Presentation 2008-03-10
Object segmentation in indoor scene by superpixel-based correspondence
Yusuke FUJITA, Tetsuya MATSUMOTO, Yoshinori TAKEUCHI, Hiroaki KUDO, Noboru OHNISHI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper proposes a system in which correspondence and segmentation are done for some pairs of indoor stereo images based on superpixels, a set of pixels with similar feature. Superpixels are obtained by applying the oversegmentation method using RGB intensity differences, and provide global features compared with just a pixel or edge. Finding corresponding superpixels between left and right images, we can reconstruct 3D information of surfaces. Superpixel-based correspondence is faster and more accurate than other methods. We show experimental results and comparison with the existing method demonstrating the validity of the proposed method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English)
Paper # IE2007-266,PRMU2007-250
Date of Issue

Conference Information
Committee PRMU
Conference Date 2008/3/3(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) Object segmentation in indoor scene by superpixel-based correspondence
Sub Title (in English)
Keyword(1)
1st Author's Name Yusuke FUJITA
1st Author's Affiliation Graduate School of Information Science, Nagoya University()
2nd Author's Name Tetsuya MATSUMOTO
2nd Author's Affiliation Graduate School of Information Science, Nagoya University
3rd Author's Name Yoshinori TAKEUCHI
3rd Author's Affiliation Information and Communications Headquarters, Nagoya University
4th Author's Name Hiroaki KUDO
4th Author's Affiliation Graduate School of Information Science, Nagoya University
5th Author's Name Noboru OHNISHI
5th Author's Affiliation Graduate School of Information Science, Nagoya University
Date 2008-03-10
Paper # IE2007-266,PRMU2007-250
Volume (vol) vol.107
Number (no) 539
Page pp.pp.-
#Pages 6
Date of Issue