Social Sensing : Image Rank using Visual Words Module II

Abstract

The proposed content-based ranking algorithm for image retrieval uses the visual features of images (visual words) to retrievemorerelevant images. Hence, making the retrieval process moreaccurate than the tag-based Image Retrieval (TBIR). Initially, the system creates a visual vocabulary and trains a classifier on a dataset of 2,400 images belonging to different categories.Next, for any given userquery, the system makes a decision to display a class of images that best matches the query. These class images are then processed in a way that we compute the relevance scores for each image and display the result based on the score. Our content-based ranking algorithm is then integrated with the tag-based ranking algorithm (Module-I).Both tag and content-based image retrieval techniques have their own advantages and disadvantages. By integrating them together, some of their disadvantages can be overcome. The existing image search engines are either tag-based or content-based, but not both. Thus, a new system with these techniques integrated together is highly needed.