From image sequences to recognized moving polyhedral objects

Describes the combination of several algorithms into a system which obtains visual motion from a sequence of images and uses it to recover the three dimensional geometry and three dimensional motion of polyhedral objects relative to the camera. The system goes on to recognize the recovered geometry as an object from a database of wireframe CAD models, a stage which also resolves the depth/speed scaling ambiguity inherent in visual motion processing and results in absolute depth and motion recovery. The system´s performance has been demonstrated on imagery from well carpentered CSG models and on natural imagery from simple wooden objects