遗传神经网络在尿沉渣识别中的应用

Application of genetic neural network to recognition of urine sediment

在使用数字图像处理和模式识别技术进行智能识别过程中,特征提取和优选成为识别率问题的关键。本文给出了一个尿沉渣显微图像识别系统,重点阐述了遗传神经网络的原理和其在尿沉渣细胞分类的特征优选中的作用,实验表明使用遗传神经网络算法可以找到最优特征子集,简化分类器的设计,提高分类效率,最终得到满意的尿沉渣识别效果。

Feature extraction and selection are important for recognition rate in intelligent recognition using technologies of digital image processing and pattern recognition. This paper introduces a cell recognition system of urine sediment micrographs. Genetic neural network is used to realize optimal feature subset selection simplification of classifier design and increase of classification efficiency and thus the satisfying result is got.