自适应模糊增强的乳腺癌CISH细胞边缘检出  

Edge detection of breast cancer CISH cells based on self-adaptive fuzzy enhancement

作　　者：刘秉瀚[1] 王伟智[1] 

机构地区：[1]福州大学数学与计算机学院,福州350002

出　　处：《中国体视学与图像分析》2010年第4期399-404,共6页Chinese Journal of Stereology and Image Analysis

基　　金：福建省自然科学基金项目(2009J01283);福建省科技计划重点项目(2008H0026)

摘　　要：目的提出一种自适应设定可调参数的双线性广义模糊增强的细胞Canny边缘检出方法。方法根据乳腺癌CISH病理图像组织灰度分布特征,自动选取阈值设置双线性广义模糊增强的可调参数,并用自适应模糊增强算法代替Canny算法中的高斯滤波器,在计算梯度的幅值和方向以及抑制非极大值梯度幅值的基础上,利用最大类别方差法求双阈值进行细胞的强弱边缘检出。结果实现了乳腺癌CISH细胞的强弱边缘检出。结论实验结果表明该方法提取细胞边缘效果理想,比较固定模糊增强参数而言,更具有效性和实用性。Objective To present a cell Canny edge detection method based on bilinear generalized fuzzy enhancement algorithm with self-adaptive adjustable parameters setting.Methods According to the tissue grey intensity distribution of breast cancer CISH pathological images,to automatically select the threshold to set the adjustable parameters of bilinear generalized fuzzy enhancement and to replace the Gaussian filter Canny algorithm with adaptive fuzzy enhancement algorithm.To calculate the dual-thresholds using maximum classification square error method and detect the edge of cells on the basis of calculating the gradient of the amplitude and direction as well as non-maxima suppression.Results The edge detection of breast cancer CISH cells is realized.Conclusions The experimental results show that the extracted cell edge using this method is satisfactory and is more effective and practical comparing with fixed fuzzy enhancement parameters.

关 键 词：CISH乳腺癌 细胞边缘 模糊增强 自适应 CANNY算子 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]