THREE-DIMENSIONAL VISION SYSTEM FOR BIN-PICKING.

A vision system for recognizing 3-D objects in a scene and evaluating their attitude is presented. This system uses a model-based approach to determine the attitude of 3-D industrial parts in the context of robot bin picking, and it is based on the matching of the extended Gaussian image (EGI) extracted from the surface normal distribution map of an observed object with the EGI generated from 3-D model of the same object using geometric modeling. It is shown how geometric modeling can be used to provide object descriptions suited to the problem of the 3-D object recognition and attitude evaluation.