Color Quantization Application Based on K-Means in Remote Sensing Image Processing

Abstract

In this paper, four kinds of RGB remote sensing images are processed using color quantization algorithm based on K-Means to reduce the number of colors in the image. The color quantization algorithm is to select the most representative color and reduce the useless color in the image as much as possible. This paper assumes that the RGB remote sensing image is composed of multiple pixels. Using the K-Means algorithm to perform unsupervised clustering on these pixels with specific colors, color quantization can be realized. The use of K-Means for color quantization of remote sensing images can reduce the number of colors in those images, so that remote sensing images can be reproduced well in lower performance computer equipment. At the same time, color quantization reduces the size of remote sensing images and improves the efficiency of remote sensing image processing.