通过语音分帧、端点检测、特征提取、DTW算法，实现多字语音识别(Through speech segmentation, endpoint detection, feature extraction and DTW algorithm, multi word speech recognition is realized.)