语音端点检测，有助于除去语音信号中的噪音部分，留下有用的携带信息的部分，有利于提取其中有用的语音特征参数，从而提高识别率。-Endpoint detection, helps remove the noise part of the speech signal, leaving useful information-bearing part, is conducive to extract the speech feature useful to improve the recognition rate.