New RGB Color Depth and Upsampling Methods for Centralized Texture Depth Packing Formats

Abstract

In the past decades the demand of the 3D entertainment grows intensively To fit the existing 2DTV broadcasting channels frame compatible 3D formats are utilized to deliver 3D video information Among these frame compatible 3D formats the formats including texture frame and its depth map are considered as a good solution for multi-view 3DTV displays Due to down and up sampling procedures which reduce the quality of the depacked 3D videos the 2D-plus-depth packing formats bring seriously degradation both in the depacked texture frame and depth map Hence to improve the quality of the depacked 3D videos the centralized texture and depth packing (CTDP) formats were proposed [1] In this thesis the performances of the CTDP formats are further improved by an adjusted RGB color depth packing method and a hierarchical evolution weighted (HEW) depth sampling method The purpose of adjusted RGB method is to reduce the distortions caused by broadcasting compression The hierarchical evolution weighted upsampling method is to recover the spike distortion in upsampled depth map The simulations are conducted for uncoded and coded CTDP 3D video sequences and the results show the refined CTDP system gives better quality of the reconstructed depth as well as the synthesized view than the original CTDP system

Date of Award	2018 Jul 23
Original language	English
Supervisor	Jar-Ferr Yang (Supervisor)