Digital image color analysis method to extract fashion color semantics from artworks

Color in artworks is an inspirational source for fashion designers. Product color is a fundamental design element to stimulate customers' emotional response as designers intend or express conceptual trends they want to infuse into products. Thus, artwork color analysis and the color selection process are crucial for product designs. This study aims to establish digital color analysis whereby fashion designers can extract colors from digital images of artworks and find relevant fashion codes with automation. The analysis method provides the image color attributes in Munsell, Practical Color Coordinate System, and Shigenobu Kobayashi Color Scale. It reveals underlying color semiotics representative of digital images. That is essential to provide color guidance to fashion experts when planning product colors. The semiotics analysis is based on Kobayashi’s 3-color combination scale and quantitatively estimates the similarity of an artwork image to the 3-color bar. Symbolic adjective of the image is found whereby image coordinates in semantic space can be estimated. This method was used to investigate the colors of 190 Monet's paintings. The results found that Monet frequently expressed subject matters in nature with colors of low saturation tone. We explore the manifestation of the painting colors in dress collections as fashion codes. This approach found that an inspiring painting and a designed dress impress the same or similar fashion codes.