DETECTABILITY, UNIQUENESS, AND RELIABILITY OF EIGEN WINDOWS FOR STABLE VERIFICATION OF PARTIALLY OCCLUDED OBJECTS

This paper describes a method for recognizing partially occluded objec ts for bin-picking tasks using eigenspace analysis, referred to as the ''eigen window'' method, that stares multiple partial appearances of an object in an eigenspace. Such partial appearances require a large a mount of memory space. Three measurements, detectability, uniqueness, and reliability, on windows are developed to eliminate redundant windo ws and thereby reduce memory requirements. Using a pose clustering tec hnique, the method determines the pose of an object and the object typ e itself. We have implemented the method and verified its Validity.