Canny算子与数学形态学算子相结合实现胸水细胞的边缘提取  

Edge detection for pleura effusion cell images based on combining of morphology and canny edge detection algorithm

作　　者：赵英红[1] 乔双[1] 董丽华[2] 

机构地区：[1]东北师范大学物理学院,吉林长春130024 [2]吉林大学第一附属医院放疗科,吉林长春130022

出　　处：《东北师大学报（自然科学版）》2008年第3期42-45,共4页Journal of Northeast Normal University(Natural Science Edition)

基　　金：吉林省自然科学基金资助项目(20050705-6)

摘　　要：针对胸水病理细胞显微图像的灰度分布不均匀性、灰度级范围窄等情况,提出了利用数学形态学方法与Canny算子相结合的算法对其进行边缘检测.首先利用Canny算子对细胞进行边缘的初次分割提取,然后利用数学形态学中的线形和菱形结构元素及填充函数对细胞边缘进行精确提取,获得了较好的实验结果,达到了对胸水细胞边缘自动检测的目的,有助于发现早期病变的细胞.A new cell image edge detection algorithm, which combines the Canny algorithm and morphology, is presented in this paper. The difficulty of the edge detection of the pleura effusion cell image lies in the complex shapes of the cells, variant and narrow range of gray levels, non-random noise. The algorithm presented tries to solve the problem effectively. It has been also shown that the algorithm produces good results, for relatively higher accuracy of pleura effusion cell image edge detection was gotten.

关 键 词：胸水细胞 数学形态学 CANNY算子 边缘检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]