Paper Abstract and Keywords |
Presentation |
2008-03-10 09:30
Extraction of Object Regions Using Shape Map and Local Features Koji Imaeda (Nagoya Univ.), Miyako Baba, Nobuyuki Shiraki, Akihiro Watanabe (Toyota Central R&D Labs.), Yasushi Hirano, Shoji Kajita, Kenji Mase (Nagoya Univ.) IE2007-265 PRMU2007-249 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
In this paper, a novel method for extraction of a precise object region from a given rough initial region is proposed. The existing method for object region extraction using a kernel density estimator has problems, e.g. influence by the background edge in the initial region. The method is considered effective only when the image texture is simple or the initial region is considerably accurate. The proposed method, even if the initial region shape is rough and the image texture is complex, can correctly extract the object area by introducing a brightness histogram as a probability density function and an object shape distribution map as a priori probability for the preprocess of the kernel density estimator. It was confirmed that the accuracy and computation speed in general cases are improved in comparison to those of existing method. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
object extraction / histogtam / kernel density estimator / joint probability density function / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 107, no. 539, PRMU2007-249, pp. 23-28, March 2008. |
Paper # |
PRMU2007-249 |
Date of Issue |
2008-03-03 (IE, 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 |
IE2007-265 PRMU2007-249 |
|