Discussion on the Optimization Method of Modern Ink Figure Painting Composition Structure Based on Geometric Algorithm

Abstract

Composition is an important factor in determining the aesthetic quality of modern ink figure painting, and the existing optimization algorithms in this field still have problems such as insufficient accuracy and rotational consistency, this paper proposes a composition optimization algorithm based on geometric algorithm. First of all, the color variety of the subject area and the color variety of the background area are chunked, and the color harmony feature is extracted by partition. Then, using the rule of thirds and the principle of visual balance as the principle, a deep convolutional neural network with VGG-16 as the model backbone is used to detect the saliency of the modern ink figure painting. In this paper, the accuracy of the composition structure optimization enrichment of modern ink figure paintings stays above 75%, and the attention and sensitivity to the three influencing factors of composition, ink color and texture, which are more important in the composition of modern ink figure painting style and aesthetics, are better, which not only provides a guarantee for maintaining the original painting style and aesthetics of the drafts after the optimization of the composition structure of the modern ink figure paintings, but also provides a theoretical basis for the modernization and transformation of the traditional paintings.