Low-memory cost belief propagation architecture for disparity estimation

Yu Cheng Tseng*, Nelson Yen Chung Chang, Tian-Sheuan Chang

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

In disparity estimation, belief propagation can deliver better disparity quality than other algorithms but suffer from large storage cost, especially at the message update processing. To reduce the storage cost, this paper proposes low-memory cost architectures for the message update PE to satisfy the real-time application. We propose four architectures which are post-normalization, shadow buffer, no memory, and no memory+double PE architectures. Compared to the previous design, the proposed no memory+double PE architecture can save 28% of the hardware cost at most for 320x240@30fps and 64 disparity levels.

Original languageEnglish
Title of host publication2009 IEEE International Symposium on Circuits and Systems, ISCAS 2009
Pages153-156
Number of pages4
DOIs
StatePublished - 26 Oct 2009
Event2009 IEEE International Symposium on Circuits and Systems, ISCAS 2009 - Taipei, Taiwan
Duration: 24 May 200927 May 2009

Publication series

NameProceedings - IEEE International Symposium on Circuits and Systems
ISSN (Print)0271-4310

Conference

Conference2009 IEEE International Symposium on Circuits and Systems, ISCAS 2009
CountryTaiwan
CityTaipei
Period24/05/0927/05/09

Fingerprint Dive into the research topics of 'Low-memory cost belief propagation architecture for disparity estimation'. Together they form a unique fingerprint.

Cite this