多信息融合的彩色细胞图像分割方法

本文就从病理切片的彩色图像中提取细胞核的有关方法作了研究，提出了有效的分割方法的策略。该方法以数学形态学为主要工具，可以利用细胞的多个特征，较好地解决了病理显微图像背景杂，光照不均及粘连团聚现象等因素给分割带来的困难。

A Segmentation of Color Cell Image of Multi-Information

An image segmentation processing for complicated color cell images is described. The mthod is based on a mathematical morphological region growing algorithm that makes use of complex features of a cell: grey level, gradients, color, and texture features. The problems and difficulty that the background of pathological microimages are complicated, illumination is not homogeneous, and there are adhesion and clustering, which make segmentation of the images difficult, can be dealt with in a better way.