Research And Design Of Image Recovery Algorithm

Image restoration technology is one of the key research topics in the field of image processing and is widely used in various fields.There are many factors that cause image blur during image acquisition and transmission,such as the relative motion between camera and object,atmospheric turbulence,noise and camera focus and so on.These factors will cause the image blur and distortion.The main purpose of image restoration is to filter the blurred image,which can not only filter out the noise but also preserve the integrity of the image.In this paper,we mainly study the Lucy-Richardson(LR)restoration algorithm.The LR algorithm assumes that the original image satisfies the Poisson distribution and the maximum likelihood of the output image reaches the ideal image through iteration.The LR algorithm has strong stability and can recover the blurred image well under the condition of uncertain noise.LR algorithm in the image restoration process,the iteration will make the noise is amplified,and thus the restored image ringing effect,while the number of iterations caused by the time-consuming.In this paper,two methods are proposed to solve the ringing effect and computational efficiency respectively.In order to solve the ringing effect,a Gaussian filter is proposed to add the classical LR algorithm to the smoothing area and the modified LR iterative formula at the edge,which ensures the image The integrity,but also inhibited the ringing phenomenon.Aiming at the problem of long computation time,this paper proposes a joint algorithm based on wavelet transform.Using the absolution feature of wavelet transform,the image is decomposed by wavelet.The low-frequency part obtained after decomposition mainly focuses on the effective information of the image.According to the principle of wavelet decomposition,the image after wavelet decomposition becomes smaller,the data involved in the processing using the restoration algorithm will be less,and the recovery efficiency of the blurred image will be improved.In this paper,the fuzzy image caused by the movement is taken as the experimental object,and the improved LR restoration algorithm is used to simulate the Matlab platform.The experimental results show that the improved LR algorithm not only has a good visual recovery effect,but also suppresses the ringing effect.At the same time,a better restoration effect can be obtained for a complex background image in a short computation time.