Implementation of an Efficient Mechanism for Improving Web Image Search based on User Intention.

Abstract-In this paper we implemented a novel web Image search method, which will overcome the drawbacks of text based retrieval system and some of the drawbacks of content based image retrieval system. In this system we have captured the user intention and then system retrieves those images which are intended to user and even we have implemented a detection of duplicate and repeated images function. The main drawback of content based retrieval is though the system forms a cluster of images according to similarity function but its result set contains bunch of duplicate images so we have focused on to overcome this drawback and to detect duplication of images and avoid them so that user can get quality image retrieval and also user can be satisfied with results. In this system we combined the information of text based query as well as visual information of image. In this system we have maintained database of Images which are categorised by some predefined properties of system lik