The Control Regularization-term Iterative Algorithm For Image Deconvolution

Digital images have been playing an important role in many fields such as medical science,aerospace and biology etc.However,object movement,defocus and other reasons lead to image blurring when digital images are acquired.Blurred images reduce the precision of segmentation,registration and pattern recognition and so on,so image deblurring turn into be an important step in image processing.In this paper,we firstly summarize the development of image nonblind deconvolution algorithm.We next analyze the advantages and disadvantages of several main non-blind deconvolution algorithms in detail.Finally,an image non-blind deconvolution algorithm is proposed which based on Weighted Difference of Anisotropic and Isotropic Total Variation Algorithm(DCA),called Control Regularization-term Iterative Algorithm(CRI).We observe that Weighted Difference of Anisotropic and Isotropic Total Variation algorithm has an overminimizing regularization term effect in the sense that it gives a solution whose regularization term is usually smaller than that of the original image.This effect is due to the pre-pondering role of the regularization term in the corresponding minimization problem and prevents from finding the exact solution of the deconvolution problem when such a solution exists.We propose a modified version of the Bregman iteration,which leads to an exact solution of the deconvolution problem if it exists and to a satisfactory approximative solution if there is no exact one.The idea is to control a varying parameter of the regularization term in the Bregman iteration.The proposed method has the advantage that the restored image has the regularization term closer to that of the original image,compared to the classical Bregman iteration.