GROUND MOBILE TARGET DISCRIMINATION TECHNIQUES WITH COMBINED USE OF NON-IMAGE AND IMAGE BASED DATA

FOR THIS SBIR EFFORT, WE WILL FUSE NON-IMAGE AND IMAGE-BASED DATA FOR EXTRACTING MOBILE-MISSILE TARGETS FROM THE BACKGROUND CLUTTER. THE PROVEN MOBILE TARGET ANALYSIS TECHNOLOGY DEVELOPED BY PLANNING RESEARCH CORPORATION USING NON-IMAGE BASED DATA WILL BE USED AS AN INITIAL APPROACH TO NARROW DOWN THE POTENTIAL TARGET SITES. THE NEXT ACTION IS TO EMPLOY LOW RESOLUTION MULTISPECTRAL IMAGERY TO REJECT THE BACKGROUND AREAS THAT HAVE A LOW PROBABILITY OF TARGET LOCATION. FOR FEATURE EXTRACTION AND CAMOUFLAGE DETECTION PROCESSED, HIGH RESOLUTION IMAGE DATA LIKE SAR, ELECTRO-OPTICAL, THERMAL IR AND EVEN VIDEO IMAGERY WILL BE EXPLOITED. PRECISION TARGETING FOR A BALLISTIC MISSILE DELIVERY SYSTEM WILL BE ACCOMPLISHED BY A SCENE MATCHING METHODOLOGY USING THE STABLE STRUCTURE AREA AROUND THE TARGET LOCATION. WITH MULTI-TEMPORAL IMAGE DATA, MOBILE MISSILES CAN BE DETECTED BY MEANS OF A STRUCTURAL CHANGE DETECTION METHODOLOGY; THIS NOVEL APPROACH IS TO PERFORM THE CANCELLATION OF THE STATIONARY OBJECTS BY MEANS OF SUBTRACTING THE SCENE'S STRUCTURAL UNITS INSTEAD OF THE PIXELS. THE BASIS FOR THE PROPOSED IMAGE EXPLOITATION TECHNOLOGIES IS THE STABLE STRUCTURE THEORY AND ITS ASSOCIATED ALGORITHMS DEVELOPED BY DR. HSU OF SR&E. SINCE THE MAJORITY OF THE PROPOSED ALGORITHMS HAVE BEEN EVALUATED AND DETERMINED AS VERY EFFECTIVE FOR EXTRACTING GROUND TARGETS FROM THE BACKGROUND CLUTTER, WE FIRMLY BELIEVE THAT THE PROPOSED METHODOLOGIES WILL WORK FOR THE MOBILEMISSILE TARGETS AS WELL.