Time-frequency analysis: theory and applications
Time-frequency analysis: theory and applications
Signal Period Analysis Based on Hilbert-Huang Transform and Its Application to Texture Analysis
ICIG '04 Proceedings of the Third International Conference on Image and Graphics
An EMD-based recognition method for Chinese fonts and styles
Pattern Recognition Letters
A novel pitch period detection algorithm based on hilbert-huang transform
SINOBIOMETRICS'04 Proceedings of the 5th Chinese conference on Advances in Biometric Person Authentication
IEEE Transactions on Signal Processing
An improved envelope algorithm for eliminating undershoots
Digital Signal Processing
Hi-index | 0.00 |
Empirical mode decomposition can decompose a complicated signal into a sum of finite intrinsic mode functions whose instantaneous frequencies computed by the analytic signal method give a physically meaningful characterization of the signal. However, it is found that computing the instantaneous frequency of an intrinsic mode function through the Hilbert transform may produce some ridiculous results. Recently, an empirical AM/FM (amplitude and frequency modulation) demodulation for computing the instantaneous frequency of an intrinsic mode function has been proposed by Huang and his coworkers. This method totally eschews the Hilbert transform and can compute the instantaneous frequency of an intrinsic mode function very accurately. However, this paper will show that the empirical FM part extracted through the empirical AM/FM demodulation may contain riding waves and is no longer an intrinsic mode function. That will make the instantaneous frequency nonsensical. To overcome this drawback, an improved method, called riding wave turnover-empirical AM/FM demodulation, is proposed in this paper. Experiments show very positive results.