Feature Integration, Multi-image Queries and Relevance Feedback in Image Retrieval
Iqbal Qasim, J.K. Aggarwal · 2003
In this paper, we explore the effect of feature integration, multi-image queries, and relevance feedback in enhancing the performance of an image retrieval system. Weighted integration of structure, color and texture features is studied. In addition, we propose a methodology of retrieval consisting of multiple query images, as opposed to the traditionallyused model of a single query image. Two different mechanism of relevance feedback are also proposed and analyzed. Integration of features and feedback significantly improves the performance of the retrieval system.