Image Denoising Using Adaptive Wavelet Packet Thresholding Function

This paper proposes a optimum adaptive wavelet packet(WP) thresh-olding function for image denoising based on the generalized Gaussiandistribution.It applies computationally efcient multilevel WP decompo-sition to noisy images to obtain the best tree or optimal wavelet basisusing Shannon entropy.It selects an adaptive threshold value which isparameters of subband coefcients.In the utilized thresholding functionwhich is based on a maximum a posteriori estimated,the modifed versionof dominant coefcients was estimated by optimal linear interpolationbetween each coefcient and the mean value of the corresponding sub-band.Experimental results,on several test images under diferent noiseintensity conditions,show that the proposed algorithm yields better peaksignal noise ratio and superior visual image quality.