摘要
针对水下信道和噪声环境相对复杂,信噪比低等问题,充分利用HHT对于非线性、非平稳信号的处理优势,以及语音和噪声的信号的统计特性,首先对含噪语音信号进行EMD分解,然后对各个IMF分量进行自相关分析,完成语音主导IMF分量的筛选,实现语音信号重构和增强。采用本文提出的HHT自相关语音增强方法,对于高斯白噪声和海洋噪声污染语音信号进行对比实验,结果表明:在相同信噪比时,海洋噪声污染信号可以实现更大程度的信噪比提升。该算法在水下语音通信系统中,增强效果相对显著。
In underwater channel and noise environment are relatively complex and the SNR is low.Make full use of processing advantages of Hilbert Huang Transform(HHT)method for non-linear and non-stationary signals,as well as statistical characteristics of noisy speech signals,a new HHT method combined with autocorrelation is proposed in this paper.Firstly,Empirical Mode Decomposition(EMD)decomposition is carried out to obtain IMF components,then auto-correlation analysis is carried out to complete the selection of speech-dominated Intrinsic Mode Function(IMF)components,so that speech reconstruction and enhancement is completed.The HHT autocorrelation speech enhancement method proposed in this paper is used to compare the speech signals polluted by white Gaussian noise and ocean noise.The results show that the marine noise polluted signal can achieve a greater degree of signal-to-noise ratio improvement at the same signal-to-noise ratio.The enhancement effect of this algorithm in underwater voice communication system is relatively obvious.
作者
王光艳
杨秀芬
祝琼珂
罗雨章
江淇
WANG Guang-yan;YANG Xiu-fen;ZHU Qiong-ke;LUO Yu-zhang;JIANG qi(School of Information Engineering,Tianjin University of Commerce,Tianjin 300134,China)
出处
《新一代信息技术》
2019年第10期10-17,共8页
New Generation of Information Technology
基金
天津市企业科技特派员项目(项目编号:18JCTPJC66900)
天津市大学生创新创业训练计划项目(项目编号:201810069005)资助课题。
