Improvement of Threshold Function Based on Wavelet De-noising

摘要

Hard threshold function is discontinuity in the threshold point, and soft threshold function have a constant error between the estimated coefficient of wavelet and the original coefficient of wavelet. To solve this problem, an improved threshold function is presented, which is continuous and high-order-differential in the threshold point. The experimental results show that the de-noising effect of this function is better than the soft threshold function, the hard threshold function and the modular square function.