Зарегистрироваться
Восстановить пароль
FAQ по входу

Zhu C., Zhao Y., Yu L., Tanimoto M. (eds.) 3D-TV System with Depth-Image-Based Rendering. Architectures, Techniques and Challenges

  • Файл формата pdf
  • размером 10,56 МБ
  • Добавлен пользователем
  • Отредактирован
Zhu C., Zhao Y., Yu L., Tanimoto M. (eds.) 3D-TV System with Depth-Image-Based Rendering. Architectures, Techniques and Challenges
Издательство Springer, 2013, -478 pp.
Riding on the success of 3D cinema blockbusters and the advancements in stereoscopic display technology, 3D video applications have been gathering momentum in recent years, which further enhance visual experience by vividly extending the conventional ‘‘flat’’ video into a third dimension. Several 3D video prototypes have been developed based on distinct techniques in 3D visualization, representation, and content production. Among them, stereoscopic 3D video systems evoke 3D perception by binocular parallax, in which scene is presented in a fixed perspective defined by two transmitted views, while further manipulations on depth perception require expensive computation with current technologies.
Depth-image-based rendering (DIBR) is being considered to significantly enhance the 3D visual experience relative to the conventional stereoscopic systems. With DIBR techniques, it becomes possible to generate additional viewpoints using 3D warping techniques to adjust the perceived depth of stereoscopic videos or to provide the necessary input for auto-stereoscopic displays that do not require glasses to view the 3D scene. This functionality is also useful for free-viewpoint video (FVV), where the viewer has the freedom to move about in front of the display, and is able to perceive natural perspective changes as if looking through a window. In recognition of progress being made in this area and a strong interest from the industry to provide equipment and services supporting such applications, MPEG is also embarking on a new phase of 3D video standardization based on DIBR techniques.
The technologies surrounding DIBR-oriented 3D video systems, however, are not mature enough at this stage to fully fulfill the above targets. Depth maps, which are central to the synthesis of virtual views, need to be either captured with specialized apparatus or estimated from scene textures using stereo matching. Existing solutions are either costly or not sufficiently robust. Besides, there is a strong need to achieve efficient storage and robust transmission of this additional information. Knowing that the depth maps and scene textures are different in nature, and that synthesized views are the ultimate information for display, DIBRoriented depth, and texture coding may employ different distortion measures for rate-distortion or rate-quality optimization, and possibly different coding principles to make better use of available bandwidth. Since view synthesis, coupled with errors introduced by depth generation and compression, may introduce new types of artifacts that are different from those of conventional video acquisition and compression systems, it is also necessary to understand the visual quality of the views produced by DIBR techniques, which is critical to ensure a comfortable, realistic, and immersive 3D experience.
This book focuses on this depth-based 3D-TV system which is expected to be put into applications in the near future as a more attractive alternative to the current stereoscopic 3D-TV system. Following an open call for chapters and a few rounds of extensive peer reviews, 15 chapters of good quality have been finally accepted, ranging from a technical review and literature survey on the whole system or a particular topic, solutions to some technical issues, to implementation of some prototypes. According to the scope of these chapters, this book is organized into four sections, namely system overview, content generation, data compression and transmission, and 3D visualization and quality assessment, with the chapters in each section summarized below.
Part I System Overview
An Overview of 3D-TV System Using Depth-Image-Based Rendering
Part II Content Generation
Generic Content Creation for 3D Displays
Stereo Matching and Viewpoint Synthesis FPGA Implementation
DIBR-Based Conversion from Monoscopic to Stereoscopic and Multi-View Video
Virtual View Synthesis and Artifact Reduction Techniques
Hole Filling for View Synthesis
LDV Generation from Multi-View Hybrid Image and Depth Video
Part III Data Compression and Transmission
3D Video Compression
Depth Map Compression for Depth-Image-Based Rendering
Effects of Wavelet-Based Depth Video Compression
Transmission of 3D Video over Broadcasting
Part IV 3D Visualization and Quality Assessment
The Psychophysics of Binocular Vision
Stereoscopic and Autostereoscopic Displays
Subjective and Objective Visual Quality Assessment in the Context of Stereoscopic 3D-TV
Visual Quality Assessment of Synthesized Views in the Context of 3D-TV
  • Чтобы скачать этот файл зарегистрируйтесь и/или войдите на сайт используя форму сверху.
  • Регистрация