摘要： 讨论建立了一真彩色血液细胞显微图像微机自动识别系统。运用图像处理技术提取血液细胞形态、光密度、色彩和纹理特征，运用遗传算法进行特征选择。在此基础上用神经网络建立分类器，对细胞图像进行分类识别。

关键词: 白血球分类, 特征选择, 遗传算法, 神经网络

Abstract: This paper discusses how to develop automatic classification system for blood cell image. Using image processing technology, it extracts four features: geometric features，density features，color features, texture features. The method based on genetic algorithm is proposed to feature selection. Finally it uses the feature sets to construct the neural network classifier. The blood cell image can be recognized.

中图分类号: 

• TP391.41