语音信号的预处理和特征提取技术

The Technology of Speech Signal Pre-processing and Feature Extraction

语音信号处理是一门多学科交叉的综合学科，它包含了语音学和数字信号处理等基础学科。这篇文章对语音信号作了两个方面的研究：语音信号的预处理和语音信号的特征提取。预处理是为了更好地进行语音信号的特征提取。是语音信号特征提取获得成功的重要保障。语音信号的预处理介绍了信号的主分量分析（PCA）技术和白化（whitening）技术．而语音信号的特征提取分为时域的和频域的特征提取。并用Madab编程实现了一段语音信号的分析处理。

Speech signal processing is a multi-disciplinary subject intersected by phonetics, digital signal processing, and so on. The article did the research about speech signal pre-processing and feature extraction. Pre-processing is prepared for the extraction, and guaranteed for the success of the feature extraction. The first part introduced the PCA and Whitening of the speech signal. The other part discussed the speech signal＇s time-domain and frequency-domain＇s features, and used Matlab to analyze one section of speech signal.