Construction and Efficacy Verification of Color Theory Optimization Model in AI Painting Assistant

Abstract

Abstract The application of artificial intelligence in the field of painting has been greatly developed, which is a new way of thinking in the digital era, bringing innovation and infinite possibilities to the art of painting. The article first researches the art and expression of AI painting, mainly taking various network models as the main form of expression. Then, color theory and color space are introduced to provide a theoretical basis for the proposal of a color optimization model. Considering that the ACE algorithm has problems of high complexity and image enlargement, the acceleration algorithm LLLUT is used to improve data processing. The AI imitation painting works “Monet’s Garden” and “Neon” are taken as case studies. In terms of color preference, AI painting prefers to use three primary colors: black, white, and red. For example, AI’s imitation painting work Han Xizai Night Banquet has three peak points higher than the average peak, respectively 0.04, 0.09, 0.15, and 0.99. The corresponding colors used are red, black, and yellow. The main colors used in the composition are red and yellow. In the application evaluation of the color model, the comparative analysis concludes that the color mixing efficiency is higher using the research method in this paper, and in the practical application, both painting professionals and amateurs have a stable effectiveness in generating color preference schemes after using the AI painting assistant based on the ACE color model optimized by LLLUT.