|
|
【答案】应助回帖
★ ★ ★ ★ ★ ★
ltx200: 金币+3, ★★★★★最佳答案 2014-03-08 16:35:10
gruyclewee: 金币+3, 感谢应助，奖励一下。 2014-03-08 21:11:37
oven1986: 检索EPI+1, 感谢应助，鼓励一下。 2014-03-09 23:38:44
Image segmentation using adaptive loopy belief propagation
作者:Xu, SJ (Xu, Sheng-Jun)[ 1,2 ] ; Han, JQ (Han, Jiu-Qiang)[ 1 ] ; Yu, JQ (Yu, Jun-Qi)[ 2 ] ; Zhao, L (Zhao, Liang)[ 2 ]
OPTIK
卷: 124
期: 22
页: 5732-5738
DOI: 10.1016/j.ijleo.2013.04.013
出版年: 2013
查看期刊信息
摘要
Loopy belief propagation (LBP) algorithm over pairwise-connected Markov random fields (MRFs) has become widely used for low-level vision problems. However, Pairwise MRF is often insufficient to capture the statistics of natural images well, and LBP is still extremely slow for application on an MRF with large discrete label space. To solve these problems, the present study proposes a new segmentation algorithm based on adaptive LBP. The proposed algorithm utilizes local region information to construct a local region model, as well as a local interaction region MRF model for image segmentation. The adaptive LBP algorithm maximizes the global probability of the proposed MRF model, which employs two very important strategies, namely, "message self-convergence" and "adaptive label pruning". Message self-convergence can improve the reliability of a pixel in choosing a label in local region, and label pruning can dismiss impossible labels for every pixel. Thus, the most reliable information messages transfer through the LBP algorithm. The experimental results show that the proposed algorithm not only obtains more accurate segmentation results but also greater speed. (C) 2013 Elsevier GmbH. All rights reserved.
关键词
作者关键词:Adaptive loopy belief propagation; Markov random fields; Local interaction region MRF; Message self-convergence; Label pruning
KeyWords Plus:MARKOV RANDOM-FIELDS; ENERGY MINIMIZATION; GRAPH CUTS; VISION
作者信息
通讯作者地址: Xu, SJ (通讯作者)
Xi An Jiao Tong Univ, MoE Key Lab Intelligent Networks & Network Secur, Xian 710049, Peoples R China.
地址:
[ 1 ] Xi An Jiao Tong Univ, MoE Key Lab Intelligent Networks & Network Secur, Xian 710049, Peoples R China
[ 2 ] Xian Univ Architecture & Technol, Sch Informat & Control Engn, Xian 710055, Peoples R China
电子邮件地址:duplin@sina.com
基金资助致谢
基金资助机构
授权号
National Natural Science Foundation for Young Scientists of China
51209167
Natural Science Foundation of Shaanxi Province, China
2012JM8026
查看基金资助信息
出版商
ELSEVIER GMBH, URBAN & FISCHER VERLAG, OFFICE JENA, P O BOX 100537, 07705 JENA, GERMANY
类别 / 分类
研究方向:Optics
Web of Science 类别:Optics
文献信息
文献类型:Article
语种:English
入藏号: WOS:000325835400098
ISSN: 0030-4026
期刊信息
Impact Factor (影响因子): Journal Citation Reports®
其他信息
IDS 号: 236YM
Web of Science 核心合集中的 "引用的参考文献": 27
Web of Science 核心合集中的 "被引频次": |
|
回复此楼