Research On Image Restoration Algorithm In Wavelet Domain

Due to the constraints of technical conditions and noises, we can only obtained some degraded images. The destination of image restoration is to recover image that has been degraded and make sure that the processed image as near as possible to the original image. Because the process of image degradation is always acting as low-pass filtering, image restoration is a typically ill-conditioned problem. Wavelet bases have flexible time-frequency dilation capability, which can characterize the image with edges and singularities very effectively. Nowadays, digital image restoration research based on wavelet analysis is a hot spot.Based on the in-depth systematic study of various image restoration methods, analysis of the essential attribute and the advantages and disadvantages of theirs, Papers brings forward image restoration methods that Fourier domain is with the combination of wavelet domain:Wavelet-domain algorithm is good to recover the edge but is so complicated to compute. Frequency-domain algorithm is simple but can't recover the edge. So two algorithms are complementary. If we restore images only in Frequency-domain or Wavelet-domain, we can't combine the effectiveness and efficiency of recovery; so this papers provide the new ideas for the restoration of degraded images, which is the restoration algorithm of Wavelet-domain and Frequency-domain. People can use an advanced Wavelet-domain regularized restoration method, which is the image restoration algorithm about Regularized inversion and Wavelet-domain threshold shrink. Experimental results show that the algorithms proposed are effective.Using Bivariate model is due to Fourier-Wavelet regularized deconvolution for ill-conditioned systems(ForWaRD) to restore images. With the help of the ForWaRD algorithm, make fuzzy and degraded images Wiener filter regularized deconvolution in Frequency-domain, then deal with the denoising. If you want to get a good effect, you can use a model whose natural image wavelet coefficient vector is Bivariate probability distribution function.