THE MASSIVE AMOUNT OF DATA PRESENT IN AN IMAGE PRESENTS PROBLEMS WHEN PROCESSING THE IMAGE USING A NEURAL NETWORK. SIGNIFICANT REDUCTIONS IN THE SIZE OF THE NETWORK CAN BE ACCOMPLISHED IF THE IMAGE IS PREPROCESSED USING AN OPTICAL FEATURE EXTRACTOR OR A DIGITAL SIMULATION OF AN OPTICAL PROCESSOR. THE PAPER DETAILS THE MODEL AND ALGORITHM USED TO REDUCE THE SIZE OF THE PROPOSED NETWORK AND TO ELIMINATE THE EFFECTS OF ANGULAR AMBIGUITY PRESENT IN THE IMAGE.
