Thesis (M.Sc.) - Cairo University - Faculty of Engineering - Department of Systems and Biomedical Engineering

One of the recent and most important topics in the field of medical diagnosis is pattern recognition which needs high speed with a high confidence level. Traditional digital pattern recognition techniques need immense computation and are relatively slow. On other hand Optical processing has obvious advantages since it is done in real time due to the inherent parallelism of processing. In this study, we have built an optical pattern recognition system using joint transform correlator (JTC) and fringe-adjusted joint transform correlator (FJTC) for real-time target recognition applications. We used the images of fingerprints to prove the feasibility of such systems in matching medical patterns. The input images of fingerprints were correlated directly with the reference fingerprints and a highly distinctive correlation signal was produced if reference and target image are similar. A negligible correlation signal was produced when the reference and target images are different

Issued also as CD