| Abstract | This article introduces an approach, based on Bayesian Networks, for the grouping of 3-D surfaces extracted from data obtained by a laser ranging sensor. A methodology for the specification of the network is presented along with an approach for determining the conditional probabilities. Determination of the conditional probabilities is based on a compatibility function that measures the uncertainty in the quality of fit of the data to a model of the features in the scene. Several compatibility functions for the grouping of 3-D surfaces are presented. These are co planarity, parallel, planarity, and proximity. These compatibility functions are used with a Bayesian Network in determining belief values of possible groupings among the surfaces, in particular grouping for continuous surfaces and corners. This operation is a form of perceptual grouping of three dimensional data and is akin to the previous studies in perceptual grouping for two dimensional images. |
|---|
