A robust indexing and retrieval method for lossy compressed color images

Recent years brought extensive growth of multimedia resources publicly accessible via world wide web. Therefore there is much labor dedicated to create a content-based image retrieval systems capable to efficiently index, retrieve and manage large scale databases. In this paper we propose a color indexing method based on Gaussian mixture model of color histograms. The model parameters serve as signatures providing fast and efficient color image retrieval. In this paper we present results showing that proposed approach is robust to color image distortions introduced by lossy compression artifacts. Therefore the proposed method is well suited for indexing and retrieval of Internet based collections of color images stored in lossy formats.