A parallel and memory-efficient mean shift filter on a regular graph

Sungchan Park*, Youngmin Ha, Hong Jeong

*Corresponding author for this work

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

Abstract

Toward real-time mean shift, a high-speed and parallel mean shift filter on a 2D regular graph is presented in this paper. For an M by N image and with L iteration times, O(NML) time complexity of sequential computation is reduced to O(NL) with M processors, and O(NM) memory complexity is reduced to O(LM) when L is smaller than N. As a result, computational speed is improved by using cascaded parallel processors. Furthermore, the proposed filter is adequate for VLSI implementation due to a linear systolic array structure. In this paper, we present quantitative and qualitative experimental results by using images in The Berkeley Image Segmentation Dataset. The proposed parallel algorithm requires 6 times smaller data access range and 2 times smaller memory size than the standard mean shift filtering at 15 iterations.

Original languageEnglish
Title of host publicationProceedings The 2007 International Conference on Intelligent Pervasive Computing, IPC 2007
Pages254-259
Number of pages6
DOIs
Publication statusPublished - 2007
Externally publishedYes
Event2007 International Conference on Intelligent Pervasive Computing, IPC 2007 - Jeju Island, Korea, Republic of
Duration: 11 Oct 200713 Oct 2007

Publication series

NameProceedings The 2007 International Conference on Intelligent Pervasive Computing, IPC 2007

Conference

Conference2007 International Conference on Intelligent Pervasive Computing, IPC 2007
Country/TerritoryKorea, Republic of
CityJeju Island
Period11/10/0713/10/07

Fingerprint

Dive into the research topics of 'A parallel and memory-efficient mean shift filter on a regular graph'. Together they form a unique fingerprint.

Cite this