A Color Conversion Method Based on l0 Gradient Preservation

Abstract

The invention discloses a color conversion method based on L0 gradient preservation. The method first clusters the original image and the reference image according to the color features and extracts the main color to establish the color mapping relationship between the original image and the reference image, and then The original image is sampled and the mapping of different main colors is converted to the corresponding sample, and the sample editing is optimized by using the similarity between the sample and the main color and the similarity between samples. Finally, the gradient of the edge of different objects is considered to describe the details of the object. An optimization model based on L0 gradient maintenance is adopted to optimize the details and edges of the image. The invention can ensure that pixels with similar colors in the original image are still similar after editing, and can maintain image details and edge structures, even if the editing differences of different main colors are relatively large.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a color conversion method based on L0 gradient preservation. Background technique [0002] Color conversion is to modify the color of the original image to be edited according to the color of the reference image provided, so that the new image will be consistent with the main color of the reference image while maintaining the content of the original image. There are two main considerations: one is color mapping, how to accurately map the color of the reference image to the original image. The second is detail preservation, how to make the details of the newly generated image consistent with the original image. [0003] Various schemes have been proposed to make the color of the new image consistent with that of the reference image, but there are still some problems to be solved. One is that the effect of the new image depends too much on the selection of the re...

AI Technical Summary

Problems solved by technology

However, in the process of color conversion, various main colors are usually mapped in different ways, and their structure will no longer be maintained at the edge of two adjacent areas with different main colors, and the existing gradient Preserving constraints or implicit filtering models treat all pixels in a uniform way, so that the edges of the image may appear blurred or the details cannot be preserved much

Images