(Back to Session Schedule)

28th Picture Coding Symposium

Session S1  Special Session 1: 3DTV/FTV
Time: 13:00 - 15:30 Wednesday, December 8, 2010
Chair: Toshiaki Fujii (Tokyo Institute of Technology, Japan)

S1-1 (Time: 13:00 - 13:30)
Title(Invited Talk) Challenges in Multiview Video - The 3 D's
AuthorAntonio Ortega (University of Southern California, U.S.A.)
Pagep. 17
Keywordmultiview video, multiview video coding, depth estimation
AbstractWe highlight some of the challenges facing wide adoption of multiview video technologies, by focusing on design issues involving displays, depth data and tools for content delivery. Our goal is to show that uncertainty about what choices will be made in each of these areas have a significant impact on the design of coding tools.

S1-2 (Time: 13:30 - 14:00)
Title(Invited Talk) Vision Field Capturing and Its Applications in 3DTV
AuthorQionghai Dai (Tsinghua University, China)
Pagep. 18
Keywordmultiview capturing, vision field capturing, 3DTV
Abstract3D video capturing acquires the visual information in 3D manner, which possesses the first step of the entire 3DTV system chain before 3D coding, transmission and visualization. As the cost of sensors reduces recently, many systems utilize multiple cameras to acquire visual information, which is called multiview capturing. We start with reviewing these multiview systems built in the past decades. Then, a new concept of vision field capturing is introduced. Finally, some new techniques using TOF(time-of-flight) camera and 3D scanner will also be presented.

S1-3 (Time: 14:00 - 14:30)
Title(Invited Talk) 3D Information Coding
AuthorJoern Ostermann (Hannover University, Germany)
Pagep. 19
KeywordMPEG, 3D reconstruction, 3DTV, stereo video, multiview display
Abstract3D visual information can be presented to viewers using stereo displays. For these displays, the two video sequences of a stereo pair are jointly coded using MVC. The extension of coding standards for use with the existing MPEG-2 or AVC infrastructure is currently developed. Auto stereoscopic displays enabling motion parallax will be supported by the future 3DV standard. Several video streams and a depth map will enable the display to render the correct views of the scene.

S1-4 (Time: 14:30 - 15:00)
Title(Invited Talk) 3D Television System Based on Integral Photography
AuthorTomoyuki Mishina (NHK, Japan)
Pagep. 20
Keyword3D image, integral photography, full-parallax, Super Hi-Vision
AbstractIntegral photography is a photographic technique which is able to reconstruct three-dimensional (3D) images having full-parallax, and the reconstructed 3D images can be seen without wearing special glasses. We have been researching into a 3D television system based on integral photography. This paper describes our system in which an ultra-high-definition imaging system called Super Hi-Vision is applied to integral photography.

S1-5 (Time: 15:00 - 15:30)
Title(Invited Talk) Progress from Stereoscopic to Three-Dimensional Displays Based on Visual Perception
AuthorSumio Yano (NHK, formerly NICT, Japan)
Pagep. 21
Keywordstereoscopic displays, multi-view stereoscopic displays, three-dimensional display, light field reproduction, human visual field
AbstractThe proper characteristics of stereoscopic and multi-view displays are described in the context of human visual perception at first. Next, the development of three-dimensional displays, matched to the function of human visual field, is described accounting for these points. These three-dimensional displays were developed based on the principle of light field reproduction. From these issues, the processing technology for large capacity image information is important for the development of the three-dimensional image systems in the future.