Presentation 2002/10/11
Segmentation of Images by Depth-First Searching of Minimum Spanning Trees
Natsuki TSUNODA, Toshinori WATANABE, Ken SUGAWARA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We study an image segmentation techique based on the minimum spanning tree (MST). In the segmentation, MST of an undirected connected graph having color pixels as nodes and color vector value differences between nodes as weights of edges is constructed. Next, eliminating those branches whose weights are larger than a given threshold value, segments are given. In this paper, noting that MST is a tree, we propose a new effective segmentation technique using the depth-first search of the MST. Noise elimination method is also discussed.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) minimum spanning tree / image segmentation / depth-first search / noise
Paper # PRMU2002-102
Date of Issue

Conference Information
Committee PRMU
Conference Date 2002/10/11(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Segmentation of Images by Depth-First Searching of Minimum Spanning Trees
Sub Title (in English)
Keyword(1) minimum spanning tree
Keyword(2) image segmentation
Keyword(3) depth-first search
Keyword(4) noise
1st Author's Name Natsuki TSUNODA
1st Author's Affiliation Graduate School of Infomation System, University of Electro-Communications()
2nd Author's Name Toshinori WATANABE
2nd Author's Affiliation Graduate School of Infomation System, University of Electro-Communications
3rd Author's Name Ken SUGAWARA
3rd Author's Affiliation Graduate School of Infomation System, University of Electro-Communications
Date 2002/10/11
Paper # PRMU2002-102
Volume (vol) vol.102
Number (no) 380
Page pp.pp.-
#Pages 4
Date of Issue