Multi-Frame Motion-Compensated Prediction for Video Transmission presents a comprehensive description of a new technique in video coding and transmission. The work presented in the book has had a very strong impact on video coding standards and will be of interest to practicing engineers and researchers as well as academics. The multi-frame technique and the Lagrangian coder control have been adopted by the ITU-T as an integral part of the well known H.263 standard and are were adopted in the ongoing H.26L project of the ITU-T Video Coding Experts Group. This work will interest researchers and students in the field of video coding and transmission. Moreover, engineers in the field will also be interested since an integral part of the well known H.263 standard is based on the presented material.

With a preface by Ton Kalker. Informed Watermarking is an essential tool for both academic and professional researchers working in the areas of multimedia security, information embedding, and communication. Theory and practice are linked, particularly in the area of multi-user communication. From the Preface: Watermarking has become a more mature discipline with proper foundation in both signal processing and information theory. We can truly say that we are in the era of "second generation" watermarking. This book is first in addressing watermarking problems in terms of second-generation insights. It provides a complete overview of the most important results on capacity and security. The Costa scheme, and in particular a simpler version of it, the Scalar Costa scheme, is studied in great detail. An important result of this book is that it is possible to approach the Shannon limit within a few decibels in a practical system. These results are verified on real-world data, not only the classical category of images, but also on chemical structure sets. Inspired by the work of Moulin and O'Sullivan, this book also addresses security aspects by studying AGWN attacks in terms of game theory. "The authors of Informed Watermarking give a well-written exposA(c) of how watermarking came of age, where we are now, and what to expect in the future. It is my expectation that this book will be a standard reference on second-generation watermarking for the years to come." Ton Kalker, Technische Universiteit Eindhoven

Appendices 133 A Mathematical Results 133 A.1 Singularities of the Displacement Error Covariance Matrix 133 A.2 A Class of Matrices and their Eigenvalues 134 A.3 Inverse of the Power Spectral Density Matrix 134 A.4 Power Spectral Density of a Frame 136 Glossary 137 References 141 Index 159 Preface This book aims to capture recent advances in motion compensation for - ficient video compression. It investigates linearly combined motion comp- sated signals and generalizes the well known superposition for bidirectional prediction in B-pictures. The number of superimposed signals and the sel- tion of reference pictures will be important aspects of the discussion. The application oriented part of the book employs this concept to the well known ITU-T Recommendation H.263 and continues with the improvements by superimposed motion-compensated signals for the emerging ITU-T R- ommendation H.264 and ISO/IEC MPEG-4 (Part 10). In addition, it discusses a new approach for wavelet-based video coding. This technology is currently investigated by MPEG to develop a new video compression standard for the mid-term future.

Traditionally, three-dimensional image analysis (a.k.a. computer vision) and three-dimensional image synthesis (a.k.a. computer graphics) were separate fields. Rarely were experts working in one area interested in and aware of the advances in the other. Over the last decade this has changed dramatically, reflecting the growing maturity of each of these areas. The vision and graphics communities are today engaged in a mutually beneficial exchange, learning from each other and coming up with new ideas and techniques that build on the state of the art in both fields. This book is the result of a fruitful collaboration between scientists at the University of NA1/4rnberg, Germany, who, coming from diverse fields, are working together propelled by the vision of a unified area of three-dimensional image analysis and synthesis. Principles of 3D Image Analysis and Synthesis starts out at the image acquisition end of a hypothetical processing chain, proceeds with analysis, recognition and interpretation of images, towards the representation of scenes by 3D geometry, then back to images via rendering and visualization techniques. Coverage includes discussion of range cameras, multiview image processing, the structure-from-motion problem, object recognition, knowledge-based image analysis, active vision, geometric modeling with meshes and splines, and reverse engineering. Also included is cutting-edge coverage of texturing techniques, global illumination, image-based rendering, volume visualization, flow visualization techniques, and acoustical imaging including object localization from audio and video. This state-of-the-art volume is a concise and readable reference for scientists, engineers, graduate students and educators working in image processing, vision, computer graphics, or visualization.

Peer-to-Peer Video Streaming describes novel solutions to enhance video quality, increase robustness to errors, and reduce end-to-end latency in video streaming systems. This book will be of use to both academics and professionals as it presents thorough coverage and solutions for current issues with Video Streaming and Peer-to-Peer architectures. The book provides an overview of today 's state-of-the art video streaming technology. It presents adaptive video coding and streaming techniques for performance enhancement of conventional client-server systems and P2P multicast. The detailed appendix incorporates various additional experiments.

With a preface by Ton Kalker. Informed Watermarking is an essential tool for both academic and professional researchers working in the areas of multimedia security, information embedding, and communication. Theory and practice are linked, particularly in the area of multi-user communication. From the Preface: Watermarking has become a more mature discipline with proper foundation in both signal processing and information theory. We can truly say that we are in the era of second generation watermarking. This book is first in addressing watermarking problems in terms of second-generation insights. It provides a complete overview of the most important results on capacity and security. The Costa scheme, and in particular a simpler version of it, the Scalar Costa scheme, is studied in great detail. An important result of this book is that it is possible to approach the Shannon limit within a few decibels in a practical system. These results are verified on real-world data, not only the classical category of images, but also on chemical structure sets.Inspired by the work of Moulin and O'Sullivan, this book also addresses security aspects by studying AGWN attacks in terms of game theory. The authors of Informed Watermarking give a well-written expose of how watermarking came of age, where we are now, and what to expect in the future. It is my expectation that this book will be a standard reference on second-generation watermarking for the years to come. Ton Kalker, Technische Universiteit Eindhoven

Multi-Frame Motion-Compensated Prediction for Video Transmission presents a comprehensive description of a new technique in video coding and transmission. The work presented in the book has had a very strong impact on video coding standards and will be of interest to practicing engineers and researchers as well as academics. The multi-frame technique and the Lagrangian coder control have been adopted by the ITU-T as an integral part of the well known H.263 standard and are were adopted in the ongoing H.26L project of the ITU-T Video Coding Experts Group. This work will interest researchers and students in the field of video coding and transmission. Moreover, engineers in the field will also be interested since an integral part of the well known H.263 standard is based on the presented material.

Appendices 133 A Mathematical Results 133 A.1 Singularities of the Displacement Error Covariance Matrix 133 A.2 A Class of Matrices and their Eigenvalues 134 A.3 Inverse of the Power Spectral Density Matrix 134 A.4 Power Spectral Density of a Frame 136 Glossary 137 References 141 Index 159 Preface This book aims to capture recent advances in motion compensation for - ficient video compression. It investigates linearly combined motion comp- sated signals and generalizes the well known superposition for bidirectional prediction in B-pictures. The number of superimposed signals and the sel- tion of reference pictures will be important aspects of the discussion. The application oriented part of the book employs this concept to the well known ITU-T Recommendation H.263 and continues with the improvements by superimposed motion-compensated signals for the emerging ITU-T R- ommendation H.264 and ISO/IEC MPEG-4 (Part 10). In addition, it discusses a new approach for wavelet-based video coding. This technology is currently investigated by MPEG to develop a new video compression standard for the mid-term future.

Traditionally, say 15 years ago, three-dimensional image analysis (aka computer vi sion) and three-dimensional image synthesis (aka computer graphics) were separate fields. Rarely were expert"

Peer-to-Peer Video Streaming describes novel solutions to enhance video quality, increase robustness to errors, and reduce end-to-end latency in video streaming systems. This book will be of use to both academics and professionals as it presents thorough coverage and solutions for current issues with Video Streaming and Peer-to-Peer architectures. The book provides an overview of today's state-of-the art video streaming technology. It presents adaptive video coding and streaming techniques for performance enhancement of conventional client-server systems and P2P multicast. The detailed appendix incorporates various additional experiments.