期刊文献+

利用脊的特征进行信号盲分离 被引量:7

Blind Source Separation Based on Ridge Characterization Extraction
下载PDF
导出
摘要 本文研究了一种新的盲信号分离方法 ,利用信号在时频变换域上的一些形态特征进行信号分量的分离和重构 .稳定位相法的理论表明 ,渐近信号在连续小波变换域上的分布呈现明显的脊的形态 ,脊上的数据表现出和原信号最相似的特性 ,各个信号分量的主要参数均可以从各自对应的脊上的信息中提取出来 ,并且完全重构信号 .本文利用这一特点对由多个非平稳信号混合而成的源信号进行盲分离 。 A new blind source separation method based on ridge characterization extraction is presented. The stationary phase method proves that in the asymptotic limit, the behavior of the continuous wavelet coefficients shows strong ridge character. By extracting the ridge characterization of each signal component, some parameters of the signal component can be extracted and used in reconstruction of the original signal. This feature can be used in blind source separation of a set of instantaneous linear mixtures of non-stationary signals. Numerical simulations are provided.
出处 《电子学报》 EI CAS CSCD 北大核心 2004年第7期1156-1159,共4页 Acta Electronica Sinica
关键词 连续小波变换 脊提取 信号盲分离 Computer simulation Feature extraction Signal reconstruction Wavelet transforms
  • 相关文献

参考文献6

  • 1A Belouchrani,M Amin.Blind source separation using t-f distributions:algorithm and asymptotic performance [A].Proc ICASSP [C].Germany:IEEE,April 1997.3469-3472. 被引量:1
  • 2C Gordan,R Reiz.Using wavelet transform for the ridge extraction of a parabolic frequency modulated signal [A].Digital Signal Processing 2002 [C].Greece:IEEE,2002.337-340. 被引量:1
  • 3N Delprat,et al.Asymptotic wavelet and gabor analysis:extraction of instantaneous frequencies [J].IEEE Trans Inform Theory,1992,38(2):644-665. 被引量:1
  • 4N Delprat.Global frequency modulation laws extraction from the Gabor transform of a signal [J].IEEE Trans Speech and Audio Processing,1997,5(1):64-71. 被引量:1
  • 5R Carmona,W L Hwang,B Torr′esani.Characterization of signals by the ridges of their wavelet transforms[J].IEEE Trans Signal Processing,1997,45(10):2586-2590. 被引量:1
  • 6R Carmona,W L Hwang,B Torr′esani.Multiridge detection and time-frequency reconstruction[J].IEEE Trans Signal Processing,1999,47(2):480-492. 被引量:1

同被引文献137

引证文献7

二级引证文献75

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部