基于小波变换的肝脏CT图像分类

摘要

A liver CT image classification method based on wavelet transform was presented. Firstly, wavelet and gray level co-occurrence matrix texture features were extracted. Secondly, Mahalonobis distance separability criterion and genetic algorithms were cornbined for feature selection and optimization. Finally, the support vector machine was used to classify the liver CT images. In this paper, features of two kind wavelets and extraction methods on the classification were discussed. Also the algorithm was simulated by software. The experiments showed that the liver CT images can be classified effectively by wavelet transform.%提出了一种基于小波变换的肝脏CT图像疾病的分类方法：首先提取小波和灰度共生矩阵纹理特征，其次结合马氏距离的可分性判据和遗传算法进行特征选择及优化，最后利用支持向量机将肝脏CT图像进行分类。讨论了2种小波以及特征提取方式对分类结果的影响，并通过软件仿真实现算法。实验表明：小波变换可对肝脏CT图像进行有效的分类。