医疗卫生装备2013，Vol.34Issue(2)：21-22,2.DOI:10.7687/J.ISSN.1003-8868.2013.02.021
基于自适应阈值的小波包非接触语音增强方法研究
Non-contact Speech Enhancement Algorithm Based on Wavelet Packet Adaptive Threshold
薛慧君 1李盛 1路国华 1张杨 1焦腾 1王健琪 1荆西京1
作者信息
摘要
Abstract
Objective To develop a new method for removing the noise of non -contact speech. Methods Based on the capability of wavelet packet for analyzing time-frequency signal, an algorithm of wavelet packet threshold was developed by using hard threshold and soft threshold for denoising radar speech. Results Compared with spectral subtraction and Wiener filter speech enhancement algorithm, the proposed algorithm could remove the component of noise availably and reserve the aboriginal pure speech signal. Conclusion This method can remove the noise of radar speech and is practical.关键词
非接触语音/小波包/阈值/语谱图Key words
non-contact speech/ wavelet packet/ threshold/ spectrogram分类
医药卫生引用本文复制引用
薛慧君,李盛,路国华,张杨,焦腾,王健琪,荆西京..基于自适应阈值的小波包非接触语音增强方法研究[J].医疗卫生装备,2013,34(2):21-22,2.
