Uncovering historical ceramic color patterns via visual analytics for design and heritage

Traditional research on ceramic colors relies heavily on qualitative descriptions and subjective judgments. This approach limits the ability to process large-scale cultural heritage datasets and extracts reproducible design knowledge difficult. To bridge this gap, this paper introduces an interactive visual analytics system that merges design principles with computational innovation. Employing computer vision and network science, the system transforms unstructured image data into structured color networks. Key techniques include color quantization for hue extraction, the construction of “ceramic-color” bipartite networks to model co-occurrence patterns, and centrality analysis to identify core color hierarchies. This approach allows researchers to visually explore color associations with cultural attributes (such as dynasties and vessel types) across large collections. Furthermore, it empowers designers to accurately extract historical color schemes for contemporary application. Ultimately, this work establishes a data-driven methodology that effectively integrates digital humanities analytics with modern design practices.