A color palette reduction method using multiobjective evolutionary clustering algorithm

When we display color images on low devices having a limited number of pixels and/or colors, the technique called “color palette reduction” which reduces the colors in color palette to represent the input image with a limited number of colors is often used. In the color reduction process, we should select representative colors by considering color distribution of the input image so that the error between the approximated image and the original one becomes minimum. In this work, we try to use a new clustering approach called MOCK [1], which uses multi-objective evolutionary algorithm as the mean of clustering, and verify the basic performance of the proposed approach through computer simulation using several benchmark color images.