Image inpainting method based on isoluxline

Abstract

The invention discloses an image inpainting method based on isoluxline. The image inpainting method based on the isoluxline comprises the following steps of improving a traditional isoluxline function through an edge-stopping function of an anisotropic propogation arithmetic and combining the improved isoluxline function with a traditional image inpainting priority function to obtain a stable priority calculation function; obtaining a priority value of a broken pixel block in a structural image through the priority calculation function; seeking a sample block with the highest similarity, namely the best-match domain block, within the scope of a whole image and completing the inpainting of the broken area in the structural image and completing the inpainting of the broken are in the structural image; inpainting a broken texture detail image by a texture detail sample block, corresponding to the best-match domain block in the structural image, in a texture detail image; superimposing the structural image completed the inpainting and the texture detail image to complete an image inpainting process. The image inpainting method based on the isoluxline effectively maintains the structural information of an image need the inpainting and obtains relatively ideal inpainting effect of a texture with illumination changes simultaneously.

Description

technical field [0001] The invention relates to the field of computer image processing, in particular to an image restoration method based on isoluminance lines. Background technique [0002] As an important part of digital image processing, the purpose of digital image inpainting is to repair the information defect area according to the known data information in the image. In recent years, with the wide application of multimedia technology in the fields of image restoration and image editing, image restoration technology has received more and more attention. At present, the technology can be roughly divided into two categories: one is the restoration method based on diffusion (Diffusion-based), and the other is the restoration method based on sample blocks (Exemplar-based). [0003] The Diffusion-based repair method is based on a parametric model or a partial differential equation, gradually smoothing the transition from the edge of the image damaged area to the inside, an...

AI Technical Summary

Problems solved by technology

[0006] The invention provides an image repair method based on isoluminance lines, which solves the error matching phenomenon that is easy to occur in the traditional Exemplar-based repair algorithm and its improved algorithm, as well as the resulting structure breakage or texture illumination distribution after repair. For common problems, see the description below:

Images