A new approach to image retrieval using multi-level generalized finite automata

A new approach to image retrieval is proposed in this paper. This approach is based on representing images by their multi-level generalized finite automata. The proposed method provides low storage overheads, overcomes the effect of partial image alterations, offers a flexible setting for retrieval accuracy and being inherently suitable for parallel processing. Experimental results which demonstrate the performance of the proposed approach are presented.