Presentation 2008-09-06
Abnormal Motion Detection at Escalator Scene based on Spatio-temporal Features
Yasuhiro Murai, Hironobu Fujiyoshi, Masato Kazui,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper presents a method for detecting abnormal motion at an escalator scene, in which the movement of the escalator provides a dynamic background. This method is based on the use of spatio-temporal features obtained by space-time patches. Our approach consists of three steps; dynamic background modeling by using the Gaussian mixture model, human region detection based on RealAdaboost, and calculation of irregularity measure by using weighted space-time gradients. The proposed method can detect abnormal motions from a scene with a dynamic background, that would be difficult to detect with the conventional method using CHLAC (Cubic Higher-order Local Auto-Correlation) features. Our experimental results show that using our method has about 27% higher performance than that of the conventional method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English)
Paper # PRMU2008-87,HIP2008-87
Date of Issue

Conference Information
Committee PRMU
Conference Date 2008/8/29(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) Abnormal Motion Detection at Escalator Scene based on Spatio-temporal Features
Sub Title (in English)
Keyword(1)
1st Author's Name Yasuhiro Murai
1st Author's Affiliation Department of Computer Science, Chubu University()
2nd Author's Name Hironobu Fujiyoshi
2nd Author's Affiliation Department of Computer Science, Chubu University
3rd Author's Name Masato Kazui
3rd Author's Affiliation Hitachi, Ltd., Hitachi Research Laboratory
Date 2008-09-06
Paper # PRMU2008-87,HIP2008-87
Volume (vol) vol.108
Number (no) 198
Page pp.pp.-
#Pages 8
Date of Issue