On Suppressing Cross-terms In WVD Via Thresholding Superimposition Of Multiple Spectrograms

The Wigner-Ville distribution(WVD) deserves universal attention as a method to analysis unstable and time-varying signals in the last decade. It has the properties of high time-frequency resolution and energy concentration as well as tracking instantaneous frequency etc.The WVD can be exploited in signal checking or analyzing, texture signal processing and image recognition for its some designed properties.The Wigner-Ville distribution (WVD) provides superb time-frequency resolution for many signals and desired mathematical properties. However, an obvious drawback is that the severe interfering cross-terms influence on the effective analysis, interpretation, and information extraction of multicomponent signals. It is one of important areas in the time-frequency analysis to suppress or reduce the cross-terms in WVD's. In this paper, a novel method is proposed to suppress the cross-terms, which is based on thresholding the sum of multiple spectrograms. First, the spectrograms with different time-frequency resolutions are superimposed and then the superimposition is thresholded to localize the autoterm support region of the WVD. Secondly, the WVD is multiplied by the indication function of the region. In this way, a novel time-frequency distribution (TDF) is obtained. Unlike the traditional kernal function methods, this novel method not only reduces the interfering cross-terms, but also preserves the superb time-frequency resolution of the WVD. The experimental results show that the method is very effective for both the multicomponent signals consisted of the LMF's and nonlinear modulation frequency chirp signals.