中文摘要：

随着信息技术和通信技术的迅速发展，说话人识别越来越受到重视，目前在说话人识别中，要提高识别率需要解决如何选取合适的有效的表征说话人特征的可靠参数和选取恰当的识别模型，文中在matlab实现了一个说话人识别系统，并在此基础上对系统进行了改进。

英文摘要：

with the development of communication and information technology, the speaker recognition attracts more and more attention from the people. Currently in the field of speaker recognition, there still exist two important unsolved problems for the enhancement of recognition rate, that is, how to select the best feature coefficient for the speaker system and how to select more effective and more reliable mode of speaker recognition. This article gives discussion of those two problems. And based on this some research and improvement is als made.