USER S PREFERENCE-BASED COLOR TRANSFORMATION USING INTERACTIVE GENETIC ALGORITHM

Abstract

In this paper, we propose a new color transform method in the image based on the user s preferences. To obtain general preferences for color, we first divided the images into two categories as favourable and unfavourable groups based on the users evaluation. From the categorized groups of images, we compute color histograms to construct general templates for 8 primary colors. To reflect the user s own preference in the color perception, we adopted an interactive genetic algorithm. In the interactive genetic algorithm, each individual consists of three chromosomes for H, S, and V and the fitness value is computed from the user s input, which is then used as criteria to iterate Genetic Algorithm for learning the user s preference. Finally, a personalized individual is generated. By transforming the colors in the image with the personalized individual, a re-colorred image is generated according to the user own preference. Survey results show that re-colorred images from our proposed personalized color templates are more satisfying than the original images.