FRACTAL IMAGE ANALYSIS, AUTOMATION, AND DETECTION FOR THE INTEGRATED UNDERSEA SURVEILLANCE SYSTEM

THE IUSS HAS AN URGENT NEED FOR MORE EFFICIENT METHODS OF SIGNAL PROCSSSING. TO MEET THIS END, SUBMARINE DETECTION AND CLASSIFICATION SYSTEMS MUST BE REEXAMINED AND MODIFIED. THIS PROPOSAL EXAMINES THE POTENTIAL OF AN INNOVATIVE TECHNIQUE CALLED FRACTAL IMAGE ANALYSIS TOAID OPERATORS TO ACCURATELY AND EFFICIENTLY DETECT AND CLASSIFY SUBMARINES IN A PASSIVE SONAR ENVIRONMENT. FRACTAL IMAGE ANALYSIS IS AN EXCELLENT CHOICE FOR UNDERSEA SURVEILLANCE BECAUSE IT HAS BEEN SUCCESSFULLY APPLIED TO OTHER IMAGES TO CLASSIFY REGIONS AND TO DETECT AND ENHANCE EDGES. FURTHER THERE IS REASON TO BELIEVE THAT SOUND RADIATED FROM SUBMARINES FOLLOWS A FRACTAL MODEL. IN PHASE I A SINGLE FRACTAL MEASURE WILL BE APPLIED TO A LIMITED SET OF LOFARGRAMS. THE PROPOSED WORK INCLUDES COLLECTION OF DATA FOR FRACTAL ANALYSIS, WRITING AND DEBUGGING OF FRACTAL ANALYSIS CODE, OPTIMIZATION OF THE PARAMETERS OF THAT CODE, EVALUATION OF THE OPTIMIZED CODE ON REAL DATA AND PREPARATION OF A PRELIMINARY PLAN FOR PHASE II WORK INTEGRATING FRACTAL ANALYSIS AND AUTOMATION. PHASE II WILL SEE AUTOMATION OF THE METHOD AND ITS APPLICATION TO OTHER FRACTAL MEASURES AND OTHER IMAGES OF INTEREST IN THE SURVEILLANCE PROBLEM. AN OPERATIONAL SYSTEM WILL BE DEVELOPED IN PHASE III.