Evaluating the performance of content-based image retrieval systems

Abstract

Content-based image retrieval (CBIR) is a new but in recent years widely-adopted method for finding images from vast and unannotated image databases. CBIR is a technique for querying images on the basis of automatically-derived features such as color, texture, and shape directly from the visual content of images. For the development of effective image retrieval applications, one of the most urgent issues is to have widely-accepted performance assessment methods for different features and approaches. In this paper, we present methods for evaluating the retrieval performance of different features and existing CBIR systems. In addition, we present a set of retrieval performance experiments carried out with an experimental image retrieval system and a large database of images from a widely-available commercial image collection.