Research On Restoration Algorithm Of Motion And Defocus Hybrid Fuzzy Image

Image blurring problem exists widely in all aspects of social life, is one of the important tasks of image processing. Motion blur and defocus blur are the most common blurring process, which has become a hot research topic in the field of image processing for some time. For motion blur and defocus blur, a lot of literatures gives the corresponding algorithms, and achieve different degrees of effect. These algorithms have been applied to the public security detection, license plate recognition, computer vision, UAV visual navigation and other fields.However, in real life, many cases of two kinds of blurring exist at the same time. For example, images captured on a moving vehicle, hybrid blur can be produced, and the ground scene captured by a fast flying drone is sometimes mixed with a blurring image. At present, the research on hybrid blurring is not too much, the literature can not be found. Therefore, it is of great significance to solve the problem of the restoration of mixed blurring images. Aiming at this problem, this paper focuses on the algorithm of hybrid blurring image restoration. In this paper, based on the former two kinds of blurring image restoration algorithm, the complexity of the algorithm process, the preparation of a large number of algorithm procedures, the twenty-three groups of experiments. A certain effect has been achieved in the restoration of the mixed blurring image. From the visual effect, which can meet the human can recognize the image content. From a technical point of view, can achieve a better level.The restoration of hybrid blurring image is the key point of this paper. In this part, twelve groups of experiments are carried out. Better recovery effect was obtained. For a given mixed blurring image, the "least square regularization model" is used to remove the blur of the mixed blurring image. The experimental results show that the R-L iterative algorithm has better effect on motion blur. At the same time also can achieve the best effect of mixed blurring.Under different noise levels, first carried out motion blurred, the defocus fuzzy processing later recovery, first remove defocus blur restoration process, get the image PSNR value than the recovery image which is carried out removing motion blurred image PSNR value average high 0.9342dB; Under different noise levels, first carried out defocus blurred, the motion blur processing later recovery, First remove motion blur process, get the image PSNR value than the recovery image which is first carried out removing defocus blurred image PSNR value average high 0.1485dBThe innovation points of this paper:This paper has drawn up some technical routes (algorithm combination). Experiments were carried out on the restoration order of the mixed blurring images with different blurring order. Through the spectrum analysis combined with experimental results, found that when recovery sequence and blurring in reverse order, the better effect of image restoration. Through the analysis of the spectrum and the experimental results, it is found that the recovery effect is better when the restoration order is opposite to the order of blurring. In the process of restoration, the relation between the sequence of the process and the blurring order is found. In order to provide a theoretical and experimental reference for further recovery of hybrid blurring images, the effect of blurring order and restoration order on image restoration is studied.