论文部分内容阅读
基于小波分析的阈值语音信号去噪技术是语音信号处理的一个重要热点技术,由于语音信号的非平稳性,传统的小波阈值去噪算法虽然能够衰减一部分语音信号中的噪声,但这些算法会不可避免地造成有用语音信号的损失。小波包可以同时对语音的低频和高频部分进行分解,能更好保留语音信息,减少噪声对语音的影响;同时,文中提出一种新的阈值函数,它对噪声清除得更加干净。仿真结果显示,该阈值函数较软、硬阈值函数具有明显的优越性。
Threshold speech signal denoising based on wavelet analysis is an important hotspot technology in speech signal processing. Because of the non-stationary speech signal, the traditional wavelet threshold denoising algorithm can attenuate some of the noise in speech signal, but these algorithms are not Avoid causing loss of useful voice signal. At the same time, a new threshold function is proposed in this paper, which removes the noise and clears the noise. The simulation results show that the threshold function is softer and the hard threshold function has obvious advantages.