This paper presents a new method for integrating di erent low level vision modules, stereo and shape from shading, in order to improve the 3D reconstruction of visible surfaces of...
Mostafa G.-H. Mostafa, Sameh M. Yamany, Aly A. Far...
Second-order priors on the smoothness of 3D surfaces are a better model of typical scenes than first-order priors. However, stereo reconstruction using global inference algorithm...
Oliver J. Woodford, Philip H. S. Torr, Ian D. Reid...
In this paper, we present a two-layer generative model that incorporates generic middle-level visual knowledge for dense stereo reconstruction. The visual knowledge is represented...
Abstract. This paper presents an algorithm for voxel-based reconstruction of objects with general reflectance properties from multiple calibrated views. It is assumed that one or m...
Abstract. Fitting parameterized 3D shape and general reflectance models to 2D image data is challenging due to the high dimensionality of the problem. The proposed method combines ...