Exploring Digital Image
Processing
This presentation provides an introduction to Digital Image Processing (DIP),
covering its origins, fundamental concepts, and diverse applications. We will
explore how digital images are processed, from their acquisition to
advanced techniques like pattern recognition and medical imaging.
by Sowmya Balakrishnan
What is Digital Image Processing?
Definition Digital Image
Digital Image Processing (DIP) involves using digital computers A digital image is a two-dimensional function f(x,y) where x, y,
to process digital images. It's a discipline where both the input and the amplitude values of f are all finite, discrete quantities.
and output are images, allowing for manipulation and analysis. This discrete nature is fundamental to how images are stored
and processed by computers.
An image is a two-dimensional light intensity function f(x, y),
where x and y are spatial coordinates, and f represents the
intensity or brightness at that point.
The Evolution of Digital Image Processing
Early 1920s
First application in newspaper industry: pictures sent
via submarine cable between London and New York,
reducing transmission time from over a week to under
1960s
3 hours.
Improvements in computing technology and the space
race led to a surge in DIP research and applications.
1970s
DIP began to be used in medical applications,
revolutionizing diagnostics.
1979
Sir Godfrey N. Hounsfield & Prof. Allan M. Cormack
shared the Nobel Prize for inventing tomography, the
1980s - Today technology behind CAT scans.
Explosion in DIP techniques for image enhancement,
medical visualization, industrial inspection, law
enforcement, and human- computer interfaces.
Diverse Applications of Digital Image
Processing
Image Sharpening Medical Field Machine/Robot
& Restoration Vision
Applications like Gamma
Enhancing images from ray imaging, PET scans, Enabling robots to see,
modern cameras, X-ray imaging, and identify objects, and
including zooming, Medical CT for detect hurdles,
blurring, sharpening, and diagnostics. contributing to the field of
color conversion. computer vision.
Color Processing
Processing colored
images, studying color
models (RGB, YCbCr,
HSV), and managing their
transmission and storage.
DIP extends beyond simple photo adjustments, impacting nearly all technical fields. It analyzes
electromagnetic waves, including those invisible to the human eye, such as X-rays and gamma rays,
providing deeper insights.
Fundamental Steps in Digital Image Processing
Image Acquisition
The initial step, involving obtaining a digital image, often including preprocessing like scaling.
Image Enhancement
Subjective techniques to improve image appearance or highlight features, like increasing contrast.
Image Restoration
Objective methods based on mathematical models to remove noise and restore original image quality.
Compression
Techniques to reduce storage size or bandwidth for images, familiar through formats like JPEG.
Segmentation
Partitioning an image into constituent parts or objects, a critical and often challenging task.
Recognition
Assigning a label to an object based on its descriptors, often following segmentation.
Components of an Image Processing System
Sensing & Digitizer Specialized Hardware
Physical devices sensitive to energy (e.g., photodiodes) and Includes digitizers and ALUs for parallel arithmetic and
digitizers converting output to digital form. logical operations on images, ensuring fast data throughput.
Computer & Software Mass Storage & Displays
A general-purpose computer (PC to supercomputer) running Essential for storing large image data (short-term, on-line,
specialized image processing modules. archival) and color TV monitors for display.
Elements of Visual Perception: The Human Eye
The human eye, a nearly spherical organ, is enclosed by three
membranes: the cornea and sclera (outer cover), the choroid
(middle layer with blood vessels), and the retina (innermost
membrane).
The retina contains two types of receptors: cones (6-7 million,
located in the fovea, for bright-light and color vision) and rods
(75-150 million, distributed over the retinal surface, for dim-light
vision without color).
Brightness Adaptation and
Discrimination
Dynamic Range
The human visual system can adapt to an enormous range of light intensities
(on the order of 10^10), from scotopic threshold to glare limit.
Brightness Adaptation
The eye adjusts its overall sensitivity to perceive this wide range, a
phenomenon known as brightness adaptation.
Contrast Sensitivity
The ability to discriminate between brightness changes is crucial. Weber's
Ratio (ΔIc/I) indicates brightness adaptation: small ratios mean good
adaptation, large ratios mean poor adaptation.
Optical Illusions
Phenomena like Mach Bands (constant intensity stripes appearing to have
gradients) and Simultaneous Contrast (inner squares of same intensity
appearing brighter against lighter backgrounds) highlight how perception is
influenced by context.
Light and the Electromagnetic Spectrum
Gamma Rays & X-rays
Highest energy, shortest wavelength. Used in medical and industrial imaging.
Ultraviolet & Visible Light
Visible light (0.43µm to 0.79µm) is a small portion. UV is higher energy.
Infrared
Radiates heat, useful for "heat signatures" in imaging applications.
Microwaves & Radio Waves
Longest wavelengths, lowest energy. Used in communication, radar,
and broadcasting.
The electromagnetic spectrum encompasses a vast range of energy, from radio waves to gamma rays. Each part of the spectrum has unique
properties and applications in imaging, depending on the wavelength and photon energy.
Image Sensing and Acquisition
Digital images are generated by sensing energy reflected or transmitted from a scene. This involves physical devices (sensors) that
convert energy into voltage, which is then digitized. Common sensor arrangements include single sensors (e.g., photodiodes),
sensor strips (e.g., flatbed scanners, airborne imaging), and 2-D sensor arrays (e.g., digital cameras). Each method has specific
applications and limitations in image acquisition.
