摘要
传统的细胞分割方法,对于淋巴结这样有着复杂背景的显微图像处理效果并不理想。本文基于相似度的自动种子像素选取、基于分水岭的图像分割及模糊C均值区域合并和基于凹点搜索的重叠粘连细胞切分,研究了淋巴结的分割处理,取得了良好的分割效果。
Due to the complex background of the microscopic images of lymph node tissue,the traditional method of cell division leads to unsatisfactory results.By using knowledge such of color image segmentation,we proposed some methods which achieved well preprocessing result.The methods are as follows:(1) automatic seed pixel selection based on the similarity;(2) image segmentation based on watershed;(3) fuzzy C mean region merging;(4) overlapping cell adhesion segmentation based on the search of concave point.
出处
《内蒙古农业大学学报(自然科学版)》
CAS
北大核心
2012年第3期226-230,共5页
Journal of Inner Mongolia Agricultural University(Natural Science Edition)
基金
内蒙古农业大学科技创新团队项目(NDPYTD2010-8)
内蒙古自治区自然科学基金项目(NO.2009MS0906)
关键词
图像分割
分水岭算法
模糊C均值
区域合并
凹点搜索
Image segmentation
watershed
fuzzy Cmean
region merging
the search of concave point