An Interactive Image Recoloring Method Based on Optimal Model

Abstract

The invention discloses an interactive image recoloring method based on an optimization model. Obtain the globally optimized matting image and the locally optimized matting image, perform binarization on the obtained globally optimized matting image and the locally optimized matting image, and take addition or The strategy of subtraction merges binarized images, calculates the area that needs to be recolored in the color image to be processed, converts the RGB color space of the colored area to the HSV color space, and adjusts the hue channel and saturation channel in the colored area in an interactive way , converting the HSV color space of the image to the RGB color space, and finally combining the recolored image with the original color image to obtain the recolored image in nine steps.

Description

technical field [0001] The invention relates to a color image recoloring method, in particular to an interactive image recoloring method based on an optimization model. Background technique [0002] Image recoloring technology refers to the process of changing the original style of the image by changing the color of the image to achieve a certain color effect. Recoloring technology usually involves many image and video processing technologies such as image segmentation, clustering, and motion estimation, which can directly generate huge economic benefits. Recoloring technology is derived from practice and is directly oriented to applications. Therefore, the current recoloring technology Research is a challenging hot topic in the neighborhood of computer graphics and computer vision, and has very broad application prospects in many engineering neighborhoods such as film and television, animation, image, and science, industry and military. The traditional recoloring technolog...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is: provide a kind of effective recoloring method based on the optimization model of the interactive color image that only needs a small amount of lines, improve the following problems existing in the existing recoloring technology based on color propagation: (1) Artificially colored lines need to cover all the colors in the image, otherwise there will be color loss when recoloring; (2) the color texture information of the recolored area is reduced; (3) the algorithm parameters need to be adjusted to match the area that needs to be recolored, The method is not intuitive; (4) the global color propagation method lacks local control, and the local color propagation method requires more user input lines; (5) when the user input line color is inappropriate and needs to be adjusted, the entire algorithm needs to be recalculated and consumes plenty of time

Images