Noise Reduction, Adaptive Threshold-based Wavelet Image

Denoising is an important step in the processing of image.In the course of collection and transmission,images are frequently affected by noises,the aim of denoising is to retain the useful information of original signal and remove the noise at the same time.In recent years,there was a great development in the domain of wavelet.Because of good characteristics of local in time and frequent domain,wavelet can give prominence to any details of researched objects.Many advantages make wavelet be widely used in image denosing field.Firstly,we introduce the development of wavelet and its application situation in image processing.And then,some traditional methods which are based on wavelet are described in detail,and these algorithms are separately compared whose disadvantages and advantages are detailed analyzed.Three wavelet denoising methods are proposed in this paper which is on the basis of deeply research on wavelet theory and traditional denoising algorithms.They are improved bayesian image denoising based on wavelets,image denoising based on lifting wavelet and subbands adaptive threshold,wavelet image denoising base on local threshold.The first method based on the bayesian denoising method focuses on the problem of estimation of paremeters,and this algorithm makes the improved model of estimation more stable and precise.The second method based on the universal threshold proposes the adaptive threshold.This algorithm fully uses the relativity between subbands and combines the advantage of lower calculation of lifting wavelet to denoise image.While the third method sufficiently makes use of the local information of image by splitting the subbands and calculating the threshold of each block,meanwhile,this algorithm possesses adaptivity.Experimental results show that the three methods proposed in this paper can obtain better denoising performance and pleased visul effect.