AUTOMATED TERRAIN INFERENCE FROM ELEVATION DATA

THE MODERN BATTLEFIELD ENVIRONMENT DEMANDS TOOLS WHICH CAN ASSIST ARMY COMMANDERS IN MAKING TACTICAL DECISIONS ABOUT A DYNAMIC ENVIRONMENT IN A TIMELY MANNER. KNOWLEDGE OF THE TERRAIN, SUCH AS AREAS OF RELATIVE HIGH GROUND, IS A CRUCIALFACTOR TO MANY OF THESE MODERN TACTICAL DECISION AIDS. IN ORDER TO AUTOMATE SUCH TACTICAL DECISION MAKING, THE COMPUTER MUST BE ABLE TO MAKE INFERENCES ABOUT THE TERRAIN FROM ELECVATION DATA WITHOUT HUMAN ASSISTANCE. COMPUTER ALGORITHMS EXIST THAT DETERMINE SEVERAL TERRAIN CHARACTERISTICS SUCH AS MASKED AREA AND LINE-OF-SIGHT. WHATIS LACKING, HOWEVER, IS A GENERAL TOOL THAT WILL INFER TERRAIN CHARACTERTISTICS FROM ELEVATION DATA AND PRODUCE A RASTER AND/OR VECTOR OUTPUT THAT IS AMENABLE TO FURTHER PROCESSING BY APPLICATIONS SUCH AS TACTICAL DECISION AIDS. MEASUREMENT CONCEPT CORPORATION (MC2), THEREFORE, PROPOSES ATECHNOLOGICALLY INNOVATIVE APPROACH TO DEVELOP SUCH AN AUTOMATED TERRAIN INFERENCE TOOL. THE PROPOSED DEVELOPMENT METHODS WILL BE BASED ON CLASSICAL IMAGE PATTERN RECOGNITIONAND/OR NEURAL NET TECHNIQUES USED IN COMBINATION WITH TERRAIN ANALYSIS METHODS.