Object recognition based on characteristic view classes

Abstract

A characteristic view (CV) class of a planar-faced solid (PFS) is a class of representative projection views of the solid such that all members of the class have isomorphic planar graph representation and can be related by a 3D geometric transformation. The viewing space of a PFS can be partitioned into regions called characteristic view domains (CVDs), such that a view of the PFS from any location within a region belongs to the same CV class and is different from all the CVs of the neighboring regions. The authors present a method of partitioning the viewing space into CVDs and finding the set of CV classes for a given PFS. Using the CV classes as the representation of the model object, they also present a method for model-based object recognition. Algorithms for representing, sorting, classifying, and graphically matching the CVs are given. All of these operations are needed in the object recognition method proposed.<>