Analyzing Video Sequences of Multiple Humans - Tracking, Posture Estimation and Behavior Recognition (THE KLUWER INTERNATIONAL SERIES IN VIDEO COMPUTING ... International Series in Video Computing)

Traditionally, scientific fields have defined boundaries, and scientists work on research problems within those boundaries. However, from time to time those boundaries get shifted or blurred to evolve new fields. For instance, the original goal of computer vision was to understand a single image of a scene, by identifying objects, their structure, and spatial arrangements. This has been referred to as image understanding. Recently, computer vision has gradually been making the transition away from understanding single images to analyz­ ing image sequences, or video understanding. Video understanding deals with understanding of video sequences, e. g. , recognition of gestures, activities, fa­ cial expressions, etc. The main shift in the classic paradigm has been from the recognition of static objects in the scene to motion-based recognition of actions and events. Video understanding has overlapping research problems with other fields, therefore blurring the fixed boundaries. Computer graphics, image processing, and video databases have obvious overlap with computer vision. The main goal of computer graphics is to gener­ ate and animate realistic looking images, and videos. Researchers in computer graphics are increasingly employing techniques from computer vision to gener­ ate the synthetic imagery. A good example of this is image-based rendering and modeling techniques, in which geometry, appearance, and lighting is de­ rived from real images using computer vision techniques. Here the shift is from synthesis to analysis followed by synthesis.

With the ubiquity of new information technology and media, more effective and friendly methods for human computer interaction (HCI) are being developed which do not rely on traditional devices such as keyboards, mice and displays. The first step for any intelligent HCI system is face detection, and one of most friendly HCI systems is hand gesture.
Face Detection and Gesture Recognition for Human-Computer Interaction introduces the frontiers of vision-based interfaces for intelligent human computer interaction with focus on two main issues: face detection and gesture recognition. The first part of the book reviews and discusses existing face detection methods, followed by a discussion on future research. Performance evaluation issues on the face detection methods are also addressed. The second part discusses an interesting hand gesture recognition method based on a generic motion segmentation algorithm. The system has been tested with gestures from American Sign Language with promising results. We conclude this book with comments on future work in face detection and hand gesture recognition.
Face Detection and Gesture Recognition for Human-Computer Interaction will interest those working in vision-based interfaces for intelligent human computer interaction. It also contains a comprehensive survey on existing face detection methods, which will serve as the entry point for new researchers embarking on such topics. Furthermore, this book also covers in-depth discussion on motion segmentation algorithms and applications, which will benefit more seasoned graduate students or researchers interested in motion pattern recognition.

Booknews

Researchers have been working under the premise that a computer can recognize information about a user's identity, state, and intent from facial expression and hand motions. Because of the immense number of variables in processing facial images (scale, location, orientations, pose, etc) the first step is getting computers to recognize faces. Gesture recognition is similarly complicated. Yang (Honda R&D Americas, Inc.) and Ahuja (computer science, U. of Illinois at Urbana-Champaign) present their own work in these related fields and summarize some recent findings by other researchers. They present an algorithm for extracting and recognizing motion patterns, present experimental results on motion patterns related to 40 American Sign Language gestures, discuss a model that extracts skin tone regions for the reduction of computation needed, present experimental results for the model, and describe a (Sparse Network of Winnows) SnoW-based face detector. Finally future directions of research are examined. Annotation c. Book News, Inc., Portland, OR (booknews.com)

This book is an excellent reference for both professional and academic researchers in the fields of computer vision and image processing. It will also be of interest to those working in video surveillance and monitoring, virtual reality, computer graphics, pattern recognition, telecommunication, human-computer interface, and general computer science.
Analyzing Video Sequences of Multiple Humans: Tracking, Posture Estimation and Behavior Recognition describes some computer vision-based methods that analyze video sequences of humans. More specifically, methods for tracking multiple humans in a scene, estimating postures of a human body in 3D in real-time, and recognizing a person's behavior (gestures or activities) are discussed. For the tracking algorithm, the authors developed a non-synchronous method that tracks multiple persons by exploiting a Kalman filter that is applied to multiple video sequences. For estimating postures, an algorithm is presented that locates the significant points which determine postures of a human body, in 3D in real-time. Human activities are recognized from a video sequence by the HMM (Hidden Markov Models)-based method that the authors pioneered. The effectiveness of the three methods is shown by experimental results.
The posture estimation method described in this book is a world-leading method in that it can estimate postures in 3D in real-time. As described in this book, the posture estimation method can be applied to avatar-based telecommunication systems, which require realistic, real-time reproduction of human images.
This book is also suitable for use in graduate classes in computer vision or image processing.

Booknews

This text describes some computer vision based methods that analyze methods for tracking multiple humans in a scene, estimating postures of a human body in 3D in real-time, and recognizing a person's gestures or activities. The five chapters detail the tracking algorithm developed by Ohya (Waseda U.), Utsumi (Advanced Telecommunications Research Institute) and Yamato (Nippon Telegraph and Telephone Corporation), which involves a non-synchronous method that exploits a Kalman filter applied to multiple video sequences. They also present the algorithm for estimating postures, and the method for recognizing human activities from a video sequence by the Hidden Markov Models. Appropriate for professional and academic researchers, as well as for use in graduate classes in computer vision or image processing. Annotation c. Book News, Inc., Portland, OR (booknews.com)

Traditionally, scientific fields have defined boundaries, and scientists work on research problems within those boundaries. However, from time to time those boundaries get shifted or blurred to evolve new fields. For instance, the original goal of computer vision was to understand a single image of a scene, by identifying objects, their structure, and spatial arrangements. This has been referred to as image understanding. Recently, computer vision has gradually been making the transition away from understanding single images to analyz­ ing image sequences, or video understanding. Video understanding deals with understanding of video sequences, e. g., recognition of gestures, activities, fa­ cial expressions, etc. The main shift in the classic paradigm has been from the recognition of static objects in the scene to motion-based recognition of actions and events. Video understanding has overlapping research problems with other fields, therefore blurring the fixed boundaries. Computer graphics, image processing, and video databases have obvious overlap with computer vision. The main goal of computer graphics is to gener­ ate and animate realistic looking images, and videos. Researchers in computer graphics are increasingly employing techniques from computer vision to gen­ erate the synthetic imagery. A good example of this is image-based rendering and modeling techniques, in which geometry, appearance, and lighting is de­ rived from real images using computer vision techniques. Here the shift is from synthesis to analysis followed by synthesis.

Written by international experts in the field, Video Registration is a survey of the state of the art in this fundamental technology. It is an essential reference to both academicians and practicing engineers working in the fields of security, surveillance, multimedia, and robotics.
The increased availability of low-cost, low-power, highly accurate video imagery has resulted in a rapid growth of the applications for this data. Video imagery has many advantages over still frame imagery; for example, it provides context and timing relationships, which are suitable for dynamic situation monitoring and action verification. Manipulation of video requires automatic processing and analysis (computer vision and image processing), vast amounts of storage and efficient search methods (databases), high bandwidth communication (networking), and real-time implementations (VLSI/hardware). Users of video imagery include disaster relief agencies, environmental monitoring and planning applications, tactical military groups, civilian agencies such as homeland security agencies, city planners, transportation (traffic management), the entertainment industry, law enforcement groups, landscape ecologists, WWW users and trainers and educators.
Video Registration is suitable as a text at the senior undergraduate and graduate levels.

Written by international experts in the field, Video Registration is a survey of the state of the art in this fundamental technology. It is an essential reference to both academicians and practicing engineers working in the fields of security, surveillance, multimedia, and robotics. The increased availability of low-cost, low-power, highly accurate video imagery has resulted in a rapid growth of the applications for this data. Video imagery has many advantages over still frame imagery; for example, it provides context and timing relationships, which are suitable for dynamic situation monitoring and action verification. Manipulation of video requires automatic processing and analysis (computer vision and image processing), vast amounts of storage and efficient search methods (databases), high bandwidth communication (networking), and real-time implementations (VLSI/hardware). Users of video imagery include disaster relief agencies, environmental monitoring and planning applications, tactical military groups, civilian agencies such as homeland security agencies, city planners, transportation (traffic management), the entertainment industry, law enforcement groups, landscape ecologists, WWW users and trainers and educators. Video Registration is suitable as a text at the senior undergraduate and graduate levels Front Matter....Pages i-xiv Video Registration: A Perspective....Pages 1-17 Automatic Camera Tracking....Pages 18-35 Motion Information in the Phase Domain....Pages 36-71 Generation and Error Characterization of Pararell-Perspective Stereo Mosaics from Real Video....Pages 72-105 Model-Based Landmark Extraction and Correspondence Finding for Aerial Image Registration....Pages 106-129 Airborne Video Registration for Activity Monitoring....Pages 130-143 Geodetic Alignment of Aerial Video Frames....Pages 144-179 Robust Video Georegistration in the Presence of Significant Appearance Changes....Pages 180-221 Panel Introduction Video Registration: Key Challenges and the Potential Impact of their Solutions to the Field of Computer Vision....Pages 222-255 Back Matter....Pages 256-257

The increasing availability of low-cost, low-power, highly accurate video imagery has brought rapid growth in the applications using this data. Video provides multiple temporal constraints, making it easier to analyze a complex and coordinated series of events that cannot be understood by looking at a single image or a few frames. Arguing for a bottom-up approach to object recognition and event detection, the authors present a framework for the extraction of meaningful information from video. The text's underlying principle is that many diverse pieces of evidence are more useful for object recognition and event detection than the most elaborate algorithm working on an impoverished image representation. Annotation c. Book News, Inc., Portland, OR (booknews.com)

This book explores the new area of computational media aesthetics to bridge the semantic gap. It brings together researchers from multiple disciplines, from computer scientist to content creators from engineers to media specialists, to highlight recent advances in this area. It examines the elements of media production and aesthetics principles in order to establish the foundation for semantics indexing, and to build innovative and effective technologies for content annotation, search, and browsing.

Media Computing: Computational Media Aesthetics outlines strategic means and methods to derive high level semantic constructs from automatic analysis of media. It is suitable for the use as a text in courses on multimedia and informational retrieval.

While issues of media archival as well as delivery on the Internet and corporate intranets are adequately addressed by improved compression standards, faster networks, and advances made in storage and streaming technologies, the challenges of automating media annotation, content indexing, segmentation, and organization for search, retrieval, and browsing applications are still being tackled. Images containing faces are essential to intelligent vision-based human computer interaction, and research efforts in face processing include face recognition, face tracking, pose estimation, and facial expression recognition.