MULTIPLE SENSORS FUSION FOR OBJECT DETECTION AND POSITION FINDING

THERE IS A NEED TO DETECT, TRACK, AND RECOGNIZE OBJECTS TO EFFECTIVELY USE ROBOTS AND TELEOPERATORS IN SPACE. A MULTIPLE SENSOR FUSION SYSTEM CAN FILL THE NEED. SUCH A FUSION SYSTEM HAS BEEN PROPOSED TO FUSE CANDIDATE DETECTION REPORTS TO PROVIDE ACCURATE TRACK AND CONFIDENT TARGET RECOGNITION. THE FUSION TRACKER IS A COMBINATION OF A MULTIPLE INPUT KALMAN FILTER AND A PROBABILISTIC LIKELIHOOD-RATIO CLASSIFIER. THE FILTER GENERATES A SINGLE TRACK ON EACH TARGET AND CLUTTER POINT, WHILE THE CLASSIFIER ACCEPTS INPUTS FROM ALL SYSTEM COMPONENTS AND MAKES THE FINAL TARGET DECISION. THE RESULT IS BETTER ACQUISITION, A MORE ACCURATE STABLE TRACK, AND CONFIDENT TARGET ID.