OBJECT DISCRIMINATION VIA BULK TEMPERATURE/VELOCITY FILTERING

AMONG THE WAYS PASSIVE ELECTRO-OPTICAL SENSORS MIGHT DISCRIMINATE BETWEEN REENTRY VEHICLES AND THE OBJECTS WHICH ACCOMPANY THEM ARE THE MEASUREMENT OF OBJECT TEMPERATURE USING THE RELATIVE BRIGHTNESS OF EACH OBJECT IN DIFFERENT SPECTRAL BANDS, AND THE MEASUREMENT OF BRIGHTNESS AND TEMPERATURE CHANGES OF EACH OBJECT AS A FUNCTION OF TIME. AN IMPORTANT PROBLEM TO BE SOLVED IS HOW TO IMPLEMENT THESE MEASUREMENTS IN A COMPUTATIONALLY EFFICIENT WAY SO THAT THEY CAN BE MADE ON VERY LARGE NUMBERS OF OBJECTS WITH PRACTICAL AMOUNTS OF HARDWARE. TO THIS END, A NEW APPROACH IS BEING DEVELOPED THAT USES BULK PROCESSING OF (POSSIBLY) MULTI-SPECTRAL IMAGES TO PROVIDE HIGH PERFORMANCE WITH RELATIVELY SIMPLE HARDWARE. BASIC ALGORITHMS ARE BEING DEMONSTRATED AND EVALUATED VIA COMPUTER SIMULATION, AND PRELIMINARY DESIGN OF A PROCESSOR ARCHITECTURE IS UNDERWAY. DESIGN AND FABRICATION OF A BRASSBOARD PROCESSOR USING LOW-COST COMMERCIAL COMPONENTS WOULD BE UNDERTAKEN IN A LATER PHASE. IN ADDITION TO ITS UTLITY IN A LARGE NUMBER OF STRATEGIC DEFENSE SYSTEMS, THIS NEW APPROACH COULD ALSO BE APPLIED TO VARIOUS MILITARY RECONNAISSANCE, SURVEILLANCE AND INTELLIGENCE SYSTEMS, INCLUDING THOSE FOR CRUISE-MISSILE DETECTION, AS WELL AS TO COMMERICAL ROBOTICS SYSTEMS.