Title :
A hybrid matching algorithm for depth estimation
Author :
Tae-Woo Kim ; Chunsoo Ahn ; Tae Min Cho ; Jitae Shin ; Hokyom Kim
Author_Institution :
Sch. of Inf. & Commun. Eng., Sungkyunkwan Univ., Suwon, South Korea
Abstract :
Depth estimation is one of the key technologies of free-viewpoint television (FTV) system because the accuracy of depth map affects directly the synthesized video quality at intermediate viewpoint. In order to obtain accurate depth information with low complexity, a hybrid matching (HM) algorithm for depth estimation method is proposed. It is composed of pixel matching (PM) and block matching (BM) mode. By using motion information in matching mode selection, the proposed algorithm is more complied with human visual system (HVS) which is more sensitive to moving regions than the static one. Experimental results show that the proposed algorithm can not only maintain the synthesized visual quality, but also reduce considerably the execution time in matching process.
Keywords :
image matching; video signal processing; BM; FTV system; HM algorithm; HVS; PM; block matching mode; depth estimation method; free-viewpoint television system; human visual system; hybrid matching algorithm; matching mode selection; synthesized video quality; synthesized visual quality; block matching; depth estimation; free-viewpoint television; human visual system; hybrid matching; motion information; pixel matching;
Conference_Titel :
Computing and Convergence Technology (ICCCT), 2012 7th International Conference on
Conference_Location :
Seoul
Print_ISBN :
978-1-4673-0894-6