iPURE: perceptual and user-friendly retrieval of images

Content-based image retrieval (CBIR) systems built around mathematical similarity models are often limited in their ability to capture user perception. On the other hand, relevance feedback approaches to CBIR in client-server environments incur a significant overhead in database search and image download time. A novel methodology for understanding user perception through a set of interactive tools is presented in this paper. The system uses a query by example methodology and uses color, texture and shape parameters for perceptual matching. It provides an object-level view of the query image through image segmentation. Traditional relevance feedback techniques on the searched results are supplemented with intra-query learning where the user perception is learnt through feedback on a set of images automatically generated from the query image itself, thus, alleviating the overheads of database search and image download. In addition, we allow the user to explicitly redefine the query point by manually modifying the query image. We demonstrate the feasibility and advantages of the approach with examples