基于谱相减改进算法的语音增强系统的实现研究  

Research and Implementation of the System of Speech Enhancement Based on the Improved Spectral Subtraction Algorithm

作　　者：覃贵礼[1] 罗云芳[1] 潘泽锴[1] 

机构地区：[1]广西职业技术学院计算机与电子信息工程系,广西南宁530226

出　　处：《河池学院学报》2016年第2期73-78,共6页Journal of Hechi University

摘　　要：针对数字通信网络中广泛应用的语音增强处理,在原有谱相减算法语音增强处理原理基础上改进该算法,并基于改进的谱相减算法采用MATLAB软件开发设计语音增强处理系统,经过测试表明系统能很好地实现带噪语音增强处理,声音的质量和可懂度提高效果显著。Speech enhancement is widely applied in the digital communication network. An improved spectral subtraction algorithm of speech enhancement is proposed and a speech enhancement system is designed by the MATLAB based on this improved algorithm. Experimental results show that the system can well achieve speech enhancement with noise. The quality and intelligibility of sounds are all improved effectively.

关 键 词：谱相减改进算法 语音增强 信噪比 MATLAB 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]