一种分割重叠粘连细胞图像的改进算法  

An Improved Segmentation Algorithm for Clustered Cells

作　　者：潘晨[1] 闫相国[1] 郑崇勋[1] 

机构地区：[1]西安交通大学生物医学信息工程教育部重点实验室,西安710049

出　　处：《中国生物医学工程学报》2006年第4期390-395,共6页Chinese Journal of Biomedical Engineering

摘　　要：为了专门分割重叠粘连细胞的图像,提出一种基于局部流域变换的迭代剥离算法。具体分割过程分为两个阶段,序贯处理胞浆和核区重叠,逐渐剥离重叠细胞。每个阶段均通过自适应距离变换分离重叠细胞,利用单个细胞形态特征控制腐蚀程度使得种子区域最大化,然后通过流域变换重建得到细胞间边界。算法能够克服细胞多核、胞浆颗粒性等干扰,适合自动化处理。骨髓和外周血涂片的分割试验表明分割精度约是传统流域算法的2倍,而运算时间并没有增加很多。In order to separate the images of clustered cells, an improved algorithm based-on local watershed transform was presented in this work. It is constructed with two similar stages. The first one separated the ceils that were overlapping in cytoplasm, the second one separated the rest cells that were overlapping in nucleus. The sequential processing reduced the difficulty of separation. In each stage, an adaptive distance transformation, which was controlled by the criteria of morphological features of single cell, was used to divide clustered cells. So the seed regions prepared for the watershed transform were almost maximum. Finally the borderline between divided cells was built by watershed transform. The new method could cope with multi-nuclear cell and granularity of the cytoplasm. The experimental results in blood and bone marrow smears demonstrated the segmentation accuracy obtained by the improved algorithm was twice as high as the traditional watershed transform, and time cost did not increase much.

关 键 词：图像处理 流域变换 重叠细胞 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]