Presentation 2015-03-03
Human Tracking Using Particle Filter and Detection Based on Multiple Likelihoods
Kotaro UDONO, Akira KUBOTA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Recently, thanks to development of low-cost and high-performance imaging equipment, object detection and tracking methods have been studied, which are expected to be applied in the field such as suspicious person detection and buying psychology. Example of object tracking methods includes Particle Filter. It works with high precision for single object tracking; however it cannot handle well to track multiple objects crossing with each other. Therefore, we propose a novel tracking method that combines Particle Filter and Template Matching based on color histogram and ORB feature for improving the precision. The experimental results on real scenes that are difficult in tracking by Particle Filter showed that the proposed method improved the precision of tracking.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Particle Filter / Template Matching / ORB / Multiple Likelihoods
Paper # IMQ2014-43,IE2014-104,MVE2014-91
Date of Issue

Conference Information
Committee IMQ
Conference Date 2015/2/24(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 Image Media Quality(IMQ)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Human Tracking Using Particle Filter and Detection Based on Multiple Likelihoods
Sub Title (in English)
Keyword(1) Particle Filter
Keyword(2) Template Matching
Keyword(3) ORB
Keyword(4) Multiple Likelihoods
1st Author's Name Kotaro UDONO
1st Author's Affiliation Faculty of Engineering, Chuo University()
2nd Author's Name Akira KUBOTA
2nd Author's Affiliation Faculty of Engineering, Chuo University
Date 2015-03-03
Paper # IMQ2014-43,IE2014-104,MVE2014-91
Volume (vol) vol.114
Number (no) 485
Page pp.pp.-
#Pages 6
Date of Issue