一种双门限语音端点检测算法

A Double Threshold Speech Endpoint Detection Algorithm

在实际的噪声环境中,传统的语音端点检测算法容易出现误判、虚判和漏判等问题。因此,本文提出了一种基于短时能量和短时自相关函数的双门限语音端点检测方法,通过分析语音短时能量和短时自相关函数,设计出一种双门限语音端点检测算法。把传统算法和本文算法应用在基于DTW算法的语音识别系统中并进行了Matlab仿真。通过对比实验数据,得出本文提出的算法能够显著提高语音识别系统的识别率的结论。

In the actual noise environment,traditional speech endpoint detection algorithms are prone to misjudgment, false judgment and missed judgment. Therefore, in this paper, a double threshold speech endpoint detection method based on short-time energy and short-time autocorrelation function is proposed. Through the analysis of speech short-term energy and short-term autocorrelation function,we design a double threshold speech endpoint detection algorithm.We apply the traditional algorithm and the algorithm in this paper to the speech recognition system based on DTW algorithm, and carry out the MATLAB simulation.By comparing the experimental data, it is concluded that the proposed algorithm can significantly improve the recognition rate of speech recognition system.