RECOGNITION OF OBJECTS IS AN IMPORTANT TASK IN COMPUTER VISION, A TASK MADE MORE DIFFICULT DUE TO OCCLUSION. THIS PAPER PRESENTS A METHOD FOR OBJECT RECOGNITION SUITABLE FOR OCCLUDED OBJECTS. THE METHOD IS DERIVED FROM FEATURE INDEXED HYPOTHESIS APPROACH, BUT PROVIDES A MUCH HIGHER EFFICIENCY WHEN ACCURACY IS IMPORTANT. A HIERARCHICAL REPRESENTATION OF THE OBJECT ENABLES THE RECOGNITION PROCESS TO BE EASILY SUBDIVIDED. AS A CONSEQUENCE OF THIS, THE PROPOSED METHOD OUTPERFORMS THE FEATURE INDEXED APPROACH. THE RECOGNITION COST FUNCTION FOR THE HIERARCHICAL APPROACH IS MATHEMATICALLY DERIVED. IT IS COMPARED TO THE FEATURE INDEXED APPROACH AND INDICATES A GAIN IN PERFORMANCE PROPORTIONAL TO THE SQUARE OF THE NUMBER OF FEATURES BEING USED. THIS ENABLES THE VISION SYSTEM TO BE ACCURATE AND EFFICIENT SIMULTANEOUSLY.