A Tactile Recognition Strategy for Planar Objects

Abstract

AN OUTSTANDING PROBLEM IN MODEL-BASED RECOGNITION OF OBJECTS BY ROBOT SYSTEMS IS HOW THE SYSTEM SHOULD PROCEED WHEN THE ACQUIRED DATA ARE INSUF- FICIENT TO IDENTIFY UNIQUELY THE MODEL INSTANCE AND MODEL POSE THAT BEST INTERPRETTTHE OBJECT. SUCH A SITUATION CAN ARISE WHEN THERE ARE MULTIPLE MODEL INSTANCES THAT COULD BE INTERPRETATIONS OF THE OBJECT, OR WHEN THERE ARE AMBIGUOUS POSES FOR A MODEL INSTANCE. THIS WORK PROPOSES A GENERIC METHOD FOR AUTOMATICALLY FINDING A PATH ALONG WHICH THE ROBOT COULD MOVE A TACTILE SENSOR, SO THAT THE ROBOT SYSTEM CAN UNIQUELY IDENTIFY THE OBJECT. USING ASSUMPTIONS THAT SATISFY REAL- WORLD SENSING CONSTRAINTS, IT IS SHOWN THAT POLYGONS ON A PLANE CAN BE IDEN TIFIED BY FINDING A LINEAR PATH THAT PASSES THROUGH AN UNSENSED FACE OF EACH INTERPRETATION. A COMPUTATIONALLY FEASIBLE METHOD IS PRESENTED FOR THE TWO-DIMENSIONAL CASE, BASED ON A CONVERSION OF THE PROBLEM TO THAT OF FINDING THE INTERSECTION OF ALL PATHS PASSING THROUGH EACH UNSENSED FACE. EXTENSIONS OF THE METHOD, INCLUDING THE IMPORTANT CASE OF RECOGNIZING POLY- HEDRAL OBJECTS, ARE BRIEFLY EXAMINED.