压缩感知理论在语音信号去噪中的应用

摘要

针对小波阈值滤波的局限性,将压缩感知理论应用到语音信号去噪中,并与小波阈值滤波方法进行了比较,仿真实验结果表明:基于压缩感知的小波滤波方法可以有效地去除语音信号中的噪声,并且去噪效果优于传统小波阈值滤波方法,对工程中音频信号的降噪具有指导意义.%According to limitation of wavelet filter, the compressed sensing theory was applied to sound denoising, and a comparison between wavelet filter based on compressed sensing and traditional wavelet filter was made through experiment. The experiment shows that the wave filter based on compressed sensing can effectively deduct speech noise, and the denosing effect of wavelet filter based on compressed sensing is better than traditional wavelet filter. The method has an engineering significance for sound signal denoising.