粘连血细胞图像分离计数算法的实现

目前相关文献中提出的分割算法很多,但都没有一种通用的图像分割算法。不论采用何种分割算法,其分割结果图像中总会存在一些细胞粘连的现象,因而影响到细胞的分析。本文利用数学形态学中的腐蚀运算对二值细胞区域进行层层剥离,然后根据剥离结果判断是否找到了细胞中心点,并对中心点进行适当的修正,解决了因细胞粘连或噪声等原因造成中心点的增加(减少),进而完成细胞的计数统计。文中主要叙述了算法的设计思想及实现,程序运行结果表明,基于vc++.net的算法设计是有效的。

Design on the Count Algorithm of Image Segmentation for Adhesive Blood Cell

There are a lot of segmentation algorithms listed in the relevant data at the moment, but it is difficult to find one that can be used in accordance with any cases. Adhesive cell will undoubtedly exist in the image of segmentation results more or less, whatever any segmentation algorithm is taken, which will affect the cell analysis. Corrosion calculation in the mathematics and morphology is used in this paper to exfoliate the cell area layer upon layer. Exfoliation results will be used to decide the correct position of center point, and amend the center point appropriately. This method will tackle the problem of increase and decrease of the center point numbers caused by cell cohesion or noise, and complete the count of the cell numbers. Design thoughts and realization are elaborated in this paper. Running results of the program prove that this algorithm design is valid on the basis of the vc++.net .