摘要
传统的小波阈值函数的去噪方法简单有效,若用在车载下的语音去噪,由于其去噪后的信号存在边缘模糊、高频信息丢失等缺陷,难以得到较好的去噪效果。为提高车载环境下的语音系统的去噪效果,首先对已有小波阈值去噪算法进行分析,针对其不足,提出改进小波阈值函数的去噪方法。其次,通过Matlab软件对不同小波及分解层数进行仿真,选取最优的小波和层数。最后,由软件仿真实验数据比较得出,在相同噪声影响下,该方法比已有方法输出的效果要好。
Traditional method of wavelet threshold de -noising function is simple and effective except being used in the vehicle＇s voice de - noising which due to the edge blur or presence of high - frequency information loss et al. In order to improve the effect of in - vehicle environment de - noising phonetic system, the first of existing wavelet threshold algorithm is analyzed for its shortcomings, and an improvement wavelet de - noising threshold function is proposed. Secondly, different wavelet and decomposition levels are simulated on Matlab and the best wavelet and layers are selected. Finally, the simula- tion experimental data shows that the output signal to noise ratio in the paper is better than the existing method in the same noise.
出处
《电声技术》
2015年第5期80-83,共4页
Audio Engineering
基金
广西千亿元产业产学研用合作项目(信科院编号0168)
广西省自然科学基金项目(2012GXNSFAA053221)
关键词
阈值函数
信号去噪
小波变换
小波选取
threshold function
signal de-noising
wavelet transform
wavelet select
