SYMBOLIC RF SIGNATURE PREDICTION

THIS EFFORT WILL ASSESS THE TECHNICAL STATE OF THE ART FOR MODELBASED VISION APPROACHES TO ANALYZING RADAR AND OTHER ACTIVE SENSOR IMAGERY. MODEL-BASED APPROACHES RELY HEAVILY ON USING SEMANTIC MODELS TO PREDICT IMAGE OBJECTS AND FEATURES IN A MANNER THAT FACILITATES RECOGNITION OF THE OBJECT IN THE IMAGE. TO DATE, RESEARCH IN THIS AREA HAS FOCUSED PREDOMINATELY ON OPTICAL IMAGERY. THE STUDY WILL DEVELOP A RESEARCH PLAN DESCRIBING AN APPROACH TO BRING MODELBASED VISION TECHNIQUES FOR RADAR IMAGE ANALYSIS TO FRUITION. THIS WILL BE ACCOMPLISHED BY REVIEWING RELEVANT TECHNOLOGY, CREATING A BIBLIOGRAPHY OF APPLICABLE RESEARCH, DEFINING KEY TAXONOMIES AND CHARACTERISTICS OF THE TECHNICAL ISSUES, AND IDENTIFYING STRENGTHS AND WEAKNESSES (OR GAPS) IN CURRENT CAPABILITIES. A KEY TASK IN ACCOMPLISHING THESE GOALS WILL BE THE ORGANIZATION OF A WORKSHOP ADDRESSING MODEL-BASED VISION FOR ACTIVE SENSORS WITH AN EMPHASIS ON PREDICTIVE AND MODEL REPRESENTATION.