摘要
基于云模型和模糊集理论,提出了一种新的边缘检测方法.该方法根据图像灰度,借助区域生长算法产生对象云;通过云运算生成边界云并构建模糊隶属带,进而在条件概率和模糊划分熵的基础上,根据最大模糊熵原则确定最优阈值,对图像模糊边界进行提取.试验结果表明,该算法能保留大量低灰度信息,可以获得较好的检测效果.
Based on the cloud model and the fuzzy set theory, a new edge detection method was proposed. Firstly, with this method, object clouds are generated using the region growth algorithm on the basis of the gray level of images. Secondly, boundary clouds and fuzzy membership regions for two or more clouds are obtained based on Boolean calculation. Thirdly, the optimal threshold is determined through maximizing the entropy of fuzzy partition based on conditional probabilities and eritropy of fuzzy partition. Finally, an edge detection procedure is executed on the edge images. The experimental results show that with the proposed method plenty of lower level gray can be preserved to obtain more favorable edge detection.
出处
《西南交通大学学报》
EI
CSCD
北大核心
2006年第1期85-90,共6页
Journal of Southwest Jiaotong University
基金
国家自然科学基金资助项目(40371098)
重庆市自然科学基金资助项目(CSTC2005BB2065)
关键词
云模型
对象云
模糊边界
模糊隶属带
边缘检测
cloud model
object cloud
fuzzy edge
fuzzy membership region
edge detection
