Human Recognition at a Distance in Video (Advances in Computer Vision and Pattern Recognition)

✍ Scribed by Bir Bhanu, Ju Han

• Publisher: Springer
• Year: 2010
• Tongue: English
• Leaves: 268
• Category: Library

✦ Synopsis

Most biometric systems employed for human recognition require physical contact with, or close proximity to, a cooperative subject. Far more challenging is the ability to reliably recognize individuals at a distance, when viewed from an arbitrary angle under real-world environmental conditions. Gait and face data are the two biometrics that can be most easily captured from a distance using a video camera.

This comprehensive and logically organized text/reference addresses the fundamental problems associated with gait and face-based human recognition, from color and infrared video data that are acquired from a distance. It examines both model-free and model-based approaches to gait-based human recognition, including newly developed techniques where the both the model and the data (obtained from multiple cameras) are in 3D. In addition, the work considers new video-based techniques for face profile recognition, and for the super-resolution of facial imagery obtained at different angles. Finally, the book investigates integrated systems that detect and fuse both gait and face biometrics from video data.

Topics and features: discusses a framework for human gait analysis based on Gait Energy Image, a spatio-temporal gait representation; evaluates the discriminating power of model-based gait features using Bayesian statistical analysis; examines methods for human recognition using 3D gait biometrics, and for moving-human detection using both color and thermal image sequences; describes approaches for the integration face profile and gait biometrics, and for super-resolution of frontal and side-view face images; introduces an objective non-reference quality evaluation algorithm for super-resolved images; presents performance comparisons between different biometrics and different fusion methods for integrating gait and super-resolved face from video.

This unique and authoritative text is an invaluable resource for researchers and graduate students ofcomputer vision, pattern recognition and biometrics. The book will also be of great interest to professional engineers of biometric systems.

✦ Table of Contents

Preface
Contents
List of Figures
List of Tables
Introduction to Gait-Based Individual Recognition at a Distance
Introduction
Gait-Based Human Recognition
Face-Based Human Recognition
Key Ideas Described in the Book
Organization of the Book
Gait-Based Individual Recognition at a Distance
Gait Representations in Video
Human Motion Analysis and Representations
Human Activity and Individual Recognition by Gait
Human Recognition by Gait
Model-Based Approaches
Model-Free Approaches
Human Activity Recognition
Model-Based Approaches
Model-Free Approaches
Gait Energy Image (GEI) Representation
Motivation
Representation Construction
Relationship with MEI and MHI
Representation Justification
Framework for GEI-Based Recognition
Silhouette Extraction and Processing
Feature Extraction
Summary
Model-Free Gait-Based Human Recognition in Video
Statistical Feature Fusion for Human Recognition by Gait
Real and Synthetic Gait Templates
Human Recognition
Experimental Results
Data and Parameters
Performance Evaluation
Human Recognition Based on Environmental Context
Walking Surface Type Detection
Classifier Design
Probabilistic Classifier Combination
Experimental Results
View-Insensitive Human Recognition by Gait
View-Insensitive Gait Templates
Human Recognition
Experimental Results
Human Repetitive Activity Recognition in Thermal Imagery
Object Detection in Thermal Infrared Imagery
Human Repetitive Activity Representation and Recognition
Experimental Results
Human Recognition Under Different Carrying Conditions
Technical Approach
Gait Energy Image (GEI)
Feature Extraction
Co-evolutionary Genetic Programming
Majority Voting
Experimental Results
Data
Experiments
Classifier Performance Comparison
Summary
Discrimination Analysis for Model-Based Gait Recognition
Predicting Human Recognition Performance
Algorithm Dependent Prediction and Performance Bounds
Body Part Length Distribution
Algorithm Dependent Performance Prediction
Upper Bound on PCR
Experimental Results
Summary
Model-Based Human Recognition-2D and 3D Gait
2D Gait Recognition (3D Model, 2D Data)
3D Human Modeling
Human Kinematic Model
Human Model Parameter Selection
Camera Model and Coordinate Transformation
World Coordinate to Camera Coordinate
Camera Coordinate to Ideal Image Coordinate
Ideal Image Coordinate to Actual Image Coordinate
Actual Image Coordinate to Computer Image Coordinate
Human Recognition from Single Non-calibrated Camera
Silhouette Preprocessing
Matching Between 3D Model and 2D Silhouette
Human Model Parameter Estimation
Stationary Parameter Estimation
Kinematic Parameter Estimation
Recognition Based on Kinematic and Stationary Features
Kinematic and Stationary Feature Classifier
Classifier Combination Strategies
Performance Evaluation on Monocular Image Sequences
Performance of Stationary Feature Classifier
Performance of Kinematic Feature Classifier
Performance with Classifier Combination
Human Recognition from Multiple Calibrated Cameras
Human Model Parameter Selection
Matching Between 3D Human Model and Multiple 2D Silhouettes
Human Model Parameter Initialization and Estimation
Performance Evaluation on Data from Multiple Cameras
Gait Recognition in 3D
Individual Recognition by Gait in 3D
Related Work
Technical Approach
3D Human Body Data
3D Human Body Model
Model Fitting
Body Axes
Torso
Arms and Legs
Head and Neck
Gait Reconstruction
Feature Matching
Experimental Results
Gait Reconstruction
Training and Testing Data
Gait Recognition
Summary
Fusion of Color/Infrared Video for Human Detection
Related Work
Hierarchical Image Registration and Fusion Approach
Image Transformation Model
Preliminary Human Silhouette Extraction and Correspondence Initialization
Automatic Image Registration
Model Parameter Selection
Parameter Estimation Based on Hierarchical Genetic Algorithm
Sensor Fusion
Registration of EO/IR Sequences with Multiple Objects
Experimental Results
Image Registration Results
Sensor Fusion Results
Summary
Face Recognition at a Distance in Video
Super-Resolution of Facial Images in Video at a Distance
Closed-Loop Super-Resolution of Face Images in Video
Related Work
Technical Approach
Bilinear Basis Images Computation
Pose and Illumination Estimation
Super-Resolution Algorithm
Experimental Results
Synthetic Data
Real Video
Super-Resolution of Facial Images with Expression Changes in Video
Related Work
Technical Approach
Tracking of Facial Regions
Local Deformation
Free Form Deformation Formulation
Cost Function
Resolution Aware Local Deformation
Deform Local Motion on High Resolution Data
Super-Resolution Methodology Requires Sub-pixel Registration
Super-Resolution Algorithm
A Match Measure for Warping Errors
Experimental Results
Data and Parameters
Results of Resolution Aware FFD
Super-Resolution Results-Global Registration vs. Global + RAIFFD Local Deformation
Quantification of Performance
Proposed Approach with Two Different SR Algorithms
Constructing Enhanced Side Face Images from Video
Enhanced Side Face Image (ESFI) Construction
Technical Approach
Acquiring Moving Head of a Person in Video
Side Face Image Alignment
Elastic Registration Method
Match Statistic
Resolution Enhancement Algorithm
The Imaging Model
Algorithm for Resolution Enhancement
Side Face Normalization
Summary
Evaluating Quality of Super-Resolved Face Images
Image Quality Indices
Integrated Image Quality Index
Gray Scale Based Quality (Qg)
Structure Based Quality (Qe)
Similarity Between Input Images (Qi)
Integrated Quality Measure (Qint)
Experimental Results for Face Recognition in Video
Experiment 1: Influence of Pose Variation on the Super-Resolved Face Image
Experiment 2: Influence of Lighting Variation on the Super-Resolved Face Image
Experiment 3: Influence of Facial Expression Variation on the Super-Resolved Face Image
Experiment 4: Influence of the Number of Images Used for Constructing the Super-Resolved Face Image for Face Recognition
Discussion
Summary
Integrated Face and Gait for Human Recognition at a Distance in Video
Integrating Face Profile and Gait at a Distance
Introduction
Technical Approach
High-Resolution Image Construction for Face Profile
The Imaging Model
The Super Resolution Algorithm
Face Profile Representation and Matching
Face Profile Extraction
Curvature-Based Fiducial Extraction
Profile Matching Using Dynamic Time Warping
Gait Recognition
Integrating Face Profile and Gait for Recognition at a Distance
Experimental Results
Face Profile-Based Recognition
Static Face Database
Experimental Results
Integrating Face Profile With Gait
Video Data
Experimental Results
Summary
Match Score Level Fusion of Face and Gait at a Distance
Introduction
Related Work
Technical Approach
Enhanced Side Face Image Construction
Gait Energy Image Construction
Human Recognition Using ESFI and GEI
Feature Learning Using PCA and MDA Combined Method
Recognition by Integrating ESFI and GEI
Experimental Results and Performance Analysis
Experiments and Parameters
Experiment 1
Experiment 2
Experiment 3
Performance Analysis
Discussion on Experiments
Performance Characterization Statistic Q
Summary
Feature Level Fusion of Face and Gait at a Distance
Introduction
Technical Approach
Human Identification Using ESFI and GEI
Feature Learning Using PCA
Synthetic Feature Generation and Classification
The Related Fusion Schemes
Fusion at the Match Score Level
Fusion at the Feature Level
Experimental Results and Comparisons
Experiments and Parameters
Experiment 1
Experiment 2
Discussion on Experiments
Summary
Conclusions for Integrated Gait and Face for Human Recognition at a Distance in Video
Conclusions and Future Work
Summary
Gait-Based Human Recognition at a Distance
Video-Based Human Recognition at a Distance
Fusion of Face and Gait for Human Recognition at Distance
Future Research Directions
References
Index