A color blindness enhancement method for remote sensing images based on palette

Abstract

The application discloses a remote sensing image color blindness enhancement method based on a palette, relates to the fields of remote sensing image processing and color science, and comprises the following steps: S1, reading a remote sensing image and converting the remote sensing image to a perceptually uniform color space, acquiring Lab pixel data of the image, and extracting a combined palette representing color distribution of the image; S2, calculating a contribution weight matrix of each pixel to each color in the combined palette based on the combined palette, color and spatial position relationship of the image pixels; S3, using a color vision disorder simulation model, taking the color distinguishability under multiple types of color vision disorders as a target, and optimizing the combined palette under color similarity constraints to obtain an optimized palette; and S4, using the weight matrix, remapping the color of the optimized palette to each pixel to generate an enhanced remote sensing image friendly to color blindness, and realizing accurate and efficient enhancement of the color of the remote sensing image.

Images