AN INTELLIGENT VISUAL PERCEPTION SYSTEM BASED ON RETINA STRUCTURE

RESEARCHERS ARE DESIGNING AN INTELLIGENT VISION PERCEPTION SYSTEM MODELED AFTER A HUMAN COGNITIVE SYSTEM (RETINA); SPECIFICALLY, A MULTILAYER PARALLEL FOCAL PLANE ARRAY IMAGE PROCESSING SYSTEM WHICH COMBINES HIGHLY-ADVANCED OPTICAL INTERCONNECTS AND HOLOGRAM TECHNIQUES, WITH A MULTILAYER SERIAL-PARALLEL STRUCTURE:A RETINALLY INSPIRED ALGOTECTURE (RETINA). THE 3-LAYER RETINA STRUCTURE AND ALGORITHMS WILL BE CAREFULLY STUDIED AND IMPLEMENTED. THE FIRST LAYER WILL PERFORM EDGE ENHANCEMENT, THE SECOND WILL REDUCE AND/OR ELIMINATE NOISE, AND THE THIRD WILL IMPLEMENT LINE CONSTRUCTION ALGORITHMS TO ENHANCE RESULTS. SOBOL, MEDIAN, AND 3-PIXEL LINE SEGMENTS WILL BE USED FOR EACH CORRESPONDING LAYER. THIS PROGRAM WILL PRODUCE: (1) A USEFUL ALGOTECTURE, (2) DETERMINATION OF HOW IT WILL WORK WITH NOISY, NONUNIFORMLY-LIT IMAGES, (3) AN UNDERSTANDING OF COMPONENT TOLERANCE, (4) THE CRITICAL WAVEGUIDE HOLOGRAM ILLUMINATOR, AND (5) A FIRST-ORDER SYSTEM DESIGN. DEVELOPERS WILL CONSTRUCT, PERFECT, AND TEST AN EDGE LIT HOLOGRAM WHICH IS THE CRITICAL PHYSICAL COMPONENT AND THEN CREATE A FIRST-ORDER DESIGN OF A FULL SYSTEM, TO BE USED IN IMPLEMENTING THE DESIRED ALGOTECTURES. THIS WILL ALLOW ASSESSMENT OF ITS SIZE, WEIGHT, POWER CONSUMPTION, ETC.