Object-based queries using color points of interest

Abstract

Image retrieval systems based on the query-by-example paradigm usually imply a global description of the image. An approach involving points of interest has been developed and comes to light with object/sub-image retrieval tasks which require more local descriptors. However, this solution is not optimal, since it is limited to gray value images. We propose a local image descriptor based on color points of interest, i.e. points extracted and characterized from a color signal at one go. This local characterization is then involved in an image indexing and retrieval approach which is dedicated to object or sub-image retrieval. The entire approach is evaluated on two image databases with various contents and some particular results of object retrieval show the relevance of our approach.