Refining Color Scheme Generation: Iterative K-Means Clustering and ARI Evaluation

Abstract

Color goes beyond mere visual sensation, holding profound sway over emotions, thoughts, and perceptions. It communicates, evokes moods, and significantly influences judgments. Research underscores its importance, with up to 90% of product assessments being based solely on color, highlighting its pivotal role in crafting memorable experiences and defining brand identities. The fusion of art and technology presents a captivating synergy within the realm of image-derived color schemes. Color palette generation from images is pivotal in graphic design, interior decoration, and digital media. This study delves into methodologies for extracting dominant colors from images and generating cohesive color schemes. Leveraging K-Means clustering with the Within-Cluster Sum of Squares (WCSS) method, we showcase superior performance compared to traditional approaches. The evaluation of palette coherence using the Adjusted Rand Index (ARI) facilitates consistency within the generated color schemes. Integrating methodologies with design tools and advanced color harmonies opens avenues for further innovation and customization. This study underscores the transformative potential of image-based color scheme generation, bridging the gap between computational analysis and creative expression. Through the convergence of artistry and technological prowess, we aim to enhance the design landscape and enrich user experiences across various applications and industries.