An Improved Approach to Threshold Function De-noising of Mobile Image in CL2 Multi-wavelet Transform Domain

A new threshold function is presented based on the traditional de-noising method put forward by Donoho. The new threshold function has many advantages over Donoho´s soft-threshold and hard-threshold functions. This new algorithm is in possession of not only the de-noising ability of the energy-adaptation, but also the capability of protecting edge information of mobile image signal when it was applied to de-noise. It is simple in expression and avoids the discontinuity of hard-threshold function; meanwhile, it is as continuous as the Donoho´s soft-threshold function. Besides, this function has a high order derivative which makes all kinds of mathematical manipulations very convenient. At the same time, the new threshold function is more elastic than the traditional soft-threshold and hard-threshold functions. All these advantages make it possible to construct a self-adaptation de-noising algorithm. Simulation results based on both single wavelet and CL2 multi-wavelet transformation indicate that the de-noising approach with the new threshold function outperforms the traditional soft-threshold function in both visional effects and signal-to-noise ratio