Paper Abstract and Keywords |
Presentation |
2008-10-24 14:50
Robust Background Subtraction by Statistic Reach Feature Kenji Iwata, Yutaka Satoh, Ryushi Ozaki, Katsuhiko Sakaue (AIST) PRMU2008-105 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
A statistic reach feature (SR feature) is defined as two arbitrary points that keep the sign of increase and decrease of the brightness in the image sequence. The SR feature realizes robust background subtraction for illumination changes and some noises. This paper describes a definition of the SR feature, and the application to background subtraction. In addition, we propose the two pass method that revalues the SR feature based on the background subtraction result. False positive appears when the foreground is contained in the reach point. The two pass method can improve by excluding the reach point that is the foreground. The experimental results show that the proposal method can segment the background with stability even if the image sequence contains illumination changes. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
background subtraction / statistic reach feature / SR feature / ISC / RRC / feature extraction / / |
Reference Info. |
IEICE Tech. Rep., vol. 108, no. 263, PRMU2008-105, pp. 97-102, Oct. 2008. |
Paper # |
PRMU2008-105 |
Date of Issue |
2008-10-16 (PRMU) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
PRMU2008-105 |