Joint use of luminance and color invariants in partial image retrieval

Object color plays an important role in object image matching. Color invariants are robust features under varying illumination conditions. In this work, we analyze the estimation process of the color invariants by Geusebroek et al. from RGB images, and propose a novel invariant feature H´ based on the basic invariants to meet the circular continuity residing in the mapping between colors and their invariants and improve the color selectivity. The proposed invariant H´ in combination with luminance, contributes to enhance the retrieval performances of partial image matching under varying illumination conditions.