A Sample-Based Fast Image Restoration Method

Abstract

The invention discloses a sample-based rapid image restoration method, which comprises: using marks to determine the area to be repaired; selecting the contour line at the junction of the area to be repaired and the known area; calculating the data value and confidence value of the template; introducing an adjustment parameters to determine the module to be repaired with the highest priority; in the known area adjacent to the module to be repaired, according to the distance from the center point of the module to be repaired, the optimal matching block is searched from near to far; the optimal matching The pixels corresponding to the block are filled to the corresponding position of the module to be repaired, and the confidence value of the newly filled pixel is updated at the same time; the above steps are repeated for the area to be repaired until all the areas to be repaired are filled. The invention optimizes from four aspects of priority calculation, data value calculation, matching area and optimal matching block search and confidence value update, the repair effect is more natural, and at the same time, the repair is only locally matched, which greatly reduces the repair time .

Description

technical field [0001] The invention relates to an image restoration method, in particular to a sample-based rapid image restoration method, and belongs to the technical field of image processing. Background technique [0002] As an important application of image processing, the main purpose of image restoration is to restore damaged images to a degree acceptable to the human eye. Its application fields include art restoration, computer animation, film and television special effects, virtual reality, etc. . [0003] The current image repair algorithms mainly focus on two categories: one is the method of image retouching (inpainting), which was first introduced into digital images by Bertalmio et al., using a repair model based on partial differential equations (PDE). There are also the total variational (TV) model proposed by Chan and the curvature-driven diffusion (CDD) model. These methods have a good repair effect on damaged images with flaws, but the effect of repairin...

AI Technical Summary

Problems solved by technology

[0006] Aiming at the problems existing in the above-mentioned prior art, the present invention provides a sample-based fast image restoration method, which can effectively solve the problem of Criminisi algorithm low-texture region repair lag, and can also ensure that the boundary lines of texture regions of different colors will follow the filling process. It is preserved and continued, and the linear structure of the image is preserved, which is conducive to the elimination of "faults", so that the sequence of restoration and synthesis is basically synthesized from the periphery to the middle, and the final restoration effect is more realistic and natural; Reduce repair time

Images