3D Super-resolution Using Generalised Sampling Expansion
Abstract
Using a probabilistic interpretation of Papoulis' generalized sampl ing theorem, an iterative algorithm has been devised for 3D reconstruction of a Lambertian surface at sub-pixel accuracy. The problem has been formulated as a n optimization one in a Bayesian framework. The latter allows for introducing { \em a priori} information on the solution, using Markov Random Fields (MRF). Th e estimated 3D features of the surface are the albedo and the height which are obtained simultaneously using a set of low resolution images.