语音编码中的新阈值小波去噪算法  

The Algorithm of Wavelet-Thresholding Denoising in Speech Encoding

作　　者：郭猛[1] 陈向东[1] 石念[1] 

机构地区：[1]西南交通大学信息科学与技术学院,四川成都610031

出　　处：《微电子学与计算机》2009年第5期138-141,共4页Microelectronics & Computer

基　　金：国家自然科学基金项目(60572026);教育部新世纪优秀人才支持计划(A0160419950120)

摘　　要：为提高语音信号的编码质量和节省宽带资源,文中在传统的CVSD编码中加入了小波阈值去噪方法.首先讨论了小波阈值去噪中估计小波系数的软阈值和硬阈值方法,然后提出了一种改进方案.该方案在阈值函数中加入因子,可以自适应地减少阈值函数中的恒定偏差,最后对重构后的语音信号再进行CVSD的编译码.Matlab仿真实验证明了该改进方案的有效性和优越性.In order to improve the encoding quality of speech signal and save the resource of bandwidth,we added the method of wavelet-thresholding denoising based on the traditional CVSD coding.We first present soft-thresholding and hard-thresholding denoising methods in wavelet thresholding denoising,then it presents a new thresholding function based on them,we add a factor to the modified project,which can be self-adaptivly reduce the fixed deviation in soft-thresholding function.Subsequently,we encode and decode the speech signal which has been reconstructed based on CVSD.In this paper,we validate the effectiveness and superority of the improved scheme by Matlab simulation.

关 键 词：CVSD 小波变换 阈值函数 

分 类 号：TN911[电子电信—通信与信息系统]