Wavelet De-Noising Based on Genetic Adaptive Threshold Algorithm
Guang Wen Wu, Chang Ming Wang · Advanced materials research · 2014
The central step of de-noising in wavelet domain is the threshold selection. To get the best de-noising result, the threshold should be selected according to the noisy observation. This paper presents a genetic adaptive threshold method, which gets the optimum threshold in the sense of least MSE by using an estimation function of the signals’ MSE function. The simulation results with standard wavelet test signals shows that the operation speed of the proposed method four times the traditional continuous searching algorithm, moreover, the optimum threshold calculated is more accurate and reasonable.