基于EMD和改进双门限法的语音端点检测

Speech Endpoint Detection Based on EMD and Improved Double Threshold Method

语音端点检测的准确与否直接影响到语音识别系统的计算复杂度和识别能力,在基于短时能量和过零率的端点检测算法中,能量计算方法不尽合理而且在低信噪比下检测效果大大降低。对此提出了一种基于经验模式分解和改进双门限法的语音端点检测算法,仿真结果表明在低信噪比情况下本文算法有更好的端点检测能力,显示了算法的优越性。

The accuracy of speech endpoint detection affects the computational complexity and the recognition performance of the speech recognition system. In the algorithm based on short-time energy and zero-crossing rate, the calculating method about energy is unreasonable and can not work well in low signal to noise ratio environments. An algorithm based on EMD and improved double threshold is proposed, and the simulation results show this algorithm has better detection capability in low signal to noise ratio environments and takes on more advantages.