进行端点检测，提取参数MFCC的基于动态时间规整的语音识别。-Endpoint detection, extraction of structured Speech recognition based on DTW which detects the endpoint and extracts MFCC as the parameter.