A Study of the Fundamental Role of Color Perception in the Design of Graphic Composition

In graphic composition design, the use of color can determine the overall feeling and communication effect of the work. This paper proposes a color image quality evaluation method based on the characteristics of color perception, so as to improve the application effect of color perception in graphic composition design. First, two images are randomly selected from the existing database to construct image pairs, and their preference class labels are estimated based on their corresponding subjective quality scores, which weakens the problem of learning image quality score prediction into a classification problem of quality preference learning. Then the three color features of color harmony, color contrast and color richness and the related luminance features of the images are extracted, the feature difference vectors of the image pairs are calculated, and the mapping problem from the feature difference vectors to the preference class labels is transformed into a binary classification problem for solving by using the machine learning classification algorithm. Finally, the TID2013 database is selected for image quality evaluation. The experimental results show that color perception can effectively improve the subjective and objective consistency of color distortion types. Applying the proposed method to the graphic composition design, the mean values of the works designed under the method are better than the ordinary works in all kinds of eye movement indexes, which indicates that color perception can improve the visual attractiveness of the works in the graphic composition design.