normal Map—Derive Displacement Map—‘‘Down-
res’’—Apply Maps—Render.
It should be noted also that the use of Normal Mapping only works with
polygonal geometry, not with NURBS. This is one major reason that NURBS
42 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.44 Student

sketchbook

modeling has fallen out of favor among artists and studios. The level of sculptural
detail that can be achieved using the normal/displacement mapping workflow has
made polygon modeling the preferred choice, particularly in the context of the
heavy technical needs of NURBS.
The important point here, which relates to the pedagogical models we are
presenting, is that the technical processes have become less intensive for the end
user. The retopology and UV layout are relatively simple procedures and the
derivation and interpretation of the normal maps are done entirely by the software,
which frees the end user to focus most of his energy on the aesthetic aspects of his
work.

3.6 Subtractive Modeling Workflow

As outlined earlier, students will assemble an array of images of their subject for
use as reference images. Using 2D imaging software such as Photoshop, or with a
pad of tracing paper, the student will identify the underlying bone structure,
muscle structure, shapes, and proportions. As with any traditional sculpture, a
number of rough sketches will be made to show these forms from various angles.
Figure 3.44 shows a page from a student sketchbook, prior to beginning on the
sculpture project.
Using a voxel-sculpting software such as Pixelogic’s ZBrush or Autodesk
Mudbox, the student begins to rough out the shape of the sculpture. In some cases,
the student will begin with a default primitive shape such as a sphere; in other
3.6 Subtractive Modeling Workflow 43

cases, they can begin with a rough shape that is part of a collection of ‘‘starter
models’’ provided by the software manufacturer. Using interactive ‘‘push/pull/
drag/carve’’ tools, the starter model is roughed into shape by the user. Although
this can generally be done with a mouse, many users prefer a pen-based tablet or
Wacom Cintiq tablet.
Figure 3.45 is an example of a default head mesh that comes with Autodesk
‘‘Mudbox’’ software. The next two images are the beginning placement of key
landing zones in the sculpture: the defined jawline, cheekbones, brow ridge, and
sternocleidomastoid tendons. With very little technical instruction or software
training, an artist can achieve a result like this within a very short period of time
(Fig. 3.46).
Note the difference in wireframe density between Figs. 3.45 and 3.47. The
default head mesh is 2,002 polygons, which is dense enough to achieve a rough
shape. As more detail is needed, the density can be increased. Each level of
increase cuts each face in half vertically and horizontally, for a 4:1 ratio. Hence,
Fig. 3.47, which represents one level increase in density, has a count of 8,008
quadrangles. There are clearly areas where this density is needed, and others where
it is not.
Figure 3.48 is a close-up of a higher-density model. With over 16,000 poly-
gons, we can see that it is possible to achieve some finer detail, particularly in the
folds of the cheek and nostril.
Ideally, we would want to have polygon edges follow these folds for more
efficient models and better muscle deformation, as the edges help to define the
shape of the creases. Any area where this happens at this stage (as indicated by the
blue line superimposed on the image) is purely by accident; more often, the creases
will cut across polygon edges randomly, as illustrated by the red lines.
Generally, large muscle masses and gross shapes can be modeled at the lowest
density level; as more details such as creases and folds in the skin are added, the
artist will increase the density stepwise, each level representing a polygon count
that is four times the previous level. The artist can move easily through the levels
of density, stepping down to change larger shapes, stepping up to add wrinkles and
other details.
In Fig. 3.49, we can see that four levels of density increase, or just over 32,000
polygons, are enough density to get a pretty decent amount of detail. However, in
the adjacent image (Fig. 3.50), we can see that much of this density does not
follow the natural lines of the face. Once we have achieved this level of com-
plexity, it is time to begin the process of retopology. It is important to identify the
point at which to stop this phase of sculpture, as the process of retopology can lead
to a significant loss of fine detail. Ultimately, we look to achieve no more detail
that is illustrated above; we can continue our sculpture on the retopologized model
more efficiently and with far less density.
44 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.45 Default head mesh

Fig. 3.46 Sculpture in-

progress, showing high-
density mesh

Fig. 3.47 Sculpture in-

progress showing no mesh,
focused on underlying facial
structure
3.6 Subtractive Modeling Workflow 45

Fig. 3.48 Increased density

‘‘forces’’ the ability to add
detail without regard to edge-
flow or modeling efficiency.
These topics will be
addressed in a later stage of
the model

Fig. 3.49 Very high-density

sculpture

Fig. 3.50 A dense sculpture

reveals a high level of detail,
and serves as a model for
retopology and the extraction
of normal and displacement
maps
46 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.51 Initial reference

image

3.6.1 Topology

As discussed in previous sections, attention to edge flow is critical in defining the

various masses and shapes of any object even more so when defining the muscles
and structures of the human face. Ridges, depressions, creases, protrusions—all
can be very efficiently modeled with relatively few polygons if the edges flow
naturally along the contour lines.
We take several approaches in strategizing our edge flow. The first, of course, is
to go back to the photographs, back to the drawing board, and sketch in the proper
placement of these edges. As an iterative process, we usually recommend that our
students do this with a pad of tracing paper or a 2D drawing software like
Photoshop which allows for multiple layers and variable transparency.
In Figs. 3.51, 3.52, and 3.53, the student has begun to sketch his polygon layout
on top of one of his reference photographs. Notice how the edges form concentric
circles around the eyes, following the form of the Orbicularis Oculi, and around
the mouth, following the form of the Obicularis Oris. Each line is carefully placed
based on the shape of the underlying anatomy (Fig. 3.54).
Another method that we use is to paint over the face, identifying broader facial
shapes based on the muscles underneath (Figs. 3.55, 3.56, 3.57).
In addition to drawing and painting over the original 2D photographs, students
can use 3D paint programs to lay out their edge flow directly on their sculpted
model (Figs. 3.58, 3.59, 3.60).
By using a combination of line sketches and painting, and comparing those
lines with information regarding the underlying structures of the face, the student
can then begin the process of retopology. In Fig. 3.53, above, we see different
students taking a variety of the recommended approaches in determining the
proper layout of the edges. In this way, complex layout ‘‘puzzles’’ can be solved
iteratively with a sketchpad before attempting the process using the modeling
software.
3.6 Subtractive Modeling Workflow 47

Fig. 3.52 Reference image

with edge-flow indicators
based on apparent lines and
understanding of underlying
structure

Fig. 3.53 Reference image

with edge-flow indicators
based on apparent lines and
understanding of underlying
structure

Fig. 3.54 Original

anatomical reference
48 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.55 Muscle groups are

subdivided into individual
polygon edges

Fig. 3.56 Digital paint is

used to define major muscle
groups

Fig. 3.57 Painting to define

muscle groups prior to
retopology
3.6 Subtractive Modeling Workflow 49

Fig. 3.58 Strategy for

retopology of sculpted model

Fig. 3.59 Completed

retopology strategy

Fig. 3.60 It’s often easier to

re-do layout details with paint
overlays prior to retopology

3.6.2 Retopology

At the time this case-study class was held, we felt that the two strongest standalone
retopology packages were ‘‘Topogun’’ and ‘‘3D-Coat.’’ Since that time, most high-
end modeling software packages have incorporated robust retopology toolsets in
their base package.
50 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.61 High density mesh

is imported into 3D-coat.
Following the underlying
muscles of the eye, the
mouth, and the jawline, the
artist begins to trace the
shapes with quadrangular
polygons

Fig. 3.62 Using previously

made illustrations and
acquired knowledge of facial
anatomy, the artist can finish
placing edges where they will
define the underlying shapes
most efficiently. Note that the
3D-coat software allows the
user to define an axis of
symmetry, across which each
new edge will be mirrored

While the toolsets may have some practical differences, the concept of reto-
pology is consistent. In each case, a high-density base mesh (like the ones we
created, or as created using a Photogrammetry process) is imported into the
software where it is traced using polygons. Our students have had plenty of
practice doing this—the placement of polygon edges follows the same workflow as
the draw-overs we have done on paper, in Photoshop, or in 3D paint packages. As
we draw with our mouse or stylus, we are drawing with edges and quadrangles. In
this manner, we can follow the sculpted forms exactly, while focusing our
attention on the technical elegance of good topology and edge flow.
This is the critical point described at the beginning of this chapter; up to now,
we have focused on anatomy, on sculpture, on expression—with no regard to any
of the technical requirements of efficient modeling. At this point, we can set the
aesthetics aside and focus on technology (Figs. 3.61, 3.62, 3.63).
It is important to reiterate the importance of finding the right cutoff, or stopping
point, in the initial sculpture phase. In this example, the student may have gone a
little too far in adding finer detail to his sculpture before retopologizing; notice in
3.6 Subtractive Modeling Workflow 51

Fig. 3.63 The retopology is

completed; the finished
model can then be exported
for further sculptural detail.
Notice how the flow of edges
and polygons match the
Obicularis Oculi (orange),
the Obicularis Oris (green)
the Occipitofrontalis (blue),
the mandible (pink) and the
Sternocleidomastoid (yellow).
The edge flow on these
defining facial ‘‘landing
zones’’ is critical

Fig. 3.64 Low resolution,

retopologized geometry

the below illustrations that while the main facial structure is intact, much of the
finer detail has been lost and will have to be redone. The new, retopologized model
can now be reimported into the sculpture package where the additional detail can
be readded. With the new, efficient edge flow, this phase of sculpture will be much
more easily accomplished (Figs. 3.64, 3.65).

3.7 Digital Sculpture: Case Studies Using Pixologic ZBrush

As the software for organic modeling has become more sophisticated, and the
hardware to support this technology has become both more robust and accessible,
the result is that the balance of emphasis on technical processes for end users has
shifted from being a cumbersome necessity to being nearly transparent. As we saw
52 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.65 Finished piece;

consisiting of the low-
resolution geometry rendered
with the normal map and
displacement map applied

in previous chapters, the ‘‘four student’’ model illustrates this balance/imbalance

between technical and aesthetic ability. The ideal student, following that matrix,
was the one who had the relatively even distribution of the two skillsets. This
sought-after ideal balance has become skewed, with far less emphasis placed on
technical ability as a part of the mix. The tools have become so automated that we
can focus almost all our energy on design and creativity without the encumbrance
of the steep technical learning curve standing in our way.
Perhaps at the forefront of digital sculpture technology is Pixologic, a privately
held software development company located in California, USA. Pixologic’s
flagship product, ZBrush, has become a benchmark standard for digital sculpture
in the 3D industry and serves as an example of the direction being taken in the
processes of both organic and inorganic modeling.
Rather than beginning with meshes that are perfectly topologized and following
a sequential subdivision process, the artist can start with a simple sphere and
follow a new methodology wherein he need not worry about topology in the initial
approach, focusing instead on shape, form, volume, and character. In fact, in the
experience of many artists working in the industry, their initial attempts to
maintain quality topology from the start frequently need to be redone as the needs
of the project or requirements of the art director change. To create the character
first and retopologize later is rapidly emerging as a much faster workflow, which
eliminates the need to address topology over and over again.
This new methodology reflects some sweeping changes in the industry that are
trickling down to education. As much of a shift that the ‘‘model—sculpt—reto-
pologize’’ paradigm has represented in industry and education over the past few
years, a case can now be made that traditional modeling, i.e., ‘‘box modeling’’ or
any of its offshoots, is becoming more of a preference rather than a necessity. Even
the relatively simple step of retopology is frequently considered by some to be a
hindrance to spontaneity and the creative process and can now be left to last (and
even automated) if necessary or preferred.
3.7 Digital Sculpture: Case Studies Using Pixologic ZBrush 53

From a pedagogical standpoint, we believe it is still important to provide stu-

dents with a strong foundation in traditional methodologies while keeping a close
eye on the industry trends that appear to be long lasting, that will shift the para-
digm. Taking the long view of modeling from its beginnings nearly a generation
ago, we see this as an important shift in approach that expands, but does not
replace, traditional methods.
Of course, in the field of scientific visualization and physics simulation, the
underlying mathematics of NURBS curves and surfaces is a necessary part of the
design process, but the kind of design that relies on this accuracy is technical in
nature. From a purely visual design standpoint, the process of dealing with curves
to derive surfaces to attach to create forms is labor-intensive and has become a
legacy process, no longer a viable option for production. The same can be said for
the additive modeling methods that incorporate a ‘‘topologize as you go’’ work-
flow. The constraints of the topology are now being replaced by sophisticated,
efficient software algorithms that do that work behind the scenes, freeing the artist
to stay focused on creativity.

3.7.1 Little Power

In his piece ‘‘Little Power,’’ (Fig. 3.66); Mariano Steiner of Pixologic demon-
strates the use of a process in ZBrush using a module called ‘‘Dynamesh.’’ Starting
with a sphere, he is able to quickly grab pieces of the sphere and with using only a
small handful of sculpting tool, can pull, push, carve, and otherwise massage this
geometry into a very sophisticated shape.
In the first three panels, a piece of this sphere has been stretched downward to
form the beginnings of the body structure. A close look at the topology reveals that
the edges are initially very distorted and stretched out, but soon become an evenly
sized array of fine quadrangular faces appropriate for continuing to add more and
more detail to this piece. A feature of Dynamesh is that it creates all quadrangles
and triangles which are all the same size, which makes it an excellent tool for fine
detail sculpting.
The change in the distribution of these faces in ZBrush is almost completely
automatic. At whatever point in the process the artist feels it is necessary, he can
select the model and select a menu item which quickly and automatically applies
this process to the model.
The first three panels show the topology on the red model, but the artist does not
need to see it at all. The process shown on the remaining five panels is one that a
traditional sculptor might take with a lump of clay—forming, shaping, and carv-
ing—and in some cases, ‘‘slapping on’’ additional bits of material which are
quickly and seamlessly incorporated into the model. In many ways, this process is
easier than working with clay—the artist does not have to worry about his material
drying out, he has an unlimited number of ‘‘undo’’ steps and needs not worry about
structural integrity or ensuring that his piece will hold together or not fall. While it
54 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.66 ‘‘Little Power’’

Fig. 3.67 Finished model, ‘‘Little Power’’, digitally sculpted using high-density sculpting
techniques

is entirely possible to display the wireframe structure of this model, it is unnec-

essary to do so at all.
As he works, the artist can quickly add detail to the sculpture by using a density
hot key, in ‘‘automatic’’ mode each quadrangle is subdivided once vertically and
once horizontally, so each step increases the overall density by a factor of four.
There are other ways in which the artist can specify density levels for specific
areas of the model as needed—but in either case, the efficiency of the software and
the ability of the hardware to effortlessly handle millions of polygons enables the
artist to add very fine detail to the model, as in the hands, shoulder structure, and
teeth of the example shown.
What is interesting about the finished product, shown here in Fig. 3.67, is that
Steiner ‘‘built in’’ specific imperfections to the legs and feet, adding to the sense
that this piece was created in clay, not pixels.
3.7 Digital Sculpture: Case Studies Using Pixologic ZBrush 55

Fig. 3.68 Batman

The intent of Mariano Steiner’s piece, ‘‘Little Power,’’ was purely sculptural,
with little need for a low-polygon-count or efficient edge flow. However, as the
next example will show, the additional steps of polygon-reduction and edge-flow
optimization for animation are just as seamless.

3.7.2 Batman

Another example of the same initial workflow can be seen in Jason Martin’s image
of Batman. The original high-resolution sculpture was created following a similar
workflow to the ‘‘Little Power’’ in the previous section.
The original sculpture Fig. 3.68 was created using ZBrush Dynamesh and
consists of millions of polygons. Using the tools in ZBrush and a subtractive
workflow, the artist was able to ‘‘decimate’’ this sculpture, i.e., reduce the polygon
count by substantial percentages.
The appropriate number of polygons in the end result is variable; it depends on
whether or not the final piece is for use in a game, a movie, or for a 3D print.
Further, retopology and reduction may be necessary at this point. Rather than use
the decimation workflow, however, which would reduce the polygon count further,
a different tool is employed called ZRemesher.
ZRemesher is a tool that will retopologize a model with close attention to edge
flow, as discussed in earlier sections of this chapter. This software can do this
completely automatically, but if the user has specific areas wherein he wants to
dictate specific edge flow to bias the automated retopology process, this
56 3 Conceptualizing Basic 3D Modeling Workflows

Fig. 3.69 ‘‘Decimating’’ a high resolution model

functionality is also built it. Rather than draw in the quads one by one; however,
the user can draw lines similar to those in the pre-modeling draw-overs.
In Fig. 3.69, the subtractive method used on this model is shown. The first
image is the decimated sculpture containing a little more than a million polygons;
by indicating specific lines on this model to guide the reduction edge flow, the
artist is able to reduce this model to several stages of polygon count.
Although detail is lost as the count gets lower, we saw in a previous section how
by generating a normal map from the high-resolution geometry, the detail can
reappear in the render.
Seeing the results of this workflow, and observing the ease with which these
sculptures are created, it is easy to see how this process is ‘‘The new wave’’ in
modeling.
Rather than beginning with meshes that are perfectly topologized and following
a sequential subdivision process, the artist can start with a simple sphere and
follow a new methodology wherein he need not worry about topology in the initial
approach, focusing instead on shape, form, volume, and character. In fact, in the
experience of many artists working in the industry, the initial attempts to maintain
quality topology are frequently required to redo this step as the needs of the project
or requirements of the art director change. To create the character first and reto-
pologize later is rapidly emerging as a much faster workflow, which eliminates the
need to address topology more than once.
During the process of digital sculpture, the artist will push, pull, inflate, and
otherwise stretch the geometry to achieve the expressive forms desired. This
naturally results in stretched topology making the addition of details very difficult.
Even after multiple subdivisions, this stretched topology will ultimately result in
an uneven and unacceptable distribution of detail across the model.
ZBrush’s mode, ‘‘DynameshTM’’ is an automatic retopologizer for the purpose
of sculpting. It uses only triangles and quadrangles which are all equal size, which
is ideal for digital sculpture.
3.7 Digital Sculpture: Case Studies Using Pixologic ZBrush 57

After stretching the model as described above, the user need only applies the
‘‘DynameshTM’’ tool as a quick step in the process to quickly and accurately
retopologize the stretched areas.
Surface topology, a necessary technical requirement of this process, has tra-
ditionally been a natural constraint to the workflow of artists; by eliminating the
need to address it so early in the process, and by automating it almost entirely at
the end, artists can be considerably freer to explore ideas in effect by ‘‘sketching’’
directly in the software.
Indeed, the need for a strong understanding of the technology has been greatly
reduced for the end user; most of the meticulous attention to mathematical surface
calculation, surface parameterization, and tangent continuity at shared edges,
‘‘edge loops’’ and edge flow is now taken away from the process by the combi-
nation of software, hardware, and the accessibility of both. As we said at the
beginning of this chapter, artists are free to be spontaneous and creative without
the constraints represented by a steep and cumbersome technical learning curve.
Image Credits
Figure 3.1 ‘‘Grays Anatomy’’
Figure 3.2 Photo courtesy of [Link]; drawover illustration courtesy of Mad-
eline Rabil
Figure 3.3 Photo courtesy of [Link]
Figure 3.4 Photo courtesy of [Link]; drawover illustration courtesy of Mad-
eline Rabil
Figure 3.5 Photo courtesy of [Link]; drawover illustration courtesy of Mad-
eline Rabil
Figure 3.6 Illustration courtesy of Madeline Rabil
Figures 3.7, 3.8, 3.9, 3.10, 3.11, 3.12, 3.13, 3.14, 3.15, 3.16, 3.17, 3.18, 3.19
Autodesk screenshots reprinted with the permission of Autodesk, Inc.
Figures 3.20, 3.21 Image courtesy of Tony Williams
Figures 3.22, 3.23, 3.24, 3.25, 3.26, 3.27 Autodesk screenshots reprinted with the
permission of Autodesk, Inc.
Figure 3.28 Illustrations courtesy of Deanna Giovinazzo
Figure 3.29 Photo courtesy of [Link]; drawover illustration courtesy of Hunter
Baule
Figure 3.30 Photo courtesy of [Link]; drawover illustration courtesy of Hunter
Baule
Figure 3.31 Image courtesy of Deanna Giovinazzo
Figure 3.32 Karl Langer, ‘‘Zur Anatomie und Physiologie der Haut. Über die
Spaltbarkeit der Cutis’’. (1861)
Figure 3.33 Photo courtesy of [Link]
Figure 3.34 Karl Langer, ‘‘Zur Anatomie und Physiologie der Haut. Über die
Spaltbarkeit der Cutis’’. (1861)
Figure 3.35 Photo courtesy of [Link]
Figure 3.37 Model courtesy of Eric Kimberly
Figure 3.38 Autodesk screenshots reprinted with the permission of Autodesk, Inc.
58 3 Conceptualizing Basic 3D Modeling Workflows

Figures 3.42, 3.43 Model courtesy of Eric Kimberly

Figure 3.45, 3.46, 3.47 Autodesk screenshots reprinted with the permission of
Autodesk, Inc., sculpturecourtesy of Ben Frazier
Figure 3.49, 3.50 Autodesk screenshots reprinted with the permission of Autodesk,
Inc., sculpturecourtesy of Eric Kimberly
Figures 3.51, 3.52, 3.53, 3.54, 3.55, 3.56 Illustrations courtesy of Eric Kimberly
Figure 3.57 Image courtesy of Daniel Pise
Figure 3.58 Image courtesy of Benjamin Frazier
Figure 3.59 Image courtesy of Andrew Turci
Figure 3.60 Image courtesy of Benjamin Frazier
Figures 3.61, 3.62, 3.63, 3.64, 3.65 Image courtesy of Eric Kimberly
Figures 3.66, 3.67,  2013 Mariano Steiner, image courtesy of Pixologic, Inc.
Figures 3.68, 3.69,  2013 Jason Martin, image courtesy of Pixologic, Inc.
Chapter 4
Photogrammetry and Design Workflows

Abstract This chapter illustrates a new paradigm in realistic 3D modeling;

combining the ability of hardware and software to handle millions of polygons
with the ability of digital photography and photogrammetry to produce these high-
density models. Various processes are now accessible to smaller studios and
universities, and this new paradigm is further easing the technological constraints
presented to digital artists.

  
Keywords Photogrammetry Autodesk 123D catch Scan Scanning Modeling 
  
Toolkit Maya Quads Subdivisions

4.1 Photogrammetry Workflows

Up to this point, our discussions have revolved around user-generated models

created by using various traditional and next-generation techniques. We have
shown how by following either a more traditional (additive) or next-generation
(subtractive) workflow, artists are able to visualize their imagined characters and
creatures with relative ease, for film, games, or any number of other applications.
The beginnings of this workflow do not, however, need to start in the computer.
Traditional sculptors, working in traditional materials, have always been a part of
any visualization pipeline. Better funded operations have long been able to use
high-end 3D digital scanning equipment to digitize these physical models for their
3D artists to reproduce in the computer, either using them as modeling reference or
going through the (previously) painstaking operation of retopology. As we have
seen in the previous chapter, softwares, such as ZBrush ZRemesher, Autodesk
Maya, 3D-Coat or Topogun, have substantially streamlined this process.
On the other end of the pipeline, the ‘‘input’’ end, the hardware, the software,
and the actual models have also become more accessible to smaller studios and to
universities. Following the matrix outlined at the beginning of Chap. 3, the net
result is that digital artists are able to fast-track the difficult and highly technical

S. Foster and D. Halbstein, Integrating 3D Modeling, Photogrammetry and Design, 59

SpringerBriefs in Computer Science, DOI: 10.1007/978-1-4471-6329-9_4,
 The Author(s) 2014
60 4 Photogrammetry and Design Workflows

Fig. 4.1 The subject is posed sitting in a chair in a reasonably well lit room

processes involved in model-making, and focus on the creative and imaginative

parts to which they may be better suited.
One example of how mainstream this process has become is the development of
Autodesk’s ‘‘123D Catch’’ software. The software is free and cloud-based; users
download a small-footprint image organization program that allows for quick file
transfer between their computers and the Autodesk servers.
Using a simple point-and-shoot camera, users photograph a subject from
multiple angles, assembling a series of 40–60 images. There are a number of
parameters that need to be followed in order to produce a successful file, but the
parameters are relatively easy to follow.
In Fig. 4.1, the subject is posed sitting in a chair in a reasonably well lit room.
Little attention is given to the lighting, however—no more than standard room
light is necessary for this capture. Note, however, that the light is constant, there is
detail in the room in the background, and the subject actually has sticky-notes
affixed to his head. This photogrammetry algorithm compares points in each shot
through triangulation; therefore, it is important to give the software enough
information to work with. Foreground and background areas are tracked from one
viewing angle to the next—so the subject must be still and the camera moved
around him.
This particular session used 80 individual photographs, as shown in the image.
Once these images are compiled into a folder and uploaded, an initial 3D model
is returned as shown in Fig. 4.2. The representation of the subject shown in the
4.1 Photogrammetry Workflows 61

Fig. 4.2 An initial 3D model Autodesk screenshots reprinted with the permission of Autodesk,
Inc

center of the image is an actual 3D model; the camera icons can be selected and
reference the associated still image at the bottom row.
This is an intermediate phase, showing the results of each calculation. Users can
make modifications to the model by deleting unneeded faces and can request that
the final calculation be performed at different resolutions.
The texture on the surface of the model, rather than using a standard UV layout,
uses a camera-projection method. The individual photographs are broken up and
blended, as seen in Fig. 4.3, with each piece being projected onto the model from
the angle of the original image. Figure 4.4 shows the final mesh resulting from this
process.
This software is very mainstream; its popularity is due to the price (free), the
ease of use, and the immediate turnaround of a very good-looking 3D represen-
tation of a subject. While the 3D mesh is very accurate and dense, there is a lot of
missing information or information that has been ‘‘averaged,’’ making this process
a little less suitable for a production pipeline than some other methods. Never-
theless, it has reached a very wide user-base, who are using this technology to
create exciting interactive 3D visualizations for the Internet, and making use of
Autodesk’s 3D print capabilities as well. Captured meshes can be returned to
Autodesk for printing in various materials including laser-cut cardboard and
polymer plastic.
62 4 Photogrammetry and Design Workflows

Fig. 4.3 The individual photographs are broken up and blended Autodesk screenshots reprinted
with the permission of Autodesk, Inc

Fig. 4.4 The final mesh

For production-ready scans, it is frequently a better option to use high-reso-

lution scanning equipment, where the finer details can be captured more accu-
rately. While the technology for this kind of scanning remains relatively costly and
thus out of reach for smaller studios, independent artists and universities, service
bureaus have formed wherein models may be purchased ‘‘off the shelf’’ or custom
scans can be contracted. This is a viable option for many and has provided a
necessary channel for the ‘‘input’’ side of the production workflow.
One such company is called ‘‘Ten 24,’’ located in Sheffield, UK. Figures 4.5
and 4.6 are examples of the kinds of models that can be purchased from them; they
4.1 Photogrammetry Workflows 63

Fig. 4.5 An inventory of scans of individual body parts Images courtesy of Ten24, Inc

Fig. 4.6 An inventory of full body scans Images courtesy of Ten24, Inc
64 4 Photogrammetry and Design Workflows

Fig. 4.7 The mesh derived

from one of the scans
purchased from Ten 24

Fig. 4.8 The beginning of

the retopology process using
Autodesk Maya 2014s new
‘‘modeling toolkit’’

maintain an inventory of full body scans and scans of individual body parts.
Individual and commercial licensing is available for these scans.
Figure 4.7 is an example of the mesh derived from one of the scans purchased
from Ten 24.
The mesh is very dense, over a million polygons, but can be utilized in any of
the ways described in Chap. 3.
The image in Fig. 4.8 is the beginning of the retopology process using Au-
todesk Maya 2014s new ‘‘modeling toolkit’’. The underlying scan data can be
4.1 Photogrammetry Workflows 65

Fig. 4.9 Roberley bell

sculpture

traced by the user using a near-standard polygon modeling process. The scanned
mesh is made ‘‘live’’ within the software, and the user can simply place points on
the surface from which quadrangular geometry can then be derived.
It is important to have a clear idea of the intended edge-flow layout before
beginning this process and to adhere to the plan. For this reason, it is equally
important (as described in Chap. 3) to strategize this edge flow using a 2D illus-
tration program like Adobe Photoshop, or even prints of the image with colored
pencils. Once the user has acceptable drawn-over lines, he can then begin the
process of applying the quads.
As points are added and quads are subdivided, the new subdivisions automat-
ically update to adhere to the underlying surface.
At points throughout this process, it may be necessary or desirable to exit the
modeling toolkit and apply more traditional polygon modeling techniques—to fill
holes, or to massage or soften edges, or to split individual polygons. Once this has
been done, the work in progress can be selected and ‘‘shrink-wrapped’’ to fit the
scanned model. In fact, primitive objects such as cylinders can be located around
large areas on the target surface and ‘‘shrink-wrapped’’ to fit. This workflow is well
suited for beginning to model arms and legs if not too much attention to edge flow
is required.
Once the base mesh has been retopologized, the scan data can be set aside and
further sculpting can be done, normal and displacement maps can be extracted, and
the processes outlined in Chap. 3 can be followed (Fig. 4.9).

4.2 Case Study: Roberley Bell

Sculpture at the intersection of the hand the machine-made.

66 4 Photogrammetry and Design Workflows

For commercial interests, a high level of control is necessary with 3D photo-

grammetry in order to achieve a predetermined output. With artist Roberley Bell,
she has an artistic process which explores and participates with some of the aes-
thetic ‘‘contributions’’ that the technical process delivers as part of the final product.
At what point does strangeness in scanning become an interesting part of a
creative, finished product? The Japanese ‘‘Wabi-Sabi’’ aesthetic is a philosophy
which centers on the acceptance of imperfections [1]. Often this aesthetic actually
values what others would see as imperfections. In a recent artist statement (below),
she discusses how her technical and creative process has led her to the point where
she is now trying photogrammetry as a component to her designs.
‘‘I am a visual artist an object maker. Over the past decade my work has
centered on the production of sculpture and site specific public projects. My work
explores the natural world both in abstracting from, and in borrowing, to reveal
hyper-realized fantastical landscapes. My current series Other Landscapes fast-
forwards the origins of organic abstraction into the twenty-first century. My forms
take their cue from Blob Design. Like the current trend in design where buildings
and form adopt an organic structure that is made possible only though computer-
aided design, my Other landscape sculptures, reveal themselves as natural forms,
though they are, in fact paradoxically based on nothing that exists in nature.
I have recently begun incorporating new technologies into my hand making
process of form. Utilizing a CAD (computer aided drawing) program rendering
blob forms via the computer and sending them to a CNC router (computer con-
trolled shaping machine.), which then digitally shapes the foam substrates of my
sculptures. Beginning with watercolor sketches and small clay models I then, with
the aid of a fabrication shop, draw the forms in a CAD program they are outputted
in foam via a CNC process at this stage I return to the hand making process
working the surfaces to articulate the forms, the nuance of the sculptures surface.
This evolution of my process engages innovation and extends the creative studio
practice beyond the tradition of hand working.’’

4.3 Case Study: Sucker Punch Studio’s

4.3.1 3D Scanning of Faces and Clothing for ‘‘Infamous:

Second Son’’ Game

[Link]
archive/2013/05/24/how-sucker-punch-photocopied-actors-for-infamous-second-
[Link]
The art style of the game Infamous combines superhero powers with a very
realistic aesthetic. In a trailer released for 2013 E3 ([Link]
watch?v=MINfjvFnzc8), they revealed clips from their upcoming game. In addi-
tion to the beautiful graphics, one of the most striking visual aspects was the
smooth and natural character facial movement. Sucker punch studios not only used
4.3 Case Study: Sucker Punch Studio’s 67

3D scanning for pedestrians and clothing, but the biggest application was for facial
animation. According to Max Chapman of Sucker Punch1:
‘‘The real reason we adopted scanning…was for our facial system. In terms of
assets the real benefit is quick and efficient data to either work from, or directly
adapt to increase realism, fairly straightforward, but not always simple. In the
greater scheme of things it’s more cost effective’’ he also went on to say that ‘‘it’s
time consuming and quite complicated.’’
Sucker Punch’s facial scanning, referred to by the acronym FACS, (facial
action coding system). Allows Sucker Punch to ‘‘…scan a host of facial features
and use the data to drive complex rigging systems to add complex micro move-
ments in our characters for greater realism. The end result is that Sucker Punch is
able to visualize and ‘‘basically able to adapt skin, bone, muscle, and fat inter-
actions on a human face. We would have never likely bought this equipment if it
wasn’t for that.’’
From your perspective as someone at a top game studio, where do you feel the
technology is going?
‘‘As far as scanning it’s the hybridization of technologies in the never ending
quest for greater realism. Real-time scanning capture solutions replacing or aug-
menting rather simulation is one of those areas. Higher resolution at faster and
faster camera speeds is the big benefit. Basically, how can you get a massive
amount of accurate data in real time and use that in conjunction with paradigms
that generate usable topology. On top of that there is the issue of capturing
accurate texture and light data for robust generation of BRDFs that could also pipe
back into real-time solutions. This is a long path of course, but it gets better and
better by the year.’’
How do these tools free and affect creativity, whether it be games or film?
‘‘…There is simply too much of a risk versus profit ratio to overcome that is
severely hampering the market as a whole, costs need to get more reasonable or the
industry will become unsustainable, meaning innovation/creative risk. At the
moment it is a thirst for realism, e.g., war games, GTA, sports, hopefully they
technology will help us move beyond that. Scanning is only one part of the
equation of course, but all in all technology almost always holds a major part of
the key to these issues.’’

Reference

1. Koren, L.: Wabi-Sabi: for Artists, Designers, Poets & Philosophers. Imperfect Publishing,
California (2008)

1
Interview with Max Chapman, Lead Character Artist, Sucker Punch Studios (October 18th
2013).  Sony Computer Entertainment America LLC 2013.
Chapter 5
3D Design and Photogrammetry

Abstract This chapter describes a framework for the design process with a spe-
cific focus on three-dimensional designs. It gives an overview of the elements of
design: color, line, shape, texture, space, and form, and then the principles of
design: unity, emphasis, placement, size and proportion, balance, symmetry and
asymmetry, repetition, proximity, and dynamics. It also investigates convergent
design paradigms from cinematography combined with 3D designs, which results
in 2D images.

 
Keywords 3D design Elements of design Color Line Shape Texture   
  
Pattern Space Principles of design Unity Gestalt, emphasis Balance  
   
Symmetry Asymmetry Repetition Proximity Dynamics Cinematography  

Lens Camera Film 
Designs are a mix of creative problem solving by applying various principles and
tools. There are various elements and principles of design that can be applied while
integrating new visual tools and technical workflows. In this section, we are dis-
cussing some of the ways design tools are integrating digital 3D forms to the
design process. We will approach the combinations by reviewing traditional design
rules by first defining the elements (or components of each image) and principles
of design while spending specific attention to ways 3D design can be applied to
these works. How elements are used will be more formally discussed in Sect. 5.2.
At the end of this chapter, we will expand upon the new workflows and oppor-
tunities that are starting to become realized by using the new photogrammetry and
retopologizing workflows.

5.1 Elements of Design

The elements of design are visual components that any image can be separated
into. Unlike the principles of design, the elements of design are almost uniformly
agreed upon set of attributes. One major distinction that can be made between

S. Foster and D. Halbstein, Integrating 3D Modeling, Photogrammetry and Design, 69

SpringerBriefs in Computer Science, DOI: 10.1007/978-1-4471-6329-9_5,
 The Author(s) 2014
70 5 3D Design and Photogrammetry

elements of 2D and 3D design is the addition of (3D) form. For this text, we will
discuss: color, line, shape, texture, space, and form. All of these elements of
design can combine to create imagery in nearly infinite combinations. As opposed
to ‘‘real life,’’ a design often uses elements in a way which abstracts reality, using
either reductive or exaggerated elements. We will discuss each element, defining
it, with a specific focus for interpretation and discussion of some of the various
common tools in most graphics programs. These elements will be discussed within
the framework of how they can fit as component elements in 3D graphics, and
photogrammetry-based 3D workflows, as well as rules of thumb for achieving
stronger graphical output.

5.1.1 Color

Colors are a breakdown of (visible) light within a small part of electromagnetic

(EM) spectrum that humans can perceive. The science and use of color is an
extensive topic. This book will give an overview, with a focus on resources for
developing color palates for use with the 3D design process.
There are two main ways of defining color space: additive or subtractive color.
Physical media such as paint are subtractive. Light at various wavelengths of color
is absorbed by paint and gets reflected back to the eye. Mixing all of the colors
subtracts (absorbs) all the color reflected to your eye until you achieve black. For
this book, we will be discussing color from the perspective of digital media, which
is additive. In additive color space, the main (or primary) colors are red, green,
and blue. When these primary colors are mixed together, the result is white. In the
figure, three colored lights overlap (add) to create ‘‘white.’’ The primary colors
red, green, and blue are mixed (as shown in Fig. 5.1) with each other to form the
secondary colors yellow, magenta, and cyan (Fig. 5.2).
Often combined with tints (whites) and shades (blacks), the application and
choice of color within design has a strong impact on the finished image. Specific
combinations of colors, or color palates, are a foundation for many designs.
Color in design is very subjective and will vary by cultural, temporal, and
national perspectives (see Sect. [Link] below). As the globe becomes more con-
nected, the greater contact with international populations highlights this subjec-
tivity by the exposure to multicultural design as well as discussion of various
meanings given to color (see discussion of Isomorphic Correspondence in Sect. 5.2
later in this chapter)
Each color can be mixed with various tones (tints and shades) (Fig. 5.3). How
the color is affected by the tonal changes depends on the amount mixed with the
color as well as the tool and display format used to present the final image.
Levels: This has become a standard tool in many programs. It allows you to
specify a point on within the white, blacks, and mid-tones of your image and
selectively stretch or contract that value. They are set with the three triangles
below the graph. To achieve the widest range of tone for each image, one goal is to
5.1 Elements of Design 71

Fig. 5.1 Additive color made with 3 lights in the program Maya

Fig. 5.2 Color strip shows primary and secondary colors

Fig. 5.3 Tone ranges (tints and shades) within a color

get the full range of lights and darks a computer can display. Some programs can
remap the tones (Fig. 5.4), by adjusting the triangles a user can chose to redefine a
greater or narrower range of tones. The dark tones are on the left and the lightest
on the right.

[Link] Relationship to 3D Form

For 3D objects, the color and form are realized by a combination of the object
color combined with any color from the light source illuminating that object. In the
real world and with some advanced computer-rendering techniques, the color of an
object further influences its surroundings by bouncing additional color off of the
object as well as potentially transforming light passing through that object
(assuming object translucency/transparency).
The perception of depth in 3D forms can be influenced by the use of light and
color. How light travels across the surface of an object describes the objects’ three
dimensionality. How light and color interact can subtly affect our perception of
depth. This is due to a biological impact of how the human brain interprets colors and
the perception of depth. This can be used stylistically within the 3D design process.
72 5 3D Design and Photogrammetry

Fig. 5.4 Levels. Adobe

product screenshot reprinted
with permission from Adobe
Systems incorporated

Warm (red–orange–yellow) colors are perceived as ‘‘pushing forward,’’ and cool

colors (blues) tend to recede. When developing 3D object designs, selective appli-
cation of color design principles can intensify or reduce the perception of depth
within 3D objects. It can also help to increase perception between foreground and
background elements. In 2D design, this is done by modifying color palates of
foreground and background elements. In 3D design, this can be done in a similar
fashion to 2D design, but with additional control of having color within 3D lights
affects foreground and background. One example demonstrating this is to set up a
scene with a character standing outside. Lights are placed in the foreground which
has a slight warmly colored ‘‘key’’ light (yellow or orange hue) directly on the
character while having the background lit with ambient color from the sky, with a
slight blue hue. The resulting image has the character with slightly more red tones,
and this helps give the character greater visibility and visual prominence in the scene.
A reverse of this process, casting cool, slightly bluish lights on the main character in
the foreground, while having warm colors in the background, can also be applied
adding a feeling of reduction or withdrawal to a character.

[Link] Color Palates

In design, there is often a reduced set of colors used when compared to the broad
spectrum available in real life. A color palate reduces the color choices for a
composition, to the main colors used (plus black and white). The color wheel (see
Fig. 5.5) can be divided in primary colors, which are colors which cannot be created
by mixing any other colors (red, green, and blue), each primary color, however, can
5.1 Elements of Design 73

Fig. 5.5 Color palates (Kuler). Adobe product screenshot reprinted with permission from Adobe
Systems incorporated

be mixed with another primary color to create the secondary colors (yellow, cyan,
and magenta), and also each secondary can mix to create the tertiary colors.
There are various starting points for color palates. Colors across from each
other on the color wheel are called ‘‘complementary.’’ In the color palate, this
color combination produces the strongest visual contrast. Analogous color palates
provide a color palate of color on a similar side of the color wheel, potentially
helping reinforce visual unity (see Sect. 5.2). Triadic colors are evenly spaced
around the color wheel (Fig. 5.5 analogous and triadic).
Joseph Albers was an accomplished artist across several disciplines (typog-
raphy, photography, and design), and he is best known for his series ‘‘Homage to
the Square,’’ which investigated color interaction between squares of different
sizes and colors. In his book Interaction of Color [1] (in 2013, transformed into an
interactive iPad application), he described many of the visual and physiological
impacts that variation in color and palates has. Many of his writings apply to
elements of technical process and design components. The interaction of colors in
a design has multiple effects on several levels: biological, cultural/stylistic, and
individual significance.

[Link] Culture, Stylistic, and Individual

The emotional and stylistic perceptions of colors are heavily influenced by the
culture a person lives in and also from an individual and temporal perspective. For
example, Irish consider the color green to be lucky, while in France, it has an
opposite perception. Green also has a strong connection to the religion of Islam.
Almost all colors have multiple meanings on cultural, individual, and temporal
levels. Often specific color palates will also be popular for some period of time.
74 5 3D Design and Photogrammetry

Fig. 5.6 Color palate derived from Vincent van Gogh’s starry night

For example, some neon color palates were popular in the 1980s in America,
whereas various shades of yellow, green, and brown were popular in the 1970s.
Known in Gestalt psychology as Isomorphic Correspondence, this principle states
that we respond to meaning. When we see an image (or color palate), we interpret
its meaning based on our personal experiences (Fig. 5.6).

[Link] Color Tools

There are many programs and apps available for developing color palates. One
program integrated into multiple Adobe programs is called Kuler. It is available
both online at [Link] and being integrated as a program extension into
many Adobe products (Photoshop, illustrator, etc.). It has specific tools for cre-
ating and saving palates, as well as helping to create starting points using color
palate rules—for example, analogous, triadic, and complementary. For example,
Vincent van Gogh’s painting Starry Night was used as a reference to drive a color
palette (Fig. 5.6).

5.1.2 Line

Line use within design often stems from analog media such as a pencil or other analog
drawing implements. Due to the shift from analog to digital art, there have been
attempts to transform approximations of analog tools into digital forms. In physical
media, a line is a mark made by a moving point. In analog media, the variation in
direction and weight produces an almost unique look for each artist (Fig. 5.7).
Lines also exist in many ways in nature, for example, lines on leaves, tigers, and
on birds. They can serve for opposite purposes also, to make objects stand out or to
help camouflage an object with its surroundings.
Our eyes are particularly geared toward understanding lines and using them as
part of object detection. Within high-contrast images, objects are often identified
by their silhouette. It has been long used within design to use lines to guide the eye
to different parts of the image [2]. This has also been established scientifically
(Motter, Brad C, and Belky).
5.1 Elements of Design 75

Fig. 5.7 Line

Fig. 5.8 Lines and form

The emotional effect that lines can generate within designs is significant.
Horizontal and vertical lines usually communicate stability and solid forms. How
the lines are used within imagery further affects the emotional impact. We will
discuss this further in the section below on composition.
Lines can be combined with colors and other lines to create textures and pat-
terns (see the Sect. 5.1.4 below for further discussion).

[Link] Relationship to 3D Form

Many 3D programs often feature the ability to create ‘‘toon lines’’. These lines are
created around each object on a 2D plane parallel to the camera (see Fig. 5.8).
Computer programs generating images of 3D objects have the ability to extract
information about the 3D object, which describes whether the faces of that object
are toward or away from the camera. Within the program Maya, the ‘‘Sampler Info
Node’’ allows the user to extract the facing ratio. The facing ratio generates a value
that is returned from 0 to 1 as to what percentage the plane of geometry is facing
the camera. Using the facing ratio to apply color or luminance information to a 3D
object which consistently faces the camera is how a custom cartoon silhouette line
is created.
When used in combination with a ‘‘toon shader,’’ which simplifies lighting, this
helps to create the perception that the 3D objects are drawn. Standard black or
white toon lines help cartoon objects to stand out from their background while
creating a visual style.
A custom toon line can give subtle or strong additional design control to the
artist. Lines can be used in 3D design to accentuate or deemphasize the 3D shape
76 5 3D Design and Photogrammetry

Fig. 5.9 Contrasting and accentuating expressive line and 3D form

Fig. 5.10 Line flow,

underlying structure

of the form. Given that a separate line pass in toon shading can be generated, a line
could be used to accentuate curvilinear or rectilinear forms. It could also be given
additional ‘‘emotional’’ (examples from Maya Visor, Fig. 5.9) design attributes.
Finally, it could also emphasize the medium being used, like the (below left),
which gives a water-colored look.
Line direction within the composition can greatly affect the emotional feeling of
the image. If the lines in an image are strongly horizontal or vertical, often the image
feels static, while diagonal lines often create greater feelings of movement. Strategies
for using digitally created lines will be discussed more in the section on composition.
In 3D design, the underlying structure of an object is made-up of connected
polygons. The direction and shape of the polygon describe the underlying form of
complex objects. In the simple as shown in Fig. 5.10, the lines describe how the
cylinder intersects the plane below.
In a more complex example, within humans, there are lines which describe how
the skin will fold and deform by default, called Langer’s lines. In addition to the skin
having default fold patterns, it also follows underlying musculature, bone as well as
how the body needs to move. Therefore, the underlying line flow (also known as edge
5.1 Elements of Design 77

Fig. 5.11 Shape

Fig. 5.12 Gestalt reification:

implied triangle shape

flow) structure of a 3D object needs to reflect how the object will deform. This is one
of the reasons for retopologizing objects as discussed further in Chap. 4.

5.1.3 Shape

Shape is a description of a two-dimensional area. Shapes can be created and

perceived in multiple ways. Often they are described by a closed group of lines (as
the black outline identifies common shapes in Fig. 4.11); however, shapes can also
be defined by a commonly filled area. As in Fig. 4.11, each shape is filled with
gray color. Shapes can be typically categorized into two categories: geometric
shapes or organic shapes.
Geometric shapes are easily described mathematically. They are regular or
precise and are regularly found in man-made (3D forms and) objects. In Fig. 5.11,
the first three shapes, square, triangle, and circle, are geometric shapes.
Organic shapes are irregular and difficult to describe with simple math and are
often found in nature.
Puddles, leaves, flowers, clouds, roots, rocks, trees, etc., and the final two
shapes in Fig. 5.11 (right) are all examples of organic shapes.
Shapes can also be implied by the use of negative space (see Fig. 5.12 Gestalt
reification). Implicit sense of line and shapes can be perceived within a negative
form. The Gestalt principle of reification is the term used for the brains ability to
construct or generate shapes and forms inside of negative space.
78 5 3D Design and Photogrammetry

Fig. 5.13 Texture

[Link] Relationship to 3D Form

As previously discussed in the Sect. 5.1.2, the shape of a 3D object can partially
be described by the shape of its silhouette. However, this can vary dramatically
depending on how the object is perceived from various angles. Highly recogniz-
able shapes are core components with 3D design. Often, just by silhouette alone,
an object with a strong, distinctly recognizable shape can be recognized. In Gestalt
psychology, this property is known as Invariance. When designing distinctive 3D
objects whether they are commercial products or character designs, one step in
checking that design is to simplify the 3D object down to a solo silhouette shape,
by turning the object completely black and the background completely white. This
is a recommended test analysis technique, especially with character designs. Most
cartoon characters are easily identified by their silhouette.
3D designers have the potential to combine the principles of reification and
invariance by modeling objects which look one way from one angle and/or have
multistable negative spaces which appear within the object.
Shapes can be separated into rounded (curvilinear) or angular (rectilinear)
forms. Rounded forms often represent gentle or kind emotions, whereas pointed
angular forms can represent exiting or negative emotions. This connection
between the forms extends beyond the visual to audio. Successful design often
attempts the greatest engagement by trying to elicit multiple senses and emotions.
There is some crossover and research between images and other senses, and this
synesthesia (combining sensory perceptions) has been determined to be universal:
‘‘Speakers cross-linguistically associate non-words that have round vowels, such
as/buba/, with round shapes, and non-words without round vowels, such as/kike/,
with spiky shapes’’ [3] (Fig. 5.13).

5.1.4 Texture/Pattern

Textures describes the surface qualities of an object. This can cover large
dichotomies of qualities: See Fig. 5.13 for some examples smooth/rough, wet/dry,
shiny/dull, warm/cold, etc. All objects have some texture. In digital 3D design,
these qualities are visually simulated. A ‘‘shader,’’ which often models qualities of
the real world, is given to the software which then interprets how the texture
should be represented based on the angle of the camera to the source(s) of light in
the scene.
5.1 Elements of Design 79

In traditional design, the texture of an object can also be touched. This is not
simply in sculpture, but the surface of a painting can also contain various textures.
Vincent van Gogh was famous for building up paint into rough peaks on the
surface of his canvas, a technique known as ‘‘impasto.’’ Other techniques that draw
attention to the texture of the object might be adding sand to paint or the use of
various types of materials in the construction of a (collage) image.
Textures can be categorized in several ways:

[Link] Man-Made

These textures are often related in some way to the function of the object as well as
the manipulation of both natural and purely constructed elements. Man-made
textures demonstrate a connection to humans in one or more ways. Often the
texture is connected to the construction process, for example, wood that has been
lathed to form components of a chair. The construction process for lathing would
aligns the grains of the wood to flow along the length of the chair which results in
maximum strength of the constructed component. The grains of wood flow along
the length of the chair leg (see Fig. 5.14). Man-made textures also reflect the
purpose of an object, such as the reflective qualities of a street sign and the
semiotic elements included on the surface.

[Link] Natural

This kind of texture tends to reflect the ‘‘organic’’ look and feel of nature. Spe-
cifically, this means they often reflect a greater connection and response to ele-
ments such as water, air, and sun.
Patterns:(see also Sect. 5.2.8)
They define an underlying structure ordering the texture. They exist in both
natural and man-made textures.

[Link] 3D Computer Graphics, Textures, and Connection to

Mathematics

3D computer graphics are based on mathematics. Simulating 3D form and texture

is a major area of research. The surface texture (especially shallow bumpiness) of
an object within 3D graphics is often not generated by tessellating simulated form;
it is often generated by a bump or normal map (see Fig. 5.15). This can be an
efficient way to simulate visual complexity with increasing the number of poly-
gons in a surface.
Our eyes are trained over time to translate and perceive a series of patterns into
meaningful information. As a culture, we are now training our computers and
80 5 3D Design and Photogrammetry

Fig. 5.14 Connecting design to reality

machines to also perceive those images and translate them into three-dimensional
forms.
Numerous mathematicians have noticed the underlying mathematical nature of
natural patterns [4]. This connection has been discussed by scientists for thousands
of years (Cook). Many of the earliest recorded observations of math within natural
patterns to spirals (see Figs. 5.16 and 5.17). Fibonacci, D’Arcy Thomas, and Alan
Turing have described natural, visual phenomena which describe complex textures
which all can be described using mathematics.
Many of the nature-inspired mathematical formulas have been applied to
computer graphics texture creation. Some of the more common applications are
‘‘fractal generators’’ which use a repeating equation to generate an image. The
Mandelbrot and Julia fractal equation sets are probably the most well known and
common fractal types; however, there are many further fractals and combinations
which have been identified.
Texturing 3D objects can be done with a mix of raster-based imagery. This is
often photographic but sometimes created by the artist in an application like
Photoshop. However, large photograph-based imagery used for textures can be
inefficient and lead to unnecessary increased resource usage. Being able to
mathematically generate natural textures has given artists increased flexibility to
alter the textural pattern, increase the visual density, or recalculate for higher
resolution. An example of this is with Perlin noise [5] (See Fig. 5.16).
Ken Perlin invented ‘‘Perlin noise’’ in 1982 and more recently simplex noise in
2001, which are algorithms for generating efficient math-based textures for effi-
cient use in graphics applications. What is visualized is a simplified type of gra-
dient noise, a complex, organic looking pattern, but at very low computer resource
costs [1].
5.1 Elements of Design 81

Fig. 5.15 400 faces versus 49,000 faces using a normal map

The connection of mathematics relates both to textures and in the form of

objects. There are many programs that use algorithms to generate 3D objects.
There are many natural objects which have tree-like branching structures which
can be represented by fractals using computer-assisted algorithms for efficiency as
an approach for generating geometry is rapidly being adopted. This rule-based
approach to generating geometry rapidly combining with computer-vision- and
photogrammetry-based approaches. While the current set of photogrammetry-
based geometry creation and retopologizing tools and workflows being discussed
in this book are human assisted, there is already work in using algorithms to
further automate is foreseeable that many of these.
82 5 3D Design and Photogrammetry

Fig. 5.16 Romanesco Broccoli

Fig. 5.17 Perlin noise

[Link] Texture Tools

There are an overwhelming number of tools for generating textures and applying
them specifically for 3D computer graphics. Choosing the correct texturing
application for a specific task is usually done by finding the tools and feature sets
which most rapidly accomplish the task and within a budget. In addition to well-
known programs such as Photoshop and internal 3D shading capabilities built into
3D programs such as Autodesk Maya (Hypershade), Mari (developed by The
5.1 Elements of Design 83

Fig. 5.18 Space

Foundry) and Substance Designer (developed by Algorithmic) are currently being

heavily used by film and games industry, respectively (Fig. 5.18).

5.1.5 Space

Space separates visual elements within an image. However, how space is per-
ceived within a design strongly influences overall feeling of created visuals.
Regardless of 2D or 3D design, the use of space in an image is designed to engage
the viewer. The way space is treated between 2D imageries relies on creating
relationships between objects on a flat 2D image (X, Y coordinate system) or in
simulating 3D imagery. Figure 5.18 shows an example of space represented using
the context of humans and an environment.
How empty (or negative) space is designed also can be used to deliver addi-
tional information about an image. The use of space in an image can create a
variety of emotional tones. For example, large gaps of space can create feelings of
loneliness or very little space can create a feeling of confinement. Overlapping
elements can help create visual engagement between objects, helping to make the
viewer question the relationship between the overlapping objects. The way space
is used can also help create rhythm or balance (see Sect. 5.2).

[Link] Relationship to 3D Form

The techniques of one-, two-, and three-point perspective used to generate within a
2D picture plane the perception of 3D space have been used for thousands of years.
3D computer graphics integrate true depth (Z) space within imagery. Whether
done within 2D space or true 3D space ‘‘depth cues’’ can create or intensify feeling
of depth within an image.
Depth Cues (see Fig. 5.19)
Color and shading: As objects recede, the color and details are reduced relative
size: Objects ‘‘near’’ are perceived as being larger.
Overlap: Objects can be perceived as being in front of or behind based on
overlap.
Linear perspective parallax: Lines that are parallel show convergence, if the
imagery shows movement—objects in the foreground move more quickly than the
background, known as perspective parallax.
84 5 3D Design and Photogrammetry

Fig. 5.19 Depth cues

Atmospheric perspective: As objects recede into the distance, they are reduced
in visibility due to water vapor in the air.
3D computer graphics technologies generate images containing 3D spaces,
often very similar to images created using moving or still-picture cameras. Later in
this chapter, in the Sect. 5.2, we will discuss some of the techniques and con-
ventions from the fields of photography and film cinematography that can be
applied to 3D computer-generated imagery.

5.1.6 Form

This book specifically deals with the 3D form. Form as a design element is any
object that can be measured with height and depth. A 3D object is understood by
how it responds to various sources of light. For example, in Fig. 5.19, the sphere
without light and dark would simply look like a (shape) circle. Additional qualities
such as surface texture and bumpiness, specular highlights, reflectivity, refraction,
and transparency further give visual clues about the form of an object.
There is significant overlap in the discussion of 2D and 3D forms. 3D objects
can be constructed in a curvilinear, rectilinear, or a combination of the two. They
can also have constructed or organic feeling to their form.
5.1 Elements of Design 85

[Link] Types of Form

Forms can be categorized in multiple ways: how they fill physical space (mass) or
how they surround empty space (volume). The mix of positive and negative spaces
is a core component of defining the composition of 3D forms. They can also be
categorized by 3D shape, for example organic, geometric, or mechanical forms.

[Link] Form Within Environment

Since 3D forms exist within 3D space, there are additional relationships that are
created within the environments the forms are placed.

[Link] Form Tools

In this book, we will spend a great deal of time discussing how we achieve and
develop form.
3D computer graphics comprehensive toolsets have the ability to achieve nearly
any type of 3D forms. As discussed in Chapter 4, many of the tools allow for a
variety of possibilities including a combination of hand-implemented sculptural,
pushing rectilinear as well as procedurally generated naturalistic sculpting as well.
As discussed in Chap. 2, the nature of technology often causes significant con-
vergence among fields. This is happening within the fields of design field as well.
When the field of 3D graphics was new simply knowing basic 3D modeling was
often enough to obtain gainful employment. Growth in the number of 3D modeling
specialists in industry combined with photogrammetry and other accelerated
workflows have reduced demand for 3D modeling jobs. In addition, as discussed in
Chap. 2, the changes in accelerating technology become disruptive.
As 3D asset creation becomes more automated, additional toolsets in design and
cinematography will become ways to differentiate 3D artists from technicians.

5.2 Principles of Design

Principles of design are used to provide organization and structure to the elements
of design. Within the design field, there is no exact list of principles of design.
However, there is a fairly consistent core group. Due to the nature of this book as a
discussion of a convergence between multiple technologies and fields of design,
we will also address some of the design components from the fields of photog-
raphy and moving picture cinematography as well as the following five principles
of design:
86 5 3D Design and Photogrammetry

unity,
emphasis,
balance,
repetition, and
dynamics.

Within this list, we will also cover several subcategories. Additionally, while in
other design books, principles of design often cover both 2D design and maybe 3D
design [6]. For the purpose of this book, we are going to specifically focus on the
application of these principles with the understanding that the final image is one
that is created using 3D forms.
Designing 3D computer models, from computer-vision-accelerated stereopho-
togrammetry as discussed in previous chapters, uses knowledge from multiple
fields. The convergent fields and technologies we will discuss in the next few
pages are photography, film, and animation and Gestalt psychology.

5.2.1 Camera: Lens, Film, and Exposure

[Link] Lens

The medium of image creation greatly affects what is perceived in the final design.
Whether a designer modifies perspective artistically or a cinematographer uses
various lenses, how depth is perceived greatly affects the final composition. Two
images taken from the same location with different lenses generate significantly
different visual output.
A long lens (telephoto, 70–300 m+) will significantly compress the perception
of depth within an image. In addition to this quality, the length of the lens also
affects the plane of focus, which is the distance from the lens where objects are in
focus.
A short lens (wide angle \21 mm) will deliver a wider angle of view. Wide-
angle lenses accentuate the feeling of depth by warping an image. All types of
lenses can be simulated in order to achieve appropriate ‘‘feeling’’ for the image
within 3D digital graphics.
Depth of field specifies what objects are in focus and out of focus. A shallow
focus can help generate images and designs which guide the viewer to a desired
object. With real-world optics, it is sometimes difficult to generate specific
sharpness of a subject with exact blur for the out-of-focus areas. Within 3D
computer graphics, users can generate an additional Z (or depth) channel within
many file types allowing the creator to specify depth of field as part of the post-
production process. In traditional film, this was impossible to accomplish. The
Lytro camera, using light field technology, is now able to capture light and depth
in a way that allows post-process planes of focus. In addition to the Lytro camera,
within the photogrammetry process, the camera, lens, and focal distance
5.2 Principles of Design 87

information are becoming standard components of metadata within calculations

used to generate photogrammetry output. As additional cameras adopt depth-
sensing technologies (such as the PrimeSense depth sensor on the Kinect), this will
give further accuracy and interactivity to combining image captures in the real
word as both photographs and 3D objects.

[Link] Film and Exposure

While photochemical film is largely a technology of the past, there are significant
parallels to the visual output generated by 3D-rendering engines to film stocks of
the past. One characteristic that parallels film type is the renderer. Each 3D
computer graphics program’s render has default ‘‘looks’’ that are generated. These
default looks can be modified. Factors similar to the speed of film (how fast it
captured light) dictated how quick the shutter needed to open and close as well as
the width of the aperture of the camera in order to expose the film. The resulting
image would be further affected by how much grain was visible within the image
as well as motion blur. Within 3D graphics, motion blur and exposure are highly
controllable elements that can be specified within the software or also done as part
of the post-production process.

5.2.2 Cinematography

While many of the principles of design that will be discussed below deal with
elements of composition, in cinematography, composition refers to what is cap-
tured by the camera in frame of the image. The mise-en-scène (from the French)
‘‘on the stage,’’ encompasses everything from actors, props, lighting, and even the
type of lens used on camera. While there is significant intentionality to the cine-
matography, however, there is potentially greater organic latitude due to the
moving and potentially unpredictable nature of the elements within the frame. In
3D computer graphics, almost all elements are currently planned. Because of this,
we will mostly focus on the still image within cinematography. We will separate
framing into the categories dynamic and static.

[Link] Dynamic Composition

In our previous discussion on the elements of design, the use of line greatly affects
the emotional impact of an image. When an image has strong diagonal lines, the
diagonals in an image create a feeling of excitement. Lines also guide the eye
around the image. The composition of an image with multiple triangular shapes
helps create visual tension or excitement, whereas an image with multiple curved
shapes can have a gentle or sensual.
88 5 3D Design and Photogrammetry

How the objects in the image sit within the frame of the image can create
dynamic or static composition. We will discuss this further in the sections below
on (2) 5.2.6 and (8) 5.2.5.

[Link] Static Composition

Strong horizontal or vertical lines are often considered static, calm, or soothing.
When the objects in the scene are also centered in the frame, this also creates a
static composition. The vertical or horizontal lines can be generated by man-made
or natural elements.

5.2.3 Unity

It is often the first and/or final consideration of any design. To achieve unity, a
variety of elements need to work together to create a cohesive image. This design
principle is connected to and supported by all of the other elements and principles
of design. It is also conceptually close to the Gestalt principles of design.
This concept deals with connecting all of the elements in the image into a
cohesive whole in a way that all should work together to support and amplify the
core idea or emotion being expressed.
An example of this would be an image that has a relaxed or gentle subject that
could be further unified by additional round shapes, smooth textures, and reduced
total contrast. Reversing this concept for an image that has a dynamic, exciting
feeling could be further unified by adding additional diagonals, triangular shapes,
rough textures, and high tonal contrast. Applying consistent post-process ‘‘on top’’
an image can help to unify it: grain, color, blur, and other post-processing on top of
the image.
Additional ways to increase unity in images as well as image sequences would
be to stage the image in a way that would visually link the imagery in a way that
would guide the eye around the image.
Unity is also highly related to design philosophy, physiology, and psychology
of Gestalt. The word Gestalt means that the organized whole is perceived in a way
that is greater than the sum of its parts. The human brain naturally looks for ways
to unify imagery [7]. Much of the Gestalt philosophy deals with how the brain
groups and categorizes visual imagery. Additional Gestalt principles can be used
to help unify elements if an image needs additional help. These are continuity,
alignment and containment, closure, proximity, and similarity.
5.2 Principles of Design 89

Fig. 5.20 Forms

Fig. 5.21 Gestalt grouping principles

5.2.4 Gestalt Grouping

A major focus of Gestalt psychology is based on the implication that the mind
translates external stimuli holistically rather than the sum of their parts.
Continuity: Objects that overlap in 3D space are often connected by the brain.
Alignment and Containment: It is also known as uniform connectedness [8].
Objects that are arranged within 3D space in ways that put them on similar planes
of height make them. Surrounding the objects within the rectangular form as
shown in the left Fig. 5.20 helps to group them, even though they are not even the
same shape.
Closure: The brain has a natural tendency to connect (see also proximity) or
close a space—do you see a circle in Fig. 5.20 left.
Proximity: Objects that are near each other are connected by the brain into
groups (Fig. 5.20 center).
Similarity: Components within an assortment of objects are cognitively
grouped if they are similar to each other. Making exact repetitions of 3D objects is
very simple in 3D computer graphics, making similarity and repetition (see also a
greater description of this design principle) to be strong unifying components
(Fig. 5.21 right) [9].
90 5 3D Design and Photogrammetry

Fig. 5.22 Rule of thirds

5.2.5 Emphasis

It refers to the point of focus or interruption. It indicates the locations in an image,

which most strongly attracts the viewer’s attention. There are usually a primary
and sometimes secondary areas of emphasis. Emphasis can be achieved in many
ways. One is to create outstanding elements by contrasting form, luminance, or
color, to create a break in rhythm or dynamics, or dominance within the visual
hierarchy (example, increased scale within a set of multiple forms).

[Link] Placement

The rule of thirds divides an image into thirds on the horizontal and vertical axes
(Fig. 5.21). Placement at one of the corners of intersection is one area where there
is emphasis. In addition, throwing the image of center, when used appropriately,
can create additional dynamic imagery with emphasis placed on areas of the
image, which further strengthens the composition (Fig. 5.22).
Focal point—what is the center of design. What is visually dominating?
In addition to how form is placed and fills the visual field, there are many
additional ways to add differentiation.

[Link] Size and Proportion

Proportion deals with the relationship between objects. The relative size of objects
within a composition helps to inform size and spatial relationships between
objects. How their relative size and scale is imaged in the context of the other
objects in the scene works to develop the visual hierarchy. Other depth cues may
be used (see Sect. 5.1), but the standard cues for size within an image are the ones
5.2 Principles of Design 91

best known to us; humans. 3D visual designs can proportionally show the human
form as huge or small against the background, or even other human forms. This
effectively generates emphasis on either the human or the environment, as well as
informing the viewer as to the size of the environment.

[Link] Contrast

The brain is specifically designed to notice differences. Using various methods to

achieve visual or conceptual contrast helps to engage the viewer. Creating contrast
within an image can be done in many ways. Each of the elements of design (color,
line, shape, value, texture, and form) can be used in ways that create contrast.
Using opposing or juxtaposing elements within the design can create contrast.
This can be specifically done with in one or several elements of design such as:
Color: Use complimentary colors (colors that are opposite from each other on
the color wheel)
Tone: Perhaps the most obvious is creating areas of strong brightness and
darkness.
Shapeand Line: contrast curvilinear and rectilinear forms, or smooth and
jagged lines.
The main contrast in an image should usually be located around the center of
interest. Having too much contrast (variety) in all areas can destroy unity as well as
create a feeling of chaos and confusion (unless this is the goal of the image).

5.2.6 Balance: Symmetry and Asymmetry

‘‘Balance is a state of equalized tension and equilibrium, which may not always be
calm’’ (White). As a design principle, balance is a visual concept which stems
from the physical world, where an object must reconcile between opposing
physical forces such as gravity, slope of the surface it sits on, shape of the object,
and that object’s center of gravity in order to achieve balance or tip over.
For a digital image of a 3D object, balance within that object can be achieved
with symmetry, with asymmetry, or with radial balance. How the 3D object is
realized as an image and achieves balance will also depend on how it sits within
the context of the image frame as well how the object is lit.

[Link] Symmetry

Similar to the same principle in physics, imagery can be defined in ways where an
image’s visual weight can be imagined. Symmetrical organization is created
through the alignment of graphics along an axis. The image plane can be imagined
as a scale, and the focal point of the image is the fulcrum (Fig. 5.23). Large
92 5 3D Design and Photogrammetry

Fig. 5.23 Symmetry

Fig. 5.24 Asymmetry

balance

symmetrical shapes when placed in the center of an image are considered in

balance. The weight of each shape within an image can also vary, depending on
how it is visualized. Light colors, transparency, and little shadows can reduce the
weight of an object in an image (Fig. 5.24).

[Link] Asymmetry

Also known as Informal Balance, asymmetry is conceptually more complex and

deals with creating balance within a composition by distributing visual weight
among elements around a fulcrum point. This can be done by placing imagery of
various visual ‘‘mass’’ on different sides of the fulcrum point. Figure 5.24 is
asymmetric, but because its center of mass is distributed, the object is centered on
the frame and appears to be resting on a surface and this creates a feeling of
balance within the image.
5.2 Principles of Design 93

Fig. 5.25 Facial symmetry

[Link] Symmetry Tools

There are many tools available for aligning objects both within the context of the
image plane and within the context of other objects. Many modeling and sculpting
workflows now include the ability to create objects symmetrically, this can greatly
accelerate initial designs. This can be done in one of the two ways. One way is to
start by building half of the object and creating a mirror image duplicate of itself.
The second way is that many programs now often include a symmetry button for
creating objects. As an example (Fig. 5.25), the face was modeled with ‘‘symmetry
on’’ every detail sculpted on one side translated to the other. In the real world,
faces are rarely 100 % symmetrical; however, it is a factor that connects to rep-
etition as well as aesthetic beauty [10]. Working using the second approach works
better because of the ability to visually check the final result.

5.2.7 Repetition

Repetition with variation is often used within design. Repetition can help unify an
object, by creating a series of repeating forms. How those forms are laid out
connects repetition to the principle of rhythm (see Discussion in Sect. 5.2.8).
However, without any variation, the repetition can become boring. One reason for
boredom is that the brain can too easily and immediately perceive the image in a
single glance. However, when variety is added within repetition, it creates a
stronger visual image which will draw the eye in and engage the brain in a way
that will cause further consideration.
94 5 3D Design and Photogrammetry

Fig. 5.26 Proximity

[Link] Repetition Tools

In most programs, there is the ability to instantly create an array of objects. When
generating the array, the user can specify how many objects are created, the
distance between objects as well as additional rotation (etc.).

[Link] Variety Within Repetition

For creating 3D objects with naturalistic or organic feeling, exact duplication often
is an undesirable element. Slight variety of rotation, scale, or color can help throw
off the eye from noticing the repetition.

5.2.8 Proximity

This principle can be applied in 2D and 3D space designs as well temporal formats,
such as film and motion graphics. How objects are placed next to each other in
space and time affects our perception of them as a connected unit. When several
objects are placed close to each other as shown in Fig. 5.26 above, the brain easily
groups them into 5 units. This principle can be connected in a temporal sense to
[11], which takes the images projected next to each other within a time proximity
(between cuts in a film) and creates an association between them. The context and
order that the images are perceived can have a great impact on the emotional impact
(Isenhour).
Visual hierarchy is the order in which humans perceive an image. This order is
created by several factors. Objects with high visual contrast to their surroundings
are perceived first. Visual contrast can be achieved by structuring the elements of
design in ways that make objects stand out. Some examples of objects that are
designed to stand out are warning signs and tennis balls. The bright fluorescent
colors catch the eye. Opposite to this would be a soldier wearing camouflage,
which has the purpose of blending and hiding the person into the background.

[Link] Proximity Tools

Spot Healing Brush: This Photoshop tool analyzes surrounding pixels to check
for similarity within surrounding colors, with the new ‘‘content-aware’’
5.2 Principles of Design 95

technology, and it looks for ‘‘proximity of pattern’’ in the surrounding area and
will attempt to create additional patterns.

[Link] Hierarchy Tools

Using a combination of tools related to the elements of design can help describe
hierarchy:
Scale: The larger the object, the more dominant it can become within the visual
hierarchy.
Contrast: Making the object stand out by contrasting it to what is surrounding it.
Position: Where the object is represented in the frame also can affect hierarchy.
The top of the frame increases importance.

5.2.9 Dynamics

This principle of design is very strongly connected to both audio and movement.
The feeling an image gives is usually related to the use of patterns and line.

[Link] Direction and Movement (Similar to Principle of Staging)

This can be accomplished by the placement of the 3D object so that the profile
creates lines that are diagonal to the frame of the image. Lines can also be a part of
the texture that covers the 3D object creating additional or contrary feelings of
direction or movement within an object.

[Link] Rhythm

Visual rhythm can be created in several ways. Individual lines can create a linear
rhythm. How lines are placed on a page affects the pace of how they are perceived by
the human eye. A strong use of repetition with any of the elements of design can
create a feeling of rhythm within an image. An example of a very simple rhythm
would be a repetition of a singular theme, and using alternation of a differentiated
sequence of repeating motifs can generate a more textured and complex rhythm.

[Link] Arrhythm

Elements presented arrhythmically (without rhythm) can give contrast to areas

with visual rhythm. The word specifically means ‘‘against rhythm.’’ Arrhythmic
96 5 3D Design and Photogrammetry

elements can be used effectively in designs to create emotional feelings which

connect to disharmony.

[Link] Random

Arrhythm taken to an extreme becomes random. Depending on the size of the

random elements, this can be effective as a naturalistic visual element.

References

1. Albers, J.: Interaction of color: Josef Albers. Yale University Press, New Haven. Also
Interaction of Color: 50th Anniversary (app) (2006)
2. Motter, B.C., Belky, E.J.: The guidance of eye movements during active visual search.
Vision. Res. 38(12), 1805–1815 (1998)
3. D’Onofrio, A.: Refining bouba-kiki: Phonetic detail and object dimensionality in sound-
shape correspondences. J. Acoust. Soc. Am. 132(3), 1968 (2012)
4. Cook, T.A: The Curves of Life: Being an Account of Spiral Formations and Their
Applications to Growth in Nature, to Science and to Art with Special Reference to the Ms. of
Leonardo Da Vinci; with 11 Plates. Courier Dover Publications, Mineola (1979)
5. Perlin, K.: Improving noise. ACM Trans. Graph. (TOG) 21(3), 681–682 (2002)
6. Stewart, M.: Launching the Imagination: A Comprehensive Guide to Three-Dimensional
Design. McGraw-Hill, New York (2006)
7. Palmer, S., Rock, I.: Rethinking perceptual organization: The role of uniform connectedness.
Psychon. Bull. Rev. 1(1), 29–55 (1994)
8. Lidwell, W., Holden, K., Butler, J.: Universal Principles of Design: 125 Ways to Enhance
Usability, Influence Perception, Increase Appeal, Make Better Design Decisions, and Tech
Through Design [25 Additional Design Principles]. Rockport publishers (2010)
9. White, A.W.: The elements of graphic design. Skyhorse Publishing Inc, New York (2011)
10. Thornhill, R., Gangestad, S.W.: Human facial beauty. Hum. Nat. 4(3), 237–269 (1993)
11. Isenhour, J.P.: The effects of context and order in film editing. Education Tech. Research
Dev. 23(1), 69–80 (1975)
Chapter 6
Conclusions and Looking Forward

Abstract This book concludes by looking at additional market leaders and how
they are further integrating technology for workflows as well as adapting and
converging.

Keywords Conclusions

As we have shown in this book, the development of 3D computer modeling

workflows has undergone substantial changes in its short history. At any point in
this development cycle, a case can be made that a ‘‘paradigm shift’’ has taken
place—the transition from NURBS to polygon modeling, for example, or the
ability to represent fine detail with normal maps.
The integration of photogrammetry, high-polygon-density interactive modeling,
and retopology represents a sweeping change that some might say will completely
overturn traditional approaches. They have certainly represented a major change in
approach for many large studios. As the tools for this workflow have proliferated
and become more accessible to more people at a lower price point, the process has
become the norm for an increasing number of smaller studios, independent artists,
and education institutions.
As this change in tools happens (as discussed in the chapter on the nature of
technology), adoption speed of new tools is also accelerating. As these new
technologies and tools become part of the status quo, we anticipate greater fidelity
within our 3D models and almost automatic retopologization within the next few
years to be ubiquitous.
Looking to Entertainment VFX market leader Weta Digital, over the last two
years, Weta upgraded the photogrammetry pipeline and added high-fidelity
LIDAR (laser scanning). For the new Hobbit films, they record ‘‘much more…
total data than ever before… for example, Avatar was approximately one petabyte,
now it is over six times that for the Hobbit.’’ They have reached the point where
they are ‘‘scanning every set that they are shooting on’’ using a Leica ScanStation
C10, and they can scan 50 million points per scan in 30 min, and about 109 scans
per set, plus high dynamic rage photographs.

S. Foster and D. Halbstein, Integrating 3D Modeling, Photogrammetry and Design, 97

SpringerBriefs in Computer Science, DOI: 10.1007/978-1-4471-6329-9_6,
 The Author(s) 2014
98 6 Conclusions and Looking Forward

As the technology for handling this quantity of data reaches consumer price
points, this opens up further design possibilities. Higher-fidelity visual 3D scans
from real life; using design principles to sculpt and retopologize output in a
multitude of forms will become standard and semiautomated workflows.
With rapid optimization and automation of 3D scanning and modeling, we
anticipate further convergence among technical and design fields. For example,
combining scanning with motion capture, augmented reality, and visual effects [1],
3D models will be combined with future generations of tools for designers to
create with.

Reference

1. Seymour, M.: The hobbit: Weta returns to middle-earth, FXGuide [Link]

featured/the-hobbit-weta/(December 2012)
Chapter 7
Resources and Application

Abstract Resources chapter—links for photogrammetry and design.

Keywords Resources  Photogrammetry  3D scanning  Color

7.1 PhotogrammetryResources

123D App Autodesk

[Link]
Arc3D
[Link]
Aero Scan
[Link]
Hypr3D
[Link]
My3D Scanner
[Link]
Microsoft Kinect SDK
[Link]
Microsoft Photosynth
[Link]
To download the Photosynth application:
[Link]
Insight 3D
[Link]
International Society for Photogrammetryand Remote Sensing
[Link]
OSM Bundler
[Link]

S. Foster and D. Halbstein, Integrating 3D Modeling, Photogrammetry and Design, 99

SpringerBriefs in Computer Science, DOI: 10.1007/978-1-4471-6329-9_7,
 The Author(s) 2014
100 7 Resources and Application

PhotoModeler Software
[Link]
ReconstructMe—Microsoft Kinect for 3D digitizing of objects
[Link]
VisualSFM: University of Seattle
[Link]
CMP SfM Web Service (Europe)
[Link]
Structure from Motion Toolkit
[Link]
Visual Size
[Link]
Good Forums
Photosynth Rivals
[Link]
Other Application:
PhotoCity (game)
[Link]
3D Scanningtechnologies
[Link]
CNO4neLpl7YCFalxOgodpE0Anw
CyArk: Digital preservation of world heritage sites
[Link]

7.2 Retopologizing Resources

Autodesk Maya (2014)

[Link]
Advanced Modeling Workflow: Retopo tools: Quad-draw, shrink wrap
[Link]
Autodesk Mudbox
[Link]
3D Coat
[Link]
Topogun
[Link]
Pixologic: ZBrush
[Link]
7.3 Color Resources 101

7.3 Color Resources

[Link]
Albers, Joseph Interaction of Color (iPad app and book)
ColrD—Free Color Chrome App

References

1. Richards, J.A.: Remote sensing digital image analysis: An introduction (09/13/2012)(ISBN:3-

642-30061-8, 978-3-642-30061-5)
2. [Link] Photogrammetry history p2
3. Dyer: Mitch Xbox’s Kinect one is legitimately awesome May 2013[Link]
articles/2013/05/22/xbox-ones-kinect-is-legitimately-awesome
4. Reeves, B-GameInformer: Infamous second son: How sucker punch photocopies pedestrians
for infamous second son May 2013[Link]
son/b/playstation4/archive/2013/05/24[Link]
[Link]
5. Seymore, M.: The hobbit Weta returns to middle-earth, FXGuide[Link]
featured/the-hobbit-weta/
6. Zbrush Webinar–ZRemesherTM [Link]
Index

0–9 F
3D coat, 19, 50 Film, 83, 86, 94
3D design, 1, 3, 69, 72, 75, 78, 83, 86 Future studies, 5
3D scanning, 98

G
A Gestalt, 74, 77, 86, 88
Additive, 20, 53
Art and design, 10
Asymmetry, 91 I
Autodesk 123D Catch, 13, 59 Introduction, 1, 2

B K
Balance, 83, 86, 91 Kinect fusion, 11
Box modeling, 30, 52

L
C Langer, 33, 34
Camera, 75, 78, 86, 87 Lens, 86, 87
Challenges and limitations, 13 Line, 69, 74–77, 87, 91, 95
Cinematography, 84, 85, 87
Color, 69–77, 83, 88–90, 92, 94, 101
Computer vision, 2, 3 M
Conclusion, 97 Major technology, 5
Convergence, 1, 83, 85, 98 Maya, 37, 64
Microsoft kinect, 10, 11
Modeling, 17–20, 23, 28, 30, 36, 42, 49–51, 56
D Modeling toolkit, 64
Design process, 4 Mudbox, 19, 40, 42
Digital sculpture, 19, 40, 51, 55, 56
Dynamics, 86, 90
N
Nature of technology, 5
E Normal maps, 37
Edge flow, 19, 20, 30, 34, 46, 50, 55, 56 NURBS, 20, 21, 23, 28–30, 41, 53
Education, 1
Elements of design, 69, 87, 91, 94, 95
Emphasis, 86, 90, 91 P
Entertainment field, 1 Patch modeling, 23, 30

S. Foster and D. Halbstein, Integrating 3D Modeling, Photogrammetry and Design, 103

SpringerBriefs in Computer Science, DOI: 10.1007/978-1-4471-6329-9,
 The Author(s) 2014
104 Index

Pattern, 75, 76, 79, 80, 95 Subdivisions, 65

Photogrammetry, 1, 2, 6, 7, 9–13, 99 Subtractive, 55, 56
Photogrammetry history, 7 Surface parameterization, 57
Photospace, 13 Symmetry, 91
Photosynth, 12
Polygons, 19–21, 36, 40, 43, 46, 50, 54–56
Principles of design, 69, 85, 88 T
Proximity, 88, 94 Technology cycles, 5
Texture, 69, 78–82, 88, 91, 95
Topology, 19, 43, 51–53, 55, 56
Q
Quads, 65
U
Unity, 73, 85, 86, 88, 91
R UV layout, 20, 42
Repetition, 86, 89, 93, 95
Resources, 99, 100
Retopology, 1, 19, 42, 45, 46, 49, 51, 52, 55 W
What a digital, 9, 10, 13
Workflows, 19, 20
S
Scan, 97
Scanning, 97 Z
Shape, 69, 75–77, 84, 90, 91 Zbrush, 18, 38, 40, 42, 51–53, 55, 56
Space, 69, 70, 77, 78, 84, 87–89, 92