Abstract

This paper proposes a method for recovery of pro jective shape and motion from multiple images by factorization of a matrix containing the images of all scene points. Compared to previous meth- ods, this method can handle perspective views and occlusions jointly. The pro jective depths of image points are estimated by the method of Sturm & Triggs [11] using epipolar geometry. Occlusions are solved by the extension of the method by Jacobs [8] for filling of missing data. This extension can exploit the geometry of perspective camera so that both points with known and unknown pro jective depths are used. Many ways of combining the two methods exist, and therefore several of them have been examined and the one with the best results is presented. The new method gives accurate results in practical situations, as demonstrated here with a series of experiments on laboratory and outdoor image sets. It becomes clear that the method is particularly suited for wide base-line multiple view stereo. Keywords: pro jective reconstruction, structure from motion, wide base- line stereo, factorization