Dynamic closest color warping to sort and compare palettes

A color palette is one of the simplest and most intuitive descriptors that can be extracted from images or videos. This paper proposes a method to assess similarity between palettes by sorting colors. While previous measures compare only colors without considering overall combination, we sort minimize geometric distance align them share common tendency. We propose dynamic closest warping (DCCW) calculate minimum sum graph connecting in other palette. evaluate proposed DCCW with several datasets demonstrate outperforms methods terms accuracy computing time. validate effectiveness technique conducting perceptual study, which indicates clear preference for results our approach. also useful applications enabled DCCW, including interpolation, navigation, image recoloring.