Automatic Construction of Junction Dictionaries and Their Exploitation for the Analysis of Range Data

Abstract

A knowledge-guided system is presented for the analysis of range data obtained from the three-dimensional world. A model of the world is defined by a finite set of prototypes of three-dimensional objects, from which the system establishes its own knowledge about the world in the form of a junction dictionary. Next the system exploits the dictionary for the prediction of missing edges in the analysis of range data originated with real three-dimensional scenes, and thus constructs concise descriptions of the scenes in terms of vertices, edges, faces and spatial relationships among them. The system is flexible in the sense that the only thing an operator has to do when the world is changed is to replace the model by a new one; the system constructs its own knowledge about the new world and utilizes it for the analysis of range data.