基于子波变换的语音去噪方法

本文在分析了随机噪声的子波变换系数在不同尺度上的传递特性和噪声信号奇异性与子波模极大值的关系后，提出了用一个尺度间变化的门限阈值来抑制带噪语音信号在不同尺度上噪声子波系数，从而实现在重构语音信号中消除噪声的目的。文中还给出了不同信噪比语音信号的子波去噪的计算机仿真结果，从结果上看出，本文的方法有较好的语音去噪、增强效果。

Speech De-noising Based on Wavelet Transform

According to the characteristics of random noise wavelet transform on the different scale and the relationshiop between the noise Lipschits and its wavelet transform modulus maxima, We present an algorithm that wavelet transform coeffcients of noise-speech signal are filtered by changing threshold on the different scale in order to reduce noise in reconstruct speech signal. We alos give some computed de-noising emulated results of some dfferent SNR speech. From the results, we effectively remove the noise in pronunciation and enhance the SNR of reconstruct speech.