This work deals with verification of local 3D geometric model recovered from noisy data. Verification is understood as making decision about consistency of the model with a set of images. To evaluate the consistency, four correlation measures with a different invariance to image intensity deformations are examined. The approach is proposed to verify a disk-like local 3D surface model in the process of bottom-up surface reconstruction.

First, the idea of verification will be breafly explained. Then we focus on studying behaviour of proposed consistency measures and decision-making algorithm on the planar object and simple conditions. And in the end of the talk the experiments with real reconstruction scenes will be shown.