Skip to main content

Anupam Agrawal

IIIT Allahabad, Information Technology, Faculty Member

Followers

119

Following

4

Public Views

Phone: +91 9455480865
Address: Room no. 4123, CC2, IIIT Allahabad, Jhalwa, Prayagraj (UP) - 211 015, India.

less

Sirvan Khalighi

Indonesian Journal of Electrical Engineering and Computer Science

Francis Xavier Engineering College

Richa Srivastava

Anna University of Technology, Coimbatore

Manish Chowdhury

IJCST Eighth Sense Research Group

Anna University

Interests

Uploads

Papers by Anupam Agrawal

Framework for Segmentation of SAR imagery based on NSCT and IABC Algorithm

Conventional methods for segmentation on Synthetic Aperture Radar (SAR) images aim to provide aut... more Conventional methods for segmentation on Synthetic Aperture Radar (SAR) images aim to provide automatic analysis and interpretation of data.Usually the task gets failed at many cases due to the influence of speckle in the segmented results. The present work proposes a Hybrid framework for segmentation which will provide approachable segmentation results in SAR imagery for the post processing. This framework uses the full advantage of Non Subsampled Contourlet to achieve promising results. The final resultsare analyzed using different metrics and it shows that our proposed framework is providing promoting results in Segmentation of SAR images.

A Soft Computing Technique for Improving the Fidelity of Thumbprints Based Identification Systems

International Journal of Intelligent Systems and Applications, 2016

With the advent of new thumbprint identification techniques, accurate personal identificat ion is... more With the advent of new thumbprint identification techniques, accurate personal identificat ion is now easy and cheaper with appro ximately zero false acceptance rates. This paper focuses on developing an advance feature for thumbprint based identificat ion systems with the help of soft computing and 2D transformation which makes the technique more flexib le and Fidel. The thu mbprint images of individuals were scanned with the help of H3 T&A terminal for co llect ing self generated datasets. The thumbprints of self generated and standard datasets were trained to form a refined set which includes linear and angular displacements of thumbprint images. The new obtained features of refined datasets were stored in the database for further identification. In the proposed technique, the minutiae coordinates and orientation angles of the thumbprint of a person to be identified are co mputed and merged together for comparison. The minutia coordinates and orientation angles of a person are compared with the minutiae trained set values stored in the database at different linear and angular rotations for identity verification. The proposed technique was tested on fifty persons self generated and standard datasets of FVC2002, FVC2004 and CASIA databases. In the experimentation and result analysis we observed that the proposed technique accurately identifies a person on the basis of minutiae features of a thumbprint with lo w FNM R (False Non-Match Rate) values.

A novel hybrid intelligence approach for 2D packing through internet crowdsourcing

2015 IEEE Technological Innovation in ICT for Agriculture and Rural Development (TIAR), 2015

Packing problems on its current state are being utilized for wide area of industrial applications... more Packing problems on its current state are being utilized for wide area of industrial applications. The aim of present research is to create and implement an intelligent system that tackles the problem of 2D packing of objects inside a 2D container, such that objects do not overlap and the container area is to be maximized. The packing problem becomes easier, when regular/rectangular objects and container are used. In most of the practical situations, the usage of irregular objects comes to existence. To solve the packing problem of irregular objects inside a rectangular container, a hybrid intelligence approach is introduced in our proposed work. The combination of machine intelligence and human intelligence is referred as the hybrid intelligence or semi-automated approach in the proposed methodology. The incorporation of human intelligence in the outcome of machine intelligence is possible to obtain using the internet crowdsourcing as we wish to handle the packing problem through internet crowdsourcing involving rural people. The proposed methodology is tested on different standard data sets and it is observed that it has clear advantage over both manual as well as fully automated heuristic based methods in terms of time and space efficiency.

Palatal Patterns Based RGB Technique for Personal Identification

International Journal of Image, Graphics and Signal Processing, 2016

Biometric system is an alternative way to the traditional identity verification methods. This res... more Biometric system is an alternative way to the traditional identity verification methods. This research article provides an overview of recently / currently used single and multiple biometrics based personal identification systems which are based on human physiological (such as fingerprint, hand geometry, head recognition, iris, retina, face recognition, DNA recognition, palm prints, heartbeat, finger veins, footprints and palates) and behavioral (such as body language, facial expression, signature verification and speech recognition) characteristics. This paper focuses on RGB based palatal pattern analysis of persons and the proposed technique uses RGB values with silhouette computes of palatal patterns for identifying a person. We have tested our proposed technique for palatal patterns of 50 persons including males & females and it is observed that RGB values based silhouette technique are accurately identifying the persons on the basis of their palatal patterns. For each person seven palatal images were taken. Out of these seven palatal images, four images were used for training dataset and last three palatal patterns were used for identifying the persons. The proposed technique is reliable & secure and it is a foolproof method which is clearly differentiating the persons on the basis of their palatal patterns.

Hardware Accelerated Multi-coordinate Viewing Framework for Volumetric Visualization of Large 3D Medical Dataset

Procedia Computer Science, 2015

The advances in Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) scanning techniques... more The advances in Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) scanning techniques are improving the resolution and size of the volume datasets. The prevalence of three dimensional volumetric data is rapidly expanding with the internal information in the different resolution sizes of the dataset. In this paper we proposed an approach that can visualize the inner organs structure of the visible human male dataset in Multi-coordinate Viewing (MCV) framework. This can help to medical experts that are able to peer inside anatomy of the human medical dataset. The volume rendering part has been carried out by the utilization of enhanced ray casting algorithm for the crossing points of 3D square strategy for voxels. We present this system using Graphics Processing Unit or GPU-accelerated Compute Unified Device Architecture (CUDA) based approach for the focusing a specific region while zooming operation. The final results would allow the doctors to diagnose and analyze the atlas of 8-bit CT-scan data using three dimensional visualization with the efficient frame rate rendering speed in multi-operations like zooming, rotating, dragging. The framework is tested for visible human male dataset prepared by National Library of Medicine (NLM, USA) of size 1.2 GB.

Video Summary Based on F-Sift, Tamura Textural and Middle Level Semantic Feature

Procedia Computer Science, 2016

The continuous generation of digital multimedia has raised the question about management of time ... more The continuous generation of digital multimedia has raised the question about management of time and data. For effective management of such a tremendous amount of data or longer videos, a number of researches have been made. Video summarization is one of important research area to solve the problem. To summarise a video of any domain, researches relied on visual features available in frames, which does not guarantee to achieve semantic meaning of original video. In this paper, we proposed an approach for getting semantic video summary of original video. Generated video summaries are based on middle level semantic features, flip-Sift and tamura's textural feature. We used clustering approach based on middle level semantic features for static key frame extraction. Tamura feature has been used for getting texture features of frames. We tested this technique on various domains like sports, news, etc.; this technique has been tested on dataset freely available at YAACVID, UT egocentric and benchmark datasets. Longer duration video is also producing good result in compare to existing technique.

Social implications of crowd sourcing in rural Scotland

Various surveys mentioned that the benefits of ubiquitous crowdsourcing are reaped by people loca... more Various surveys mentioned that the benefits of ubiquitous crowdsourcing are reaped by people located in metro and smaller cities. The reach of crowdsourcing to rural population is questionable. The aim of this research is to bridge widening urban and rural divide by providing knowledge-intensive crowdsourcing tasks to rural work force which could provide long term benefits to them as well as improve supporting infrastructure. This paper reports an initial study of the demographic of small samples of twenty two rural homeworkers in Scotland, their motivation to do crowdsourcing work, present main occupation, computer skills, views on rural infrastructure and finally their skills on solving three spatial visualization tests. The survey shows that flexible hours of working, extra income, and work life balance are the three important factors emphasized as motivational constructs to do crowdsourcing work. Their skills on solving a spatial visualization test is equivalent to the literatur...

Out-Of-Core Construction and 3D Visualization of Level-Of-Detail Terrains Populated with Large Collection of Heterogeneous Objects

Proceedings of the First International Conference on Computer Graphics Theory and Applications, 2006

Interactive visualization of very large-scale terrain data in scientific visualization, GIS or si... more Interactive visualization of very large-scale terrain data in scientific visualization, GIS or simulation and training applications is a hard problem. The grid digital terrain elevation and texture data are not only too large to be rendered in real-time but also exceed physical main memory capacity. Therefore out-of-core management of digital terrain data is an essential requirement. Further to bring photorealism in visualization, it is required to place multiple collections of man-made objects such as buildings, lampposts etc. as well as natural objects such as trees, grass etc. on top of the terrain surface. In this paper we have proposed an integrated approach for effective out-of-core visualization of terrains populated with large collection of static heterogeneous objects. We have developed an efficient tile-based out-of-core viewdependent Level of Detail (LOD) mesh simplification algorithm for real-time rendering of large terrains. Instead of manipulating individual triangles, the algorithm operates on clusters of geometry called blocks of aggregate triangles. Hence the amount of work CPU must perform is greatly reduced. The formation of long triangle strips for LOD blocks also solves the CPU-to-Card bandwidth problem. The tile-based multiresolution terrain geometry framework has been extended to support large satellite or aerial imagery textures. To display large collection of objects over the terrain while maintaining the real-time frame rate, an efficient object handling method has been proposed using paging technique and object instantiation. User is allowed to control the objects locations, scales and orientations. The algorithms have been implemented using Visual C++ and OpenGL 3D API and successfully tested on different real-world height maps and satellite phototextures of sizes upto 16K*16K coupled with thousands of static objects on PCs.

Move Table: An Intelligent Software Tool for OptimalPath Finding and Halt Schedule Generation

Defence Science Journal, Sep 21, 2007

This study aims to help army officials in taking decisions before war to decide the optimal path ... more This study aims to help army officials in taking decisions before war to decide the optimal path for army troops moving between two points in a real world digital terrain, considering factors like traveled distance, terrain type, terrain slope, and road network. There can optionally be one or more enemies (obstacles) located on the terrain which should be avoided. A tile-based A* search strategy with diagonal distance and tie-breaker heuristics is proposed for finding the optimal path between source and destination nodes across a real-world 3-D terrain. A performance comparison (time analysis, search space analysis, and accuracy) has been made between the multiresolution A* search and the proposed tile-based A* search for large-scale digital terrain maps. Different heuristics, which are used by the algorithms to guide these to the goal node, are presented and compared to overcome some of the computational constraints associated with path finding on large digital terrains. Finally, a halt schedule is generated using the optimal path, weather condition, moving time, priority and type of a column, so that the senior military planners can strategically decide in advance the time and locations where the troops have to halt or overtake other troops depending on their priority and also the time of reaching the destination.

Devanagari Character Recognition towards natural Human-Computer Interaction

Electronic Workshops in Computing, 2010

Human-computer interaction is a growing research area. There are several ways of interaction with... more Human-computer interaction is a growing research area. There are several ways of interaction with the computer. Handwriting has continued to persist as a means of communication and recording information in the day to day life even with the introduction of new technologies. Due to the growth of technology in India, it becomes important to devise ways that allow people to communicate with computer in Indian languages. Hindi being the national language of India, we present a way to communicate with the computer in Hindi or more precisely, 'Devanagari script'. Due to absence of a global font to represent Devanagari characters, it is important that the computer recognizes the characters written by the user in order to interact with him. The algorithm implemented for character recognition first segments the image containing Devanagari text fed to the software into lines, lines to words and words to characters. The obtained characters are then brought down to a standard size. The Kohonen Neural Network based recognizer then comes into action and recognizes the text character by character and provides the output in Unicode format. The network has been designed with no hidden layer to support quick recognition. Apart from text recognition from an image, we also provided the option to recognize individual handwritten characters drawn using a mouse. Such a system provides keyboard less computer interaction. The technique is implemented using Java. The overall recognition rate for a fixed font machine printed characters is 90.26% and for hand written characters, it is 83.33%.

Evaluation schema for SAR image Segmentation based on Swarm Optimization in Neutrosophic domain

2014 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT), 2014

The present paper proposes an evaluation schema for SAR image Segmentation. The segmentation of S... more The present paper proposes an evaluation schema for SAR image Segmentation. The segmentation of SAR images becomes a crucial step because the segmented results will be used as the input for the post-processing like target detection, change detection applications. The preprocessing task segmentation is highly influenced by inherent noise known as speckle. The proposed system introduces the evaluation schema for segmentation of SAR images in Neutrosophic domain, which overcome the problem of speckle. The system is evaluated for different Swarm Optimization algorithms and the promoting results are obtained through the proposed evaluation schema. The results are discussed on the basis of different parameters.

Fusion of Multi-Sensor Satellite Images Using Non-Subsampled Contourlet Transform

Procedia Computer Science, 2015

The presented research work proposes fusion of multi-sensor satellite images using non subsampled... more The presented research work proposes fusion of multi-sensor satellite images using non subsampled contourlet transform. In the proposed work, trade-off between the spectral distortion and enhancement of spatial information is witnessed while fusing two multi-sensor images. The ills of wavelet based fusion techniques such as limited directionality, lack of phase information and shift invariant are addressed with the help of Non subsampled contourlet transform. The Non subsampled contourlet helps to retain the intrinsic structural information while decomposing and reconstructing the image components. Decision based rules are applied for component substitution for fusion. The experiments are carried out against the current state of art and observed that the proposed system provides promising results visually and quantitatively. The efficiency of the proposed system in the fused product is analysed qualitatively by Isodata classification algorithm.

Crowdsourcing solutions to 2D irregular strip packing problems from Internet workers

International Journal of Production Research, 2015

Many industrial processes require the nesting of 2D profiles prior to the cutting, or stamping, o... more Many industrial processes require the nesting of 2D profiles prior to the cutting, or stamping, of components from raw sheet material. Despite decades of sustained academic effort, algorithmic solutions are still sub-optimal and produce results that can frequently be improved by manual inspection. However, the Internet offers the prospect of novel 'human-in-the-loop' approaches to nesting problems that uses online workers to produce packing efficiencies beyond the reach of current CAM packages. To investigate the feasibility of such an approach, this paper reports on the speed and efficiency of online workers engaged in the interactive nesting of six standard benchmark data-sets. To ensure the results accurately characterise the diverse educational and social backgrounds of the many different labour forces available online, the study has been conducted with subjects based in both Indian IT service (i.e. Rural BPOs) centres and a network of homeworkers in Northern Scotland. The results (i.e. time and packing efficiency) of the human workers are contrasted with both the baseline performance of a commercial CAM package and recent research results. The paper concludes that online workers could consistently achieve packing efficiencies roughly 4% higher than the commercial based-line established by the project. Beyond characterising the abilities of online workers to nest components, the results also make a contribution to the development of algorithmic solutions by reporting new solutions to the benchmark problems and demonstrating methods for assessing the packing strategy employed by the best workers.

CORSISCA : classification of remotely sensed images - a soft computing approach

Proceedings of the IEEE INDICON 2004. First India Annual Conference, 2004.

The classification of multispectral satellite images is a challenging problem and has a number of... more The classification of multispectral satellite images is a challenging problem and has a number of applications such as feature identification, change detection, etc. We apply modified Neural Network Algorithms: GA-BP (Genetic Algorithm as precursor to the Back Propagation) and Modular Artificial Neural Network (MNN) to classify the LISS-3 image of Allahabad area. We also classify the resolution merged image (LISS-3 with PAN) using the same algorithms. By using Genetic algorithm as a precursor to ANN, we increase the probability of reaching to the global minimum, thus reducing the problem of a stuck neural network in the local minimum. MNN models the human brain more closely to apply task decomposition to the satellite images as well. The output of the above techniques are generated and analyzed.

An Assessment of the Impact of Dimensionality Reduction on the Speed and Accuracy of Hyperspectral Image Classification

This paper investigates the extent to which the accuracy and speed of classifying Hyperspectral r... more This paper investigates the extent to which the accuracy and speed of classifying Hyperspectral remote sensing images are affected by the application of varying degrees of dimensionality reduction. Three methods have been used for dimensionality reduction: PCA, ICA and random band selection; SVM has been used for classification. The results have been evaluated on both natural and man-made scenarios.

Towards crowdsourcing spatial manufacturing tasks from rural India

Many tasks involving production software that involve spatial reasoning can only produce "good" r... more Many tasks involving production software that involve spatial reasoning can only produce "good" rather than optimum solutions. It is frequently possible for humans to improve on algorithmically generated solutions computed by CAD/CAM software. The thesis motivating this research is that large numbers of industrial optimisation tasks involving spatial reasoning (such as 2D part nesting) can be outsourced as human intelligence tasks to rural workers to provide a sustainable source of skilled employment. We hypothesis that 3D spatial reasoning ability is essential to solve spatial optimization tasks. In this paper we reported the 3D spatial manipulation ability of eighty rural workers in four rural business process outsourcing (BPO) centres in India. The assessments reveal that although the average spatial manipulation ability is less than the literature reported benchmark, there are talented workers identified in all the four rural centres, and the results identify priority activities required to enable the proposed approach.

Real Time Multiple Hand Gesture Recognition System for Human Computer Interaction

International Journal of Intelligent Systems and Applications, 2012

With the increasing use of computing devices in day to day life, the need of user friendly interf... more With the increasing use of computing devices in day to day life, the need of user friendly interfaces has lead towards the evolution of different types of interfaces for human computer interaction. Real time vision based hand gesture recognition affords users the ability to interact with computers in more natural and intuitive ways. Direct use of hands as an input device is an attractive method which can communicate much more information by itself in comparison to mice, joysticks etc allowing a greater number of recognition system that can be used in a variety of human computer interaction applications. The gesture recognition system consist of three main modules like hand segmentation, hand tracking and gesture recognition from hand features. The designed system further integrated with different applications like image browser, virtual game etc. possibilities for human computer interaction. Computer Vision based systems has the potential to provide more natural, non-contact solutions. The present research work focuses on to design and develops a practical framework for real time hand gesture Index Terms-Real time, gesture recognition, human computer interaction, tracking.

Nonparametric Neural Network Model Based on Rough-Fuzzy Membership Function for Classification of Remotely Sensed Images

Lecture Notes in Computer Science, 2006

A nonparametric neural network model based on Rough-Fuzzy Membership function, multilayer percept... more A nonparametric neural network model based on Rough-Fuzzy Membership function, multilayer perceptron, and back-propagation algorithm is described. The described model is capable to deal with rough uncertainty as well as fuzzy uncertainty associated with classification of remotely sensed multi-spectral images. The input vector consists of membership values to linguistic properties while the output vector is defined in terms of rough fuzzy class membership values. This allows efficient modeling of indiscernibility and fuzziness between patterns by appropriate weights being assigned to the backpropagated errors depending upon the Rough-Fuzzy Membership values at the corresponding outputs. The effectiveness of the model is demonstrated on classification problem of IRS-P6 LISS IV images of Allahabad area. The results are compared with statistical (Minimum Distance), conventional MLP, and FMLP models.

Real Time Gesture Recognition System for Interaction in Dynamic Environment

Procedia Technology, 2012

Human Computer Interaction techniques have become a bottleneck in the effective utilization of th... more Human Computer Interaction techniques have become a bottleneck in the effective utilization of the available information flow. The development of user interfaces influences the changes in the Human-Computer Interaction (HCI). Human hand gestures have been a mode of non verbal interaction widely used. Naturalistic and intuitiveness of the hand gesture has been a great motivating factor for the researchers in the area of HCI to put their efforts to research and develop the more promising means of interaction between human and computers. This paper designs a system for gestural interaction between a user and a computer in dynamic environment. The gesture recognition system uses image processing techniques for detection, segmentation, tracking and recognition of hand gestures for converting it to a meaningful command. The interface being proposed here can be substantially applied towards different applications like image browser, games etc.

A Vision based Hand Gesture Interface for Controlling VLC Media Player

International Journal of Computer Applications, 2010

Human Computer Interaction can acquire several advantages with the introduction of different natu... more Human Computer Interaction can acquire several advantages with the introduction of different natural forms of device free communication. Gestures are a natural form of actions which we often use in our daily life for interaction, therefore to use it as a communication medium with computers generates a new paradigm of interaction with computers. This paper implements computer vision and gesture recognition techniques and develops a vision based low cost input device for controlling the VLC player through gestures. VLC application consists of a central computational module which uses the Principal Component Analysis for gesture images and finds the feature vectors of the gesture and save it into a XML file. The Recognition of the gesture is done by K Nearest Neighbour algorithm. The theoretical analysis of the approach shows how to do recognition in static background. The Training Images are made by cropping the hand gesture from static background by detecting the hand motion using Lucas Kanade Pyramidical Optical Flow algorithm. This hand gesture recognition technique will not only replace the use of mouse to control the VLC player but also provide different gesture vocabulary which will be useful in controlling the application.

Framework for Segmentation of SAR imagery based on NSCT and IABC Algorithm

Conventional methods for segmentation on Synthetic Aperture Radar (SAR) images aim to provide aut... more Conventional methods for segmentation on Synthetic Aperture Radar (SAR) images aim to provide automatic analysis and interpretation of data.Usually the task gets failed at many cases due to the influence of speckle in the segmented results. The present work proposes a Hybrid framework for segmentation which will provide approachable segmentation results in SAR imagery for the post processing. This framework uses the full advantage of Non Subsampled Contourlet to achieve promising results. The final resultsare analyzed using different metrics and it shows that our proposed framework is providing promoting results in Segmentation of SAR images.

A Soft Computing Technique for Improving the Fidelity of Thumbprints Based Identification Systems

International Journal of Intelligent Systems and Applications, 2016

With the advent of new thumbprint identification techniques, accurate personal identificat ion is... more With the advent of new thumbprint identification techniques, accurate personal identificat ion is now easy and cheaper with appro ximately zero false acceptance rates. This paper focuses on developing an advance feature for thumbprint based identificat ion systems with the help of soft computing and 2D transformation which makes the technique more flexib le and Fidel. The thu mbprint images of individuals were scanned with the help of H3 T&A terminal for co llect ing self generated datasets. The thumbprints of self generated and standard datasets were trained to form a refined set which includes linear and angular displacements of thumbprint images. The new obtained features of refined datasets were stored in the database for further identification. In the proposed technique, the minutiae coordinates and orientation angles of the thumbprint of a person to be identified are co mputed and merged together for comparison. The minutia coordinates and orientation angles of a person are compared with the minutiae trained set values stored in the database at different linear and angular rotations for identity verification. The proposed technique was tested on fifty persons self generated and standard datasets of FVC2002, FVC2004 and CASIA databases. In the experimentation and result analysis we observed that the proposed technique accurately identifies a person on the basis of minutiae features of a thumbprint with lo w FNM R (False Non-Match Rate) values.

A novel hybrid intelligence approach for 2D packing through internet crowdsourcing

2015 IEEE Technological Innovation in ICT for Agriculture and Rural Development (TIAR), 2015

Packing problems on its current state are being utilized for wide area of industrial applications... more Packing problems on its current state are being utilized for wide area of industrial applications. The aim of present research is to create and implement an intelligent system that tackles the problem of 2D packing of objects inside a 2D container, such that objects do not overlap and the container area is to be maximized. The packing problem becomes easier, when regular/rectangular objects and container are used. In most of the practical situations, the usage of irregular objects comes to existence. To solve the packing problem of irregular objects inside a rectangular container, a hybrid intelligence approach is introduced in our proposed work. The combination of machine intelligence and human intelligence is referred as the hybrid intelligence or semi-automated approach in the proposed methodology. The incorporation of human intelligence in the outcome of machine intelligence is possible to obtain using the internet crowdsourcing as we wish to handle the packing problem through internet crowdsourcing involving rural people. The proposed methodology is tested on different standard data sets and it is observed that it has clear advantage over both manual as well as fully automated heuristic based methods in terms of time and space efficiency.

Palatal Patterns Based RGB Technique for Personal Identification

International Journal of Image, Graphics and Signal Processing, 2016

Biometric system is an alternative way to the traditional identity verification methods. This res... more Biometric system is an alternative way to the traditional identity verification methods. This research article provides an overview of recently / currently used single and multiple biometrics based personal identification systems which are based on human physiological (such as fingerprint, hand geometry, head recognition, iris, retina, face recognition, DNA recognition, palm prints, heartbeat, finger veins, footprints and palates) and behavioral (such as body language, facial expression, signature verification and speech recognition) characteristics. This paper focuses on RGB based palatal pattern analysis of persons and the proposed technique uses RGB values with silhouette computes of palatal patterns for identifying a person. We have tested our proposed technique for palatal patterns of 50 persons including males & females and it is observed that RGB values based silhouette technique are accurately identifying the persons on the basis of their palatal patterns. For each person seven palatal images were taken. Out of these seven palatal images, four images were used for training dataset and last three palatal patterns were used for identifying the persons. The proposed technique is reliable & secure and it is a foolproof method which is clearly differentiating the persons on the basis of their palatal patterns.

Hardware Accelerated Multi-coordinate Viewing Framework for Volumetric Visualization of Large 3D Medical Dataset

Procedia Computer Science, 2015

The advances in Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) scanning techniques... more The advances in Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) scanning techniques are improving the resolution and size of the volume datasets. The prevalence of three dimensional volumetric data is rapidly expanding with the internal information in the different resolution sizes of the dataset. In this paper we proposed an approach that can visualize the inner organs structure of the visible human male dataset in Multi-coordinate Viewing (MCV) framework. This can help to medical experts that are able to peer inside anatomy of the human medical dataset. The volume rendering part has been carried out by the utilization of enhanced ray casting algorithm for the crossing points of 3D square strategy for voxels. We present this system using Graphics Processing Unit or GPU-accelerated Compute Unified Device Architecture (CUDA) based approach for the focusing a specific region while zooming operation. The final results would allow the doctors to diagnose and analyze the atlas of 8-bit CT-scan data using three dimensional visualization with the efficient frame rate rendering speed in multi-operations like zooming, rotating, dragging. The framework is tested for visible human male dataset prepared by National Library of Medicine (NLM, USA) of size 1.2 GB.

Video Summary Based on F-Sift, Tamura Textural and Middle Level Semantic Feature

Procedia Computer Science, 2016

The continuous generation of digital multimedia has raised the question about management of time ... more The continuous generation of digital multimedia has raised the question about management of time and data. For effective management of such a tremendous amount of data or longer videos, a number of researches have been made. Video summarization is one of important research area to solve the problem. To summarise a video of any domain, researches relied on visual features available in frames, which does not guarantee to achieve semantic meaning of original video. In this paper, we proposed an approach for getting semantic video summary of original video. Generated video summaries are based on middle level semantic features, flip-Sift and tamura's textural feature. We used clustering approach based on middle level semantic features for static key frame extraction. Tamura feature has been used for getting texture features of frames. We tested this technique on various domains like sports, news, etc.; this technique has been tested on dataset freely available at YAACVID, UT egocentric and benchmark datasets. Longer duration video is also producing good result in compare to existing technique.

Social implications of crowd sourcing in rural Scotland

Various surveys mentioned that the benefits of ubiquitous crowdsourcing are reaped by people loca... more Various surveys mentioned that the benefits of ubiquitous crowdsourcing are reaped by people located in metro and smaller cities. The reach of crowdsourcing to rural population is questionable. The aim of this research is to bridge widening urban and rural divide by providing knowledge-intensive crowdsourcing tasks to rural work force which could provide long term benefits to them as well as improve supporting infrastructure. This paper reports an initial study of the demographic of small samples of twenty two rural homeworkers in Scotland, their motivation to do crowdsourcing work, present main occupation, computer skills, views on rural infrastructure and finally their skills on solving three spatial visualization tests. The survey shows that flexible hours of working, extra income, and work life balance are the three important factors emphasized as motivational constructs to do crowdsourcing work. Their skills on solving a spatial visualization test is equivalent to the literatur...

Out-Of-Core Construction and 3D Visualization of Level-Of-Detail Terrains Populated with Large Collection of Heterogeneous Objects

Proceedings of the First International Conference on Computer Graphics Theory and Applications, 2006

Interactive visualization of very large-scale terrain data in scientific visualization, GIS or si... more Interactive visualization of very large-scale terrain data in scientific visualization, GIS or simulation and training applications is a hard problem. The grid digital terrain elevation and texture data are not only too large to be rendered in real-time but also exceed physical main memory capacity. Therefore out-of-core management of digital terrain data is an essential requirement. Further to bring photorealism in visualization, it is required to place multiple collections of man-made objects such as buildings, lampposts etc. as well as natural objects such as trees, grass etc. on top of the terrain surface. In this paper we have proposed an integrated approach for effective out-of-core visualization of terrains populated with large collection of static heterogeneous objects. We have developed an efficient tile-based out-of-core viewdependent Level of Detail (LOD) mesh simplification algorithm for real-time rendering of large terrains. Instead of manipulating individual triangles, the algorithm operates on clusters of geometry called blocks of aggregate triangles. Hence the amount of work CPU must perform is greatly reduced. The formation of long triangle strips for LOD blocks also solves the CPU-to-Card bandwidth problem. The tile-based multiresolution terrain geometry framework has been extended to support large satellite or aerial imagery textures. To display large collection of objects over the terrain while maintaining the real-time frame rate, an efficient object handling method has been proposed using paging technique and object instantiation. User is allowed to control the objects locations, scales and orientations. The algorithms have been implemented using Visual C++ and OpenGL 3D API and successfully tested on different real-world height maps and satellite phototextures of sizes upto 16K*16K coupled with thousands of static objects on PCs.

Move Table: An Intelligent Software Tool for OptimalPath Finding and Halt Schedule Generation

Defence Science Journal, Sep 21, 2007

This study aims to help army officials in taking decisions before war to decide the optimal path ... more This study aims to help army officials in taking decisions before war to decide the optimal path for army troops moving between two points in a real world digital terrain, considering factors like traveled distance, terrain type, terrain slope, and road network. There can optionally be one or more enemies (obstacles) located on the terrain which should be avoided. A tile-based A* search strategy with diagonal distance and tie-breaker heuristics is proposed for finding the optimal path between source and destination nodes across a real-world 3-D terrain. A performance comparison (time analysis, search space analysis, and accuracy) has been made between the multiresolution A* search and the proposed tile-based A* search for large-scale digital terrain maps. Different heuristics, which are used by the algorithms to guide these to the goal node, are presented and compared to overcome some of the computational constraints associated with path finding on large digital terrains. Finally, a halt schedule is generated using the optimal path, weather condition, moving time, priority and type of a column, so that the senior military planners can strategically decide in advance the time and locations where the troops have to halt or overtake other troops depending on their priority and also the time of reaching the destination.

Devanagari Character Recognition towards natural Human-Computer Interaction

Electronic Workshops in Computing, 2010

Human-computer interaction is a growing research area. There are several ways of interaction with... more Human-computer interaction is a growing research area. There are several ways of interaction with the computer. Handwriting has continued to persist as a means of communication and recording information in the day to day life even with the introduction of new technologies. Due to the growth of technology in India, it becomes important to devise ways that allow people to communicate with computer in Indian languages. Hindi being the national language of India, we present a way to communicate with the computer in Hindi or more precisely, 'Devanagari script'. Due to absence of a global font to represent Devanagari characters, it is important that the computer recognizes the characters written by the user in order to interact with him. The algorithm implemented for character recognition first segments the image containing Devanagari text fed to the software into lines, lines to words and words to characters. The obtained characters are then brought down to a standard size. The Kohonen Neural Network based recognizer then comes into action and recognizes the text character by character and provides the output in Unicode format. The network has been designed with no hidden layer to support quick recognition. Apart from text recognition from an image, we also provided the option to recognize individual handwritten characters drawn using a mouse. Such a system provides keyboard less computer interaction. The technique is implemented using Java. The overall recognition rate for a fixed font machine printed characters is 90.26% and for hand written characters, it is 83.33%.

Evaluation schema for SAR image Segmentation based on Swarm Optimization in Neutrosophic domain

2014 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT), 2014

The present paper proposes an evaluation schema for SAR image Segmentation. The segmentation of S... more The present paper proposes an evaluation schema for SAR image Segmentation. The segmentation of SAR images becomes a crucial step because the segmented results will be used as the input for the post-processing like target detection, change detection applications. The preprocessing task segmentation is highly influenced by inherent noise known as speckle. The proposed system introduces the evaluation schema for segmentation of SAR images in Neutrosophic domain, which overcome the problem of speckle. The system is evaluated for different Swarm Optimization algorithms and the promoting results are obtained through the proposed evaluation schema. The results are discussed on the basis of different parameters.

Fusion of Multi-Sensor Satellite Images Using Non-Subsampled Contourlet Transform

Procedia Computer Science, 2015

The presented research work proposes fusion of multi-sensor satellite images using non subsampled... more The presented research work proposes fusion of multi-sensor satellite images using non subsampled contourlet transform. In the proposed work, trade-off between the spectral distortion and enhancement of spatial information is witnessed while fusing two multi-sensor images. The ills of wavelet based fusion techniques such as limited directionality, lack of phase information and shift invariant are addressed with the help of Non subsampled contourlet transform. The Non subsampled contourlet helps to retain the intrinsic structural information while decomposing and reconstructing the image components. Decision based rules are applied for component substitution for fusion. The experiments are carried out against the current state of art and observed that the proposed system provides promising results visually and quantitatively. The efficiency of the proposed system in the fused product is analysed qualitatively by Isodata classification algorithm.

Crowdsourcing solutions to 2D irregular strip packing problems from Internet workers

International Journal of Production Research, 2015

Many industrial processes require the nesting of 2D profiles prior to the cutting, or stamping, o... more Many industrial processes require the nesting of 2D profiles prior to the cutting, or stamping, of components from raw sheet material. Despite decades of sustained academic effort, algorithmic solutions are still sub-optimal and produce results that can frequently be improved by manual inspection. However, the Internet offers the prospect of novel 'human-in-the-loop' approaches to nesting problems that uses online workers to produce packing efficiencies beyond the reach of current CAM packages. To investigate the feasibility of such an approach, this paper reports on the speed and efficiency of online workers engaged in the interactive nesting of six standard benchmark data-sets. To ensure the results accurately characterise the diverse educational and social backgrounds of the many different labour forces available online, the study has been conducted with subjects based in both Indian IT service (i.e. Rural BPOs) centres and a network of homeworkers in Northern Scotland. The results (i.e. time and packing efficiency) of the human workers are contrasted with both the baseline performance of a commercial CAM package and recent research results. The paper concludes that online workers could consistently achieve packing efficiencies roughly 4% higher than the commercial based-line established by the project. Beyond characterising the abilities of online workers to nest components, the results also make a contribution to the development of algorithmic solutions by reporting new solutions to the benchmark problems and demonstrating methods for assessing the packing strategy employed by the best workers.

CORSISCA : classification of remotely sensed images - a soft computing approach

Proceedings of the IEEE INDICON 2004. First India Annual Conference, 2004.

The classification of multispectral satellite images is a challenging problem and has a number of... more The classification of multispectral satellite images is a challenging problem and has a number of applications such as feature identification, change detection, etc. We apply modified Neural Network Algorithms: GA-BP (Genetic Algorithm as precursor to the Back Propagation) and Modular Artificial Neural Network (MNN) to classify the LISS-3 image of Allahabad area. We also classify the resolution merged image (LISS-3 with PAN) using the same algorithms. By using Genetic algorithm as a precursor to ANN, we increase the probability of reaching to the global minimum, thus reducing the problem of a stuck neural network in the local minimum. MNN models the human brain more closely to apply task decomposition to the satellite images as well. The output of the above techniques are generated and analyzed.

An Assessment of the Impact of Dimensionality Reduction on the Speed and Accuracy of Hyperspectral Image Classification

This paper investigates the extent to which the accuracy and speed of classifying Hyperspectral r... more This paper investigates the extent to which the accuracy and speed of classifying Hyperspectral remote sensing images are affected by the application of varying degrees of dimensionality reduction. Three methods have been used for dimensionality reduction: PCA, ICA and random band selection; SVM has been used for classification. The results have been evaluated on both natural and man-made scenarios.

Towards crowdsourcing spatial manufacturing tasks from rural India

Many tasks involving production software that involve spatial reasoning can only produce "good" r... more Many tasks involving production software that involve spatial reasoning can only produce "good" rather than optimum solutions. It is frequently possible for humans to improve on algorithmically generated solutions computed by CAD/CAM software. The thesis motivating this research is that large numbers of industrial optimisation tasks involving spatial reasoning (such as 2D part nesting) can be outsourced as human intelligence tasks to rural workers to provide a sustainable source of skilled employment. We hypothesis that 3D spatial reasoning ability is essential to solve spatial optimization tasks. In this paper we reported the 3D spatial manipulation ability of eighty rural workers in four rural business process outsourcing (BPO) centres in India. The assessments reveal that although the average spatial manipulation ability is less than the literature reported benchmark, there are talented workers identified in all the four rural centres, and the results identify priority activities required to enable the proposed approach.

Real Time Multiple Hand Gesture Recognition System for Human Computer Interaction

International Journal of Intelligent Systems and Applications, 2012

With the increasing use of computing devices in day to day life, the need of user friendly interf... more With the increasing use of computing devices in day to day life, the need of user friendly interfaces has lead towards the evolution of different types of interfaces for human computer interaction. Real time vision based hand gesture recognition affords users the ability to interact with computers in more natural and intuitive ways. Direct use of hands as an input device is an attractive method which can communicate much more information by itself in comparison to mice, joysticks etc allowing a greater number of recognition system that can be used in a variety of human computer interaction applications. The gesture recognition system consist of three main modules like hand segmentation, hand tracking and gesture recognition from hand features. The designed system further integrated with different applications like image browser, virtual game etc. possibilities for human computer interaction. Computer Vision based systems has the potential to provide more natural, non-contact solutions. The present research work focuses on to design and develops a practical framework for real time hand gesture Index Terms-Real time, gesture recognition, human computer interaction, tracking.

Nonparametric Neural Network Model Based on Rough-Fuzzy Membership Function for Classification of Remotely Sensed Images

Lecture Notes in Computer Science, 2006

A nonparametric neural network model based on Rough-Fuzzy Membership function, multilayer percept... more A nonparametric neural network model based on Rough-Fuzzy Membership function, multilayer perceptron, and back-propagation algorithm is described. The described model is capable to deal with rough uncertainty as well as fuzzy uncertainty associated with classification of remotely sensed multi-spectral images. The input vector consists of membership values to linguistic properties while the output vector is defined in terms of rough fuzzy class membership values. This allows efficient modeling of indiscernibility and fuzziness between patterns by appropriate weights being assigned to the backpropagated errors depending upon the Rough-Fuzzy Membership values at the corresponding outputs. The effectiveness of the model is demonstrated on classification problem of IRS-P6 LISS IV images of Allahabad area. The results are compared with statistical (Minimum Distance), conventional MLP, and FMLP models.

Real Time Gesture Recognition System for Interaction in Dynamic Environment

Procedia Technology, 2012

Human Computer Interaction techniques have become a bottleneck in the effective utilization of th... more Human Computer Interaction techniques have become a bottleneck in the effective utilization of the available information flow. The development of user interfaces influences the changes in the Human-Computer Interaction (HCI). Human hand gestures have been a mode of non verbal interaction widely used. Naturalistic and intuitiveness of the hand gesture has been a great motivating factor for the researchers in the area of HCI to put their efforts to research and develop the more promising means of interaction between human and computers. This paper designs a system for gestural interaction between a user and a computer in dynamic environment. The gesture recognition system uses image processing techniques for detection, segmentation, tracking and recognition of hand gestures for converting it to a meaningful command. The interface being proposed here can be substantially applied towards different applications like image browser, games etc.

A Vision based Hand Gesture Interface for Controlling VLC Media Player

International Journal of Computer Applications, 2010

Human Computer Interaction can acquire several advantages with the introduction of different natu... more Human Computer Interaction can acquire several advantages with the introduction of different natural forms of device free communication. Gestures are a natural form of actions which we often use in our daily life for interaction, therefore to use it as a communication medium with computers generates a new paradigm of interaction with computers. This paper implements computer vision and gesture recognition techniques and develops a vision based low cost input device for controlling the VLC player through gestures. VLC application consists of a central computational module which uses the Principal Component Analysis for gesture images and finds the feature vectors of the gesture and save it into a XML file. The Recognition of the gesture is done by K Nearest Neighbour algorithm. The theoretical analysis of the approach shows how to do recognition in static background. The Training Images are made by cropping the hand gesture from static background by detecting the hand motion using Lucas Kanade Pyramidical Optical Flow algorithm. This hand gesture recognition technique will not only replace the use of mouse to control the VLC player but also provide different gesture vocabulary which will be useful in controlling the application.