On the recognition of curved objects

The problem of determining the identity and pose of occluded objects from noisy data is determined. Previous work has used the local geometry of small patches of an object´s surface in a constrained search process to solve this problem for the case of rigid polygonal objects using 2-D sensory data or rigid polyhedral objects using 3-D data. The method is extended in 2-D to curved objects and applies to the recognition of 2-D objects from 2-D data or of 3-D objects in stable positions from 2-D data