Presentation 1998/9/18
Real-time estimation for global/local motion affine parameters : for a real-time implementation by software
Masakatsu Kourogi, Yoichi Muraoka,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we propose a method of estimating global/local motion of affine parameters, which requires so little computational cost that its real-time implementation can be achieved. This method firstly tries to estimate the global motion of an image sequence by using pixel-wise matching scheme to verify motion vectors.Each motion vectors are generated based on random or spatio-temporal gradient of brightness.Then the method estimates affine parameters from the verified vectors by M-estimation. Secondly it compensates the estimated global motion and then extracts the pixels with large prediction error. The method similarly estimates the local motion parameters on the pixels. It repeats the local motion estimation until all local motion are compensated.We implement the method on a general PC and apply it on real image sequences. As a result, it proves to be effective and able to achieve real-time computation.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Motion estimation / Affine motion / Motion compensation / Motion segmentation / Real-time processing
Paper # PRMU98-91
Date of Issue

Conference Information
Committee PRMU
Conference Date 1998/9/18(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) Real-time estimation for global/local motion affine parameters : for a real-time implementation by software
Sub Title (in English)
Keyword(1) Motion estimation
Keyword(2) Affine motion
Keyword(3) Motion compensation
Keyword(4) Motion segmentation
Keyword(5) Real-time processing
1st Author's Name Masakatsu Kourogi
1st Author's Affiliation School of Science and Engineering, Waseda University()
2nd Author's Name Yoichi Muraoka
2nd Author's Affiliation School of Science and Engineering, Waseda University
Date 1998/9/18
Paper # PRMU98-91
Volume (vol) vol.98
Number (no) 275
Page pp.pp.-
#Pages 8
Date of Issue