SHAPE FROM VIDEO V.S. STILL IMAGES

In this paper we compare two dense matching approaches. The fi rst one has been developed in the context of shape from video using minimal path search. The second one is a PDE-based approach, and would be expected to give better results for shape from (a small number of) still images and for wide baseline situations. Both methods use as input the fully calibrated camera parameters, that have been obtained after structure form motion recovery from uncalibrated images. Emphasis lies on the usage of only a small amount of (high resolution) images instead of a (low resolution) video sequence. We use ground truth synthetic data as well as real data to compare the two algorithms.