A computational model of view degeneracy

We quantify the observation by Kender and Freudenstein [24] that degenerate Views occupy a significant fraction of the viewing sphere surrounding an o bject. For a perspective camera geometry, we introduce a computational mode l that can be used to estimate the probability that a view degeneracy will occur in a random view of a polyhedral object. For a typical recognition sy stem parameterization, view degeneracies typically occur with probabilities of 20 percent and, depending on the parameterization, as high as 50 percen t. We discuss the impact of view degeneracy on the problem of object recogn ition and, for a particular recognition framework, relate the cost of objec t disambiguation to the probability of view degeneracy. To reduce this cost , we incorporate our model of view degeneracy in an active focal length con trol paradigm that balances the probability of View degeneracy with the cam era field of view. In order to validate both our view degeneracy model as w ell as our active focal length control model, a set of experiments are repo rted using a real recognition system operating on real images.