Presentation 2004/12/9
Segmentation of Microscopic Images based on a Novel Edge Interpolation and Enhancement Technique
Toshiyuki YOSHIDA, Shunsuke OHMAE, Kozo KANEDA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Grain size analysis for microscopic images has been used as an evaluation technique for the mechanical strength of steel materials. The measurement of grain size distribution requires a segmentation for crystalline granularity, and is currently done manually by specialists. This paper hence proposes an automatic segmentation technique for crystalline granularity based on the watershed algorithm. To alleviate oversegmentation problem and suppress errorneous region merging due to broken edges, a novel edge interpolatin and enhancement technique based on electric-circuit simulation, which has been proposed by one of the authors, is used together with the watershed algorithm. Experimental results illustrate that the proposed algorithm offers precise segmentation results for those images with relatively clear boundaries of regions.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) microscopic images / crystalline grain / segmentation / edge interpolation / edge enhancement / electrical-circuit model
Paper # SIS2004-45
Date of Issue

Conference Information
Committee SIS
Conference Date 2004/12/9(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 Smart Info-Media Systems (SIS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Segmentation of Microscopic Images based on a Novel Edge Interpolation and Enhancement Technique
Sub Title (in English)
Keyword(1) microscopic images
Keyword(2) crystalline grain
Keyword(3) segmentation
Keyword(4) edge interpolation
Keyword(5) edge enhancement
Keyword(6) electrical-circuit model
1st Author's Name Toshiyuki YOSHIDA
1st Author's Affiliation Dept. of Information Science, Faculty of Eng., Fukui University()
2nd Author's Name Shunsuke OHMAE
2nd Author's Affiliation Dept. of Information Science, Faculty of Eng., Fukui University
3rd Author's Name Kozo KANEDA
3rd Author's Affiliation Information System Division, MITANI Corporation
Date 2004/12/9
Paper # SIS2004-45
Volume (vol) vol.104
Number (no) 511
Page pp.pp.-
#Pages 8
Date of Issue