Presentation 2000/10/13
Contour Segmentation for overlapped Object Region
Katsumi MORI, Eiji Wtanabe, Shigekazu KATAGIRI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In scene images, objects are overlapping each other according to their positions. Then, contours of regions become range of different objects outlines. Therefore, it is necessary to divide every different object outline for the object recognition. In this method, the division is made by using contours of neighboring regions on the assumption that space of adjoining contour is almost fixed. First, contours are individually approximated by polygons and dominant points are obtained. Next, dominant points at which over three regions are contact are detected as division points. As a result of the experiment, it was confirmed that the segmentation of long and smooth curves was well carried out.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) contour segmentation / neighboring region / Maximum Deviation Method / curvature
Paper # PRMU2000-97
Date of Issue

Conference Information
Committee PRMU
Conference Date 2000/10/13(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) Contour Segmentation for overlapped Object Region
Sub Title (in English)
Keyword(1) contour segmentation
Keyword(2) neighboring region
Keyword(3) Maximum Deviation Method
Keyword(4) curvature
1st Author's Name Katsumi MORI
1st Author's Affiliation Department of Information Processing Engineering, Fukuyama University()
2nd Author's Name Eiji Wtanabe
2nd Author's Affiliation Department of Information Systems and Management Science, Konan University
3rd Author's Name Shigekazu KATAGIRI
3rd Author's Affiliation Department of Information Processing Engineering, Fukuyama University
Date 2000/10/13
Paper # PRMU2000-97
Volume (vol) vol.100
Number (no) 360
Page pp.pp.-
#Pages 8
Date of Issue