Shape from Shading: Provably Convergent Algorithms and Uniqueness Results

An explicit representation for the surface corresponding to a shaded image is presented and proven to be correct (under standard conditions). Uniqueness of the surface is an immediate consequence. Using this representation, various iterative algorithms for shape reconstruction are derived. It has been proven that all these algorithms converge monotonically to the correct surface reconstruction, and they have been shown experimentally to be fast and robust. Some of the results of this paper extend previous ones to the case of illumination from a general direction.

[1]  Tyrone E. Duncan,et al.  Numerical Methods for Stochastic Control Problems in Continuous Time (Harold J. Kushner and Paul G. Dupuis) , 1994, SIAM Rev..

[2]  Alex Pentland,et al.  A simple algorithm for shape from shading , 1992, Proceedings 1992 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[3]  B. Saxberg.  A Modern Differential Geometric Approach to Shape from Shading , 1989 .

[4]  John Oliensis,et al.  A global algorithm for shape from shading , 1993, 1993 (4th) International Conference on Computer Vision.

[5]  P. Dupuis,et al.  Direct method for reconstructing shape from shading , 1991, Proceedings 1992 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[6]  Anna R. Bruss.  The Eikonal equation: some results applicable to computer vision , 1982 .

[7]  Berthold K. P. Horn,et al.  Shape from shading , 1989 .

[8]  B. Saxberg,et al.  An application of dynamical systems theory to shape from shading , 1989 .

[9]  P. Dupuis,et al.  An Optimal Control Formulation and Related Numerical Methods for a Problem in Shape Reconstruction , 1994 .

[10]  E. Rouy,et al.  A viscosity solutions approach to shape-from-shading , 1992 .

[11]  Paul Dupuis,et al.  Direct method for reconstructing shape from shading , 1991, Optics & Photonics.

[12]  J. Oliensis,et al.  Shape from shading as a partially well-constrained problem , 1991, Proceedings. 1991 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.