摘要： 基于统计学习理论的支持向量机分类算法，提出一种X光胸片异常筛查系统，能够自动判别胸片的正常和异常。为了提高SVM算法的效率，利用小波变换等预处理手段去除对判读无用的图像冗余信息，采用二维主成分分析进一步降低图像特征维数。实验结果表明，SVM用于医学X光片异常筛查可行且有效、识别率高。

关键词: X光片, 图像分类, 支持向量机, 二维主成分分析

Abstract: Based on Support Vector Machine(SVM), the system for the abnormality judgment of X-ray chest file is presented, which can classify the X-ray picture normal and abnormal automatically. In order to improve the efficiency of the SVM, the wavelet transform is adopted in the system to eliminate the redundancy information in image. Two-Dimensional Principal Component Analysis(2DPCA) is used for feature extraction. Experimental results show that the SVM-based method is feasible in X-ray abnormality judgment, and has good classification ability.

Key words: X-ray file, image classification, Support Vector Machine(SVM), Two-Dimensional Principal Component Analysis(2DPCA)

中图分类号: 

• TP391