Region-interior painting in Contour Based Depth map Coding System

Minghui Wang*, Xun He, Xin Jin, Satoshi Goto

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

Depth Image Based Rendering (DIBR) is a coding strategy for Multiview Video Coding (MVC). Depth video is involved to DIBR to represent another dimensional of the traditional 2D video (texture video). The physics meaning of depth video is different from texture video. But the traditional block based coding system use the same procedure to code both depth video and texture video. A new coding system that dedicate to depth video coding is proposed in this paper, which is called Contour Based Depth map Coding system (CBDC). To meet the feature of sharp-edge and smooth-interior, the proposed CBDC performs contour coding and interior coding separately. This paper focus on introduce the interior coding part. Linear painting modes and non-linear painting mode are introduced and compared. Experimental results indicate that proposed painting method performs an equivalent PSNR performance to the block based coding strategy (JMVC).

Original languageEnglish
Title of host publication2011 International Symposium on Intelligent Signal Processing and Communications Systems: "The Decade of Intelligent and Green Signal Processing and Communications", ISPACS 2011
DOIs
Publication statusPublished - 2011
Event19th IEEE International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2011 - Chiang Mai
Duration: 2011 Dec 72011 Dec 9

Other

Other19th IEEE International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2011
CityChiang Mai
Period11/12/711/12/9

Keywords

  • contour
  • depth
  • DIBR
  • MVC

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Signal Processing

Fingerprint

Dive into the research topics of 'Region-interior painting in Contour Based Depth map Coding System'. Together they form a unique fingerprint.

Cite this