Research On Image Inpainting Algorithm Based On D-S Evidence Theory And Bat Algorithm

With the rapid development of information technology,image inpainting is a an increasingly significant research direction in the field of digital image processing.The inpainting technique which is valuable in application is mainly used for three aspects including the repair of damaged photos,the removal of excess objects on the image and the protection of ancient artifacts.The image inpainting method is mainly composed of three categories:inpainting method based on partial differential equation,inpainting method based on texture synthesis and image inpainting method based on sparse representation.This paper describes the advantages and disadvantages of these three types of inpainting algorithms,analyzing the evaluation criteria of image inpainting quality,focusing on the image inpainting method based on texture synthesis,and focuses on the study of Criminisi algorithm,and three improvements aiming at the shortcomings have been put forward.(1)Aiming at the deficiency of the block effect in the Criminisi algorithm,an modified Criminisi repair algorithm is proposed.In order to fully consider the texture feature of the image,the calculation of the priority is improved,and the confidence degree and the data item are improved in the same way.And a new confident method is proposed to reduce the error accumulated in the process of updating the confidence due to the difference between the target block and the sample block.Experiments show that this method can maintain the overall continuity of the image and improve the quality of image inpainting.(2)The D-S evidence theory is introduced to update the calculation formula of the priority,and the optimal distance of the pixel is added to the matching criterion to select the optimal matching result,which can reduce the inpainting error.The effectiveness of the algorithm is verified by experiments.In the restoration of natural images and murals,the effect is more reasonable.(3)Based on the D-S evidence theory,this paper proposes it that adaptively determining the size of the matching template according to the gradient change,so that the template window can be selected flexibly,which can improve the quality and efficiency of the repair.In search of the best matching template and the corresponding improvement of the filling,the employment of intelligent element heuristic algorithm-bat algorithm makes the SSD functioning as a target function to search,improving the efficiency of the inpainting algorithm.Experiments show that,compared with the traditional inpainting algorithm,the improved algorithm proposed in this paper is more accurate and the time is also reduced.On the basis of sufficient theoretical analysis,this paper has carried on a lot of simulation experiments to different degree of damaged images,verified the feasibility of the improved inpainting algorithm,it can preserve the structure and keep the propagation of the texture at the same time and improved the repair efficiency.