Color image enhancement method based on L0 gradient minimization

Abstract

The invention discloses a color image enhancement method based on L0 gradient minimization, which belongs to the color image processing field, and solves the problems that according to an existing enhancement method, the illumination of an image is not natural, when the scale selection is not appropriate, the edge-preserving smoothing of the illumination image is insufficient, a halo phenomenon occurs or the texture information and the structure information in the image cannot be well distinguished, so that the estimated illumination image is inaccurate, and the detail loss of the finally enhanced image is caused. The method comprises the following steps of converting a color image from an RGB space to an HSV space, and extracting an intensity component V as a grayscale image; constructing an optimal function of the L0 norm constraint sparse item, and inputting the gray level image into the optimal function; solving an optimal solution of an optimal function by using a semi-quadratic term separation method, and taking the optimal solution as an estimated illumination image; obtaining a reflection image based on the grayscale image and the illumination image; and carrying out reconstruction by using the illumination image and the reflection image, and converting the reconstructed illumination image and reflection image back to the RGB space to obtain a final enhancement result. The method is used for color image enhancement.

Description

technical field [0001] A color image enhancement method based on L0 gradient minimization is used for color image enhancement and belongs to the field of color image processing. Background technique [0002] In the process of image acquisition, due to the limitations of the image acquisition equipment and the influence of the ambient light in the acquisition environment, the acquired images often have some low image quality, such as underexposure and overexposure, noise pollution, object out of focus and color light confusing question. These problems are very unfavorable for human eye observation and subsequent image processing. In order to solve the problem of low image quality of the image, it is necessary to enhance the image. [0003] Traditional image processing methods, such as gamma transformation, filter mask and histogram equalization, are very simple to implement and have a good enhancement effect on most grayscale images. However, for color images, these method...

AI Technical Summary

Problems solved by technology

[0005] In view of the problems of the above research, the purpose of the present invention is to provide a color image enhancement method based on the minimization of the L0 gradient, which solves the enhancement method in the prior art, which will cause unnatural illumination of the image, and when the scale is not selected properly, the illuminance will be reduced. The edge-preserving smoothing of the image is insufficient, resulting in halo phenomenon or the inability to distinguish the texture information and structural information in the image well, resulting in inaccurate estimated illuminance images, resulting in the loss of details in the final enhanced image

Images