显微细胞图像的自动分割

An Automatic Segmentation Method of Microscopic Cell Image

由于细胞组织本身的复杂特性以及显微镜的影响,细胞图像的分割成为图像分割中的一大难题。考虑到边缘检测可以准确的为分割提供边缘点,提出基于区域一致性测度的边缘检测算法,针对细胞图像的特点,实现其初始轮廓的自动获取,运用活动轮廓模型对初始轮廓进行优化,从而达到在大样本条件下直接对彩色细胞图像自动分割的目的。

Because of the nature of cell image and the influence of microscope,automatic segmentation for microscopic cell image becomes one difficult question.Because edge detection accurately indicates the edge points of the image, we present the edge detection algorithm based on region homogeneity measure. Then we can attain the initial contour automatically and fit the initial contour to true contour based on edge points using active contour models. At last we can realize the automatic segmentation of cell image.