Wavelet threshold de-noising based on higher-order statistics in attenuating random noise
Zhaocai Wu, Tianyou Liu, Caihong Hua · 2006
Wavelet threshold de-noising is a useful method for reducing random noise. However, it becomes more difficult to select a suited threshold with SNR of signal decreasing. We used a higher-order correlation method for random noise elimination. In this approach, we applied a third-order correlation technique for identification of wavelet coefficients uncorrupted by noise by calculating tripe correlation coefficients of wavelet-signal correlations. Because the higher than second-order moment of the Gaussian probability function is zero, tripe correlation coefficients does better than wavelet coefficients in reducing random noise. Model and real data processing result show that can extract weak reflected signal effectively and improve SNR of signal.