云密切度判定法及其在分水岭分割中的应用

Cloud Affinity Judging Method and Its Application in Watershed Segmentation

针对传统的云模型在进行概念提升时,对什么样的子云适合进行合并没有给出一个简便有效的方法问题,提出了云密切度的概念,同时从云模型相对距离和交叉部分大小的角度进行了定义,形成云密切度判定方法;将使用该判定方法的云综合算法应用于图像分水岭分割算法中,成功解决后者易产生过分割的问题。选择两组医学图像进行分割实验,结果表明改进的分水岭算法获得了良好的分割效果,同时验证了云密切度判定方法的实用性。

Cloud model needs to merge cloud in concept generalizing.In traditional merge process there is a lack of a simple and effective method to decide what kind of cloudis suitable for merging.To solve this problem a new concept of cloud offinity is proposed.This concept makes the decision more reasonable because it is defined from the point of view of cloud model relative distance and cross section size.Cloud synthesis algorithm with this judgment method can successfully solve the over segmentation problem of image watershed segmentation algorithm.Choosing two groups of medical image for segmentation experiment,the results show that the improved watershed algorithm gives better segmentation effect.At the same time the practicability of affinity degree decision method is verified.