Content Based Image Search over the World Wide Web

Most web pages typically contain both images and text. However, most current search engines index documents based on text only. In order to facilitate effective search for images on the web, we need to complement text with the visual content of the images. We often look for images containing specific objects having some particular spatial and topological relations among them. In this paper, we describe a system which enables the user to effectively search for images using the image content information including color, component objects and their relations in addition to associated text.