Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (28): 149-151.DOI: 10.3778/j.issn.1002-8331.2010.28.042

• 数据库、信号与信息处理 • Previous Articles     Next Articles

Application of Hilbert-Huang transform to fine time-frequency analysis of speech signal

SONG Qian-qian,YU Feng-qin   

  1. College of Communications and Control Engineering,Jiangnan University,Wuxi,Jiangsu 214122,China
  • Received:2009-03-02 Revised:2009-04-22 Online:2010-10-01 Published:2010-10-01
  • Contact: SONG Qian-qian



  1. 江南大学 通信与控制工程学院,江苏 无锡 214122
  • 通讯作者: 宋倩倩

Abstract: Speech signal is a typical non-stationary signal,its characteristic and feature parameters change along with time.Time-Frequency is a powerful tool for the analysis of time-varying spectrum.Hilbert-Huang Transform(HHT) is a new and self-adaptable method for time-frequency analysis.Its physical meaning is clear to nonlinear and non-stationary signal,it can get three dimensional distribution of time-frequency-amplitude.The theory of HHT is studied and the decomposition accuracy is improved by adopting the proper method to deal with end effect.Simulation experiments show that it can get finer time-frequency structure compared with STFT,WVD and Choi-Williams.

Key words: speech signal, time-frequency analysis, Hilbert-Huang Transform(HHT), Hilbert spectrum

摘要: 语音信号是一种典型的非平稳信号,其特性及表征本质特征的参数均是随时间变化的,而时频分析是分析时变谱的有力工具,Hilbert-Huang变换是一种新型的具有自适应性的时频分析方法,对于非线性、非平稳信号有清晰的物理意义,通过HHT变换,能够得到信号的时间-频率-振幅三维分布特征。分析了HHT算法的原理,采用了合适的端点效应处理方法提高了EMD的分解精度,通过仿真实验得到了语音信号更加精细的时频结构,并与STFT、WVD及Choi-Williams分布进行了对比,显示了HHT算法的优越性。

关键词: 语音信号, 时频分析, Hilbert-Huang变换, Hilbert谱

CLC Number: