Presentation 1997/12/19
Affine Parameter Estimation and Region Split of Moving Regions based on Principle of Superposition
Masato Kazui, Tomoaki Shirakawa, Nobuhiro Miki, Yoshihiko Ogawa,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Segmentation of dynamic scene using optical flow is an important research field in image recognition. However the estimation of the optical flow is sensitive to noise, many conventional methods are required to segment the image iteratively by using clustering algorithm. In contrast to this approach, simultaneous and direct estimation of multiple attribute in the image is possible by means of principle of superposition. In this paper, we propose a new method that simultaneously estimate a set of affine parameters of moving regions in dynamic image based on principle of superposition. We also propose noise reduction technique using spatio-temporal BPF to reduce aliasing in input data. Experimental result from real image sequence shows the effectiveness of the proposed method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) affine parameter / principle of superposition / region split / aliasing
Paper # PRMU97-189
Date of Issue

Conference Information
Committee PRMU
Conference Date 1997/12/19(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) Affine Parameter Estimation and Region Split of Moving Regions based on Principle of Superposition
Sub Title (in English)
Keyword(1) affine parameter
Keyword(2) principle of superposition
Keyword(3) region split
Keyword(4) aliasing
1st Author's Name Masato Kazui
1st Author's Affiliation Faculty of Engineering, Hokkaido University()
2nd Author's Name Tomoaki Shirakawa
2nd Author's Affiliation Faculty of Engineering, Hokkaido University
3rd Author's Name Nobuhiro Miki
3rd Author's Affiliation Faculty of Engineering, Hokkaido University
4th Author's Name Yoshihiko Ogawa
4th Author's Affiliation Faculty of Engineering, Hokkaido University
Date 1997/12/19
Paper # PRMU97-189
Volume (vol) vol.97
Number (no) 459
Page pp.pp.-
#Pages 6
Date of Issue