RECOVERING THE 3D STRUCTURE OF TUBULAR OBJECTS FROM STEREO SILHOUETTES

Although silhouette-based image understanding is attractive from an en gineering viewpoint, recovering 3D shape from a single stereo pair of silhouette images of a generic multiple-object scene is a highly under constrained problem. With respect to a gray-level-based approach, the the loss of data due to mutual visual occlusions are even more severe. These problems are alleviated when the observed objects can be assume d to belong to some restricted class. In this paper we consider the ca se of almost vertical tubular objects (AVTOs), i.e. generalized cylind ers with some restrictions on their axis' shape and pose relative to t he stereo pair. This restriction, together with the assumption that th e scene must be explained with the minimum number of objects consisten t with the observations, allows one to devise an effective reconstruct ion algorithm. The object shape/location parameters are estimated by r ecursive least-squares (Kalman) filtering. Constrained blind tracking is performed on the occluded sections by feeding the filters with the most likely parameter values compatible with the constraints induced b y the observed images. The case of AVTOs with circular cross-section i s analyzed in some detail, with examples taken from an actual implemen tation of the algorithm in the field of agricultural automation. (C) 1 997 Pattern Recognition Society. Published by Elsevier Science Ltd.