白细胞图像的柔性组合分割算法  

Flexible combination segmentation algorithm for leukocyte images

作　　者：顾广华[1] 崔冬[1] 

机构地区：[1]燕山大学信息科学与工程学院,秦皇岛066004

出　　处：《仪器仪表学报》2008年第9期1977-1981,共5页Chinese Journal of Scientific Instrument

摘　　要：根据细胞图像的饱和度分量S和绿分量G中白细胞核分布的特点,构造组合加权图像,柔性调节权系数,提取细胞核,实现白细胞的初步定位。针对图像中红、白细胞的粘连现象,利用细胞图像红分量R和蓝分量B的分布,构造组合图像,提取白细胞。本文提出的柔性组合分割算法简单,计算量小,具有很好的分割效果,且分量权系数可自动调节,具有普适性。The paper constructs a weighted combined image according to the distribution characteristics of leukocyte nuclei in the saturation and green component of the cell images. The weight coefficients of the combination algorithm can be adjusted flexibly. According to the combined cell image, nuclei can be extracted exactly, and utimately the primary location of leukocyte is identified. Aiming to the overlapping cell image in which the conglutination phenomenon between leukocyte and erythrocytes exists, the paper also constructs a combined image according to the characteristics of the red and blue component images to extract the leukocyte. It should be noted that the computation complexity in the segmentation algorithm is rather low. The leukocyte are segmented perfectly. The weight coefficients can be adjusted automatically, so the algorithm has some universality.

关 键 词：白细胞图像 柔性组合 细胞分割 细胞粘连 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]