Research On Blind Restoration Algorithm Of A Single Motion Blurred Image

Image deblurring is a basic but challenging problem in computer vision and image processing.Blurred images are mainly caused by camera shake,object movement,and camera out of focus.In recent years,due to the popularity of portable camera equipment and the popularization of photographers,motion blurred images are more common due to camera shake,object motion,etc.motion blurred images due to camera shake,object movement,and other reasons are more common.At the same time,the blur kernel causing the image blur in most cases is unknown.Therefore,the research on the blind restoration of motion blur images under unknown blur kernels has certain application value and research value.According to the causes of motion blur,motion blur images can be divided into global and partial motion blur.Global motion blur refers to motion blur caused by camera shake(here refers specifically to the situation where the scene is stationary during camera shooting and only the camera shakes),and local motion blur refers to motion blur due to object movement(this point refers to,The camera is stationary,and there is only one moving object in the scene).This article studies these two kinds of motion blur separately.The main work includes the following two aspects:1.Aiming at the problems of dark channel prior deblurring method which are sensitive to noise and prone to ringing artifacts,a global motion blurred image blind restoration method based on the combination of dark channel and color image gradient is proposed.First,the color image gradient and dark channel prior are combined to build a motion blurred images blind restoration model;Then,the deblurring process is based on alternating iterations of image estimations and blurring kernel estimations.The clear image is obtained by L0 minimization method based on variable separation;Finally,the blurring kernel is calculated by the Least squares method.Experimental results show that the proposed method effectively estimates the blurring kernels.The fidelity of the deblurred images is better than other blind image deblurring methods.2.To deal with the serious visual artifacts caused by the consistent restoration algorithm on partial motion blurred image restoration using globally consistent blurred image restoration method,a two-stage blur region detection and restoration method based on matting is proposed.First,Combining the blur features of frequency domain and space domain,a blur feature criterion is constructed to guide the coarse localization of partial blurred region;Then,the refinement detection of blurred region is realized by improving the KNN Matting algorithm;Finally,based on the restoration of the blurred areas,the final image is obtained by fusing the restoration results and the clear areas.Experimental results show that the proposed algorithm can detect the partial blurred region quickly and effectively,and it has outperformed other traditional methods in the application of blur region detection and image restoration as well.