介绍了一种基于 MA T L A B的多个特定人连接词语音识别的方法，并提出了在进行端点检测时，引入平均的概念能进一

步提高识别率。 此设计是以 L P C C系数、 D T W 算法为核心的基于图形界面的设计。 通过大量的实验测试 ，表明该方法基本达

到屏蔽外界环境的影响 ， 具有非常高的精度识别-: In this paper based on a number of specific persons MA TLAB conjunctions speech recognition method, and presented during endpoint detection, the introduction of the concept of the average recognition rate can be further improved. This design is LPCC coefficients, DTW algorithm as the core design based on graphical interface. Through a large number of experimental tests show that the method is basically to shield the external environment, with very high accuracy recognition