本文以语音信号LPC倒谱系数作为识别特征,运用矢量量化技术实现自动说话人识别.在一个42人、7700个语音的数据库上,我们进行了系统的识别实验,研究了不同系统参数对识别率的影响,得到了不少有参考价值的结果.本文还介绍了在此基础上研制成功的一个实时语声交互式身份确认系统,该系统作为计算机的语音锁,正确识别率达到95％以上. In this paper, we use the LPC Cepstral Coefficient of speech signal as recognition feature and use vector quantization technology to realize automatic speaker recognition.In a database of 42 persons and 7700 voices, we conducted a system identification experiment and studied the effect of different system parameters on recognition rate , We get a lot of results with reference value.This paper also introduces a real-time interactive voice recognition system based on this system, which is used as the voice lock of computer, and the correct recognition rate is more than 95%.