摘要
在采集语音时常常不可避免地混入背景噪声，使语音质量下降，从而影响了语音处理中各种算法的效果。因此研究语音消噪系统是非常必要的。该文分析了多通道语音增强系统，提出一种采用两级滤波器的消噪系统。计算机模拟结果表明，其消噪性能明显优于Ｚｅｌｉｎｓｋｉ法［１］，听音试验也证实了这一点。
The speech received under many circumstances is often disturbed by various noises. This seriously decreases the intelligibility and articulation of the received speech. Therefore, it is essential to use some method to extract the clean speech. In this paper, we propose a new algorithm with two stage Wiener filter to suppress noise, both its SNR and the intelligibility of the filtered speech are better than those of Zelinskis method.
出处
《南京理工大学学报》
EI
CAS
CSCD
1997年第4期305-308,共4页
Journal of Nanjing University of Science and Technology
关键词
噪声控制
信噪比
滤波器
语音信号处理
speech sounds, signal processing, noise suppression
signal noise ratio, filter
