Presentation 2010-03-16
Recognition of specific behaviors in crowded video sequences based on human trajectory
Masaki TAKAHASHI, Mahito FUJII, Masahiro SHIBATA, Shin'ichi SATOH,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We describe a method that can detect specific human behaviors even in crowded surveillance video scenes. Our proposed method uses a HOG descriptor and SVM classifier to detect people, and it tracks the regions that contain people, called human regions, by calculating two-dimensional color histograms. The method recognizes specific behaviors by the features of human region trajectories. It identifies human behaviors by estimating the similarities to the reference trajectory of each behavior in the trajectory feature space. Verification techniques such as backward tracking and calculating optical flows helped reduce false alarms. Experiments showed our system could recognize several specific behaviors robustly even in crowded scenes.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Human behavior recognition / Surveillance / HOG
Paper # PRMU2009-304,HIP2009-189
Date of Issue

Conference Information
Committee HIP
Conference Date 2010/3/8(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 Human Information Processing (HIP)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Recognition of specific behaviors in crowded video sequences based on human trajectory
Sub Title (in English)
Keyword(1) Human behavior recognition
Keyword(2) Surveillance
Keyword(3) HOG
1st Author's Name Masaki TAKAHASHI
1st Author's Affiliation NHK Science and Technology Research Laboratories()
2nd Author's Name Mahito FUJII
2nd Author's Affiliation NHK Science and Technology Research Laboratories
3rd Author's Name Masahiro SHIBATA
3rd Author's Affiliation NHK Science and Technology Research Laboratories
4th Author's Name Shin'ichi SATOH
4th Author's Affiliation National Institute of Informatics
Date 2010-03-16
Paper # PRMU2009-304,HIP2009-189
Volume (vol) vol.109
Number (no) 471
Page pp.pp.-
#Pages 6
Date of Issue