Consistent topographic surface labelling

This paper describes work aimed at consistently labelling surface facets us ing topographic classes derived from mean and Gaussian curvature measuremen ts. There are two distinct contributions.]Firstly, we develop a statistical model which allows label probabilities to be assigned to the different top ographic classes. These probabilities capture uncertainties in the computat ion of surface curvature from raw surface normal information. The probabili ties are computed using propagation of variance from the surface normal mea surements. The second contribution is to demonstrate how topographic surfac e labelling can be realised using probabilistic relaxation. The key ingredi ent is to develop a constraint dictionary for the feasible configurations o f the topographic labels that can occur on neighbouring faces of the surfac e mesh. These constraints relate to the legal adjacency of different topogr aphic structures together with the smoothness and continuity of uniform reg ions. (C) 1999 Pattern Recognition Society. Published by Elsevier Science L td. All rights reserved.