Multi-features description for an efficient image retrieval

This paper describes an multi-features indexing system for use in Content Based Image Retrieval. The standard CBIR approach is simple and usually use a single information such as shape, scale or color, which leads to recognition problems in some cases, to remedy this problematic, we use additional features to combine types of information. From many points of view local descriptors are relatively different, the purpose of this paper is to have an overall view over some existing features descriptors and calculate the similarities between a query image and video frames in the database from different points of view.