Research on Children's Illustration Color Matching Perception Based on Clustering Algorithm and Color Network

ABSTRACT Color expression in children's illustrations plays a vital role in the display. Children can enjoy the colors and experience the emotion conveyed by the illustration theme through their perception of color. Based on the K‐means clustering algorithm and color network, this paper analyzed color clustering of children's illustration samples with the theme of “Spring” and constructed a color network model for these samples. Using the semantic difference method and Likert scale, a semantic division questionnaire was developed to derive representative vocabulary for the illustration samples. The color systems obtained from clustering were then matched with the representative vocabulary to form a color perceptual analysis questionnaire. Through factor analysis, typical vocabularies were identified, and corresponding color systems were created as experimental illustration samples. This paper explores the expressive forms of color systems under different vocabularies and verifies the feasibility and reliability of the experimental illustration samples within the target group, thereby providing new directions and references for color design and color education in children's illustrations.