基于子波变换的自适应滤波语音增强方法  

A Speech Enhancement Method Using Adapted Filter Based on Wavelet Transform

作　　者：沈亚强[1] 

机构地区：[1]浙江师范大学交通学院,金华321004

出　　处：《仪器仪表学报》2005年第z1期195-196,214,共3页Chinese Journal of Scientific Instrument

摘　　要：在对噪声语音信号进行多分辨率子波分析的基础上,对分解后子波变换系数在尺度上和尺度间进行自适应滤波处理,并用它们的线性组合重构其语音信号,获得一种语音增强方法。Based on the noise speech signal analyzing with multi-resolution wavelet transform, the speech signals are rebuild with linear combination of decomposed wavelet transform coefficients which are processed on scales and between scales with adapted filter, and a way can be obtained to remove the noises effectively from the noise speech signals, and both the rate between signal and noise and the subjective audio quality are improved very evidently.

关 键 词：子波变换 多分辨率分析 自适应滤波 语音增强 

分 类 号：TH7-55[机械工程—仪器科学与技术]