Region-Based Automated Relevance Feedback in Algae Image Retrieval
Hassan Tabout, Ali Souissi, Abderrahmane Sbihi · 2007
In this paper, we present a content based image retrieval system with automatic relevance feedback. The proposed technique is based on color interest points extraction and on the calculation of the local descriptors in interest regions. Global and local features are grouped to feed a two level Self Organizing Maps Network in order to discriminate between retrieved images without user interaction. An implemented prototype system has demonstrated a promising retrieval performance for a test database containing more than 1000 color algae images. Key words: