Design And Implementation Of Image Restoration System Based

Artists had begun to manually repair paintings which had been damaged since Renaissance. With the development of digital computer technique, people expected and started using electronic technology to repair damaged images, thus the “digital image inpainting” emerged.Image inpainting algorithm based on correspondence maps is a texture based image inpainting algorithm. The essential part of the algorithm is to find out an optimal solution of maps between pixels in the damaged areas and pixels in undamaged areas. The algorithm begins with random mappings, then calculating and comparing the Euclidean distance between pixels through randomly neighborhood matching. My software implements and optimizes the algorithm to make it fully consider the enviroment of each pixel. The evolved algorithm aims at finding optimal mapping based on the enviroment, and enhancing the efficiency by scan conversion methods.Based on consistent mapping of image patches is one constituency and automatic repair system contains. The system allows users the freedom to choose the selected image area to be repaired, and erasing selected area, directly select a rectangular area and the like. After scanning the storage system selection, run inpainting, and display the results in a short time, the results are saved automatically.User client also needs to build a fully functional application software, including userfriendly interface to perform function selection, picture and drag to view the results as a basis to provide a good user experience. In the arithmetic processing section, the results of the system should each iteration algorithm are displayed on the system,allowing users to run the algorithm has an intuitive understanding. And it allows the user to control the number of iterations, and the results are automatically saved to a file. This article results of the project lies in the innovative image processing algorithms based on consistent mapping of the introduction of the concept of neighborhood levels and randomly selected points, and the use of sub-block to take point at random to take points to the restrictive effect at random. Scanning pixel, inorder to deal with the situation of irregular regions, to take a two-way scanning method, flexibility and fast was realized from the boundary scan order caused by an internal algorithm requirements and defines the area of the storage formats of the scan. In the area selected, using a selection and clearing operations and inter-regional and non-regional operations to achieve a high degree of freedom and precision. It provides statistical results algorithm running in terms of image restoration, giving the user a visual reference operating results. Overall software from interaction with the user to the image restoration algorithm performance are good, very good to complete some of the features of an image repair software should, instead of doing so manually repair work picture, and it can be completed within a short period of time a job.The image inpainting system takes the image inpainting algorithm based on correspondence maps as the core, and it’s fully functional and easy to use. The software provides the user interface to select areas to be repaired, so that the user can freely select areas to be repaired, and automatically fix these areas, to achieve the purpose of image inpainting through electronic meathods. The system has characteristics of convenient operation, high efficiency and effect running result,which is a convenient and effective image inpainting software.