基于模糊算子的形态学神经网络及其应用  

Study on the Morphological Neural Network over Fuzzy Operators and Its Application

作　　者：程科[1] 

机构地区：[1]江苏科技大学电子信息学院,江苏镇江212003

出　　处：《江南大学学报（自然科学版）》2007年第6期874-877,共4页Joural of Jiangnan University (Natural Science Edition)　

摘　　要：构造了一种模糊形态学击中/丢失算子.结合形态学神经网络,提出了一种基于模糊击中/丢失算子的模糊形态学网络,并采用模糊形态学网络实现了血涂片图像中白细胞的定位/计数.实验结果表明,这种网络可以较好地处理细胞图像处理中出现的背景干扰、细胞粘连和部分遮挡的情况.The paper presents a new kind of fuzzy operator, a kind of fuzzy morphological neural morphological hit/miss operator. Based on the network （FMSNN） is developed. During its application on locating/counting of white blood cell, the results show that the FMSNN is capable of dealing with such conditions as background, cell connection and the case partially occluded.

关 键 词：白细胞检测 模糊形态学 模糊击中/丢失 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]