首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于显著性的OTSU局部递归分割算法
引用本文:汪国有,邹玉兰,凌勇.基于显著性的OTSU局部递归分割算法[J].华中科技大学学报(自然科学版),2002,30(9):57-59.
作者姓名:汪国有  邹玉兰  凌勇
作者单位:华中科技大学图像识别与人工智能研究所图像信息处理与智能控制教育部重点开放实验室
基金项目:武器装备预研基金资助项目
摘    要:针对OTSU方法的局限性,提出了一种基于显著性的OTSU局部递归分割算法。该算法的思路是把一次OTSU分割得到的目标区域作为新的图像进行下一次OTSU分割,依次递归下去,直到满足一定条件为止,然后把每次分割得到的目标区域按照一定加权原则累加起来,得到分割结果图像。实验结果表明,在不同的冷背景下,基于显著性的OTSU局部递归分割算法能够稳定地实现目标分割提取,具有较好的鲁棒性。

关 键 词:OTSU局部递归分割算法  图像分割  阈值  显著性  目标分割  OTSU准则
文章编号:1671-4512(2002)09-0057-03
修稿时间:2002年3月25日

An algorithm for salience-based local recursive OTSU segmentation
Wang Guoyou Zou Yulan Ling Yong Prof., Institute for Pattern Recognition & AI,Huazhong Univ. of Sci. & Tech.,Wuhan ,China..An algorithm for salience-based local recursive OTSU segmentation[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,2002,30(9):57-59.
Authors:Wang Guoyou Zou Yulan Ling Yong Prof  Institute for Pattern Recognition & AI  Huazhong Univ of Sci & Tech  Wuhan  China
Institution:Wang Guoyou Zou Yulan Ling Yong Prof., Institute for Pattern Recognition & AI,Huazhong Univ. of Sci. & Tech.,Wuhan 430074,China.
Abstract:In view of the limitations of OTSU algorithm, an algorithm for salience based local recursive OTSU segmentation (SBLR OTSU) is proposed. The main technique of SBLR OTSU consists in that the object area resulting from OTSU is regarded as the whole image to be segmented at the next OTSU segmentation. The process goes on sequentially until definite conditions are satisfied. Then all the object areas resulting from OTSU are accumulated to yield the image of segmentation results. It is shown by experimental results that, under different cold background conditions, the algorithm in question is robust against background clutter and capable of realizing object segmentation and extraction.
Keywords:image segmentation  OTSU algorithm  threshold  local recursion  salience
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号