关键词: 小波分析, 心音信号识别, 支持向量机, 特征提取

Abstract: Recognition of heart sound signal plays an important role in the diagnosis of heart disease. In order to improve the performance of heart sound recognition, this paper presents an automatic recognition method of heart sound signal based on support vector machine. Firstly, wavelet analysis is used to reduce the noise of the heart sound signal, and then Mel frequency cepstral coefficients are extracted as the feature of heart sound, finally, the support vector machine (SVM) is used to build classifier of heart sound signal and the performance is tested by using heart sound data. The results show that the recognition accuracy of the heart sound signal is as high as 93% for the proposed method, it can automatically and correctly identify the normal and abnormal heart sound signal.


Key words:  wavelet analysis, heart sound signal recognition, support vector machine, feature extraction