基于小波基函数的语音增强研究

小波分析理论是一种新兴的信号处理理论，在时间上和频率上都有很好的局部性，小波变换可同时进行时域和频域分析，具有多分辨特性，特别适合于处理非平稳信号。小波基是由小波函数经过平移和伸缩得到，具有简单、灵活、随意的特性和多分辨分析的功能。通过计算机仿真小波基函数对语音去噪的实现，并分析了不同阀值函数在语音处理中的效果，仿真结果表明，这些方法的运用能有效改善语音的失真程度，具有一定的实践指导意义。

A Research of Speech Enhancement Based on Wavelet Base Function

Wavelet analysis theory, a new theory of signal process,has good localization in both frequency and time domains. Wavelet transforming can be carried out in time domain and frequency domain analysis at the same time and has properties of multi-resolution, so it is particularly suited to handle the non-stationary signals. Wavelet is obtained through translation and stretch, it has the characteristics of simplicity, flexibility and random and the function of multiresolution analysis. In this paper, a computer simulation of wavelet function is used to realize noise-removal, and effects of speech processing in different threshold functions. The simulation results show that the use of these methods can effectively improve the level of voice distortion, with some practical guiding significance.