非高斯零均值噪声环境下的语音增强算法  

OFDM Techniques of European Digital Television Transmission Standard DVB-T

作　　者：蔡萍[1] 

机构地区：[1]闽江学院物理学与电子信息工程系,福州350108

出　　处：《湖南理工学院学报（自然科学版）》2015年第1期26-28,共3页Journal of Hunan Institute of Science and Technology(Natural Sciences)

摘　　要：语音增强的目的就是尽量从带噪语音信号中提取纯净语音.谱减法由于其实现简单,实时性好等优点被广泛应用于语音增强中.然而,传统谱减法只能对高斯零均值的噪声信号进行增强.针对这一问题,提出一种能适应非高斯零均值噪声环境的改进算法.该算法处理后的语音具有较好的可懂度和信噪比,并大大削弱了音乐噪声.The purpose of speech enhancement is to extract pure speech signal from polluted speech signal as much as possible. Spectral subtraction algorithm is widely used in speech enhancement for its simple realization and good real-time character. However, traditional spectral subtraction algorithm can only be applied on Gaussian and zeromean noise. To solve this problem, put forward an improved algorithm which is adapted to non-Gaussian and non-zeromean noise environment. Being processed by this algorithm, speech signal has relatively high understandability and signal-to-noise ratio, and has less influence of musical noise.

关 键 词：语音增强 谱减法 信噪比 

分 类 号：TN912.3[电子电信—通信与信息系统]