Modular optical systems for nonlinear data processing and manipulation

The paper reviews the usage of modular optical systems for nonlinear data processing and manipulation. The paper starts with presenting an algorithm for multi stage image processing in which a derivation of instillation having N stages of binary filters is used in order to generate general transformation and then to apply it for pattern recognition. Then, optical configuration that implements second order of Volterra operator is used for non-linear control processes. In the third part of the paper a fuzzy logic based algorithm is used in order to optimize the design of multi stage interconnection network (MIN). The optimization is done in a sense of minimal number of switching modules and minimal number of switching operations per routing stage. In the fourth part we present a technique for optical generation of fuzzy based rules. In the last part of the paper we suggest a fuzzy logic based technique for improving the depth of focus in an imaging system using as single-phase only filter. All the suggested pmcessing configurations are backed up with either simulations or experimental verifications.