An algebraic approach to shape-from-image problems
Artificial Intelligence
A Computational Approach to Edge Detection
IEEE Transactions on Pattern Analysis and Machine Intelligence
A Shape From Shading Analysis for a Single Perspective Image of a Polyhedron
IEEE Transactions on Pattern Analysis and Machine Intelligence
Lipschitzian optimization without the Lipschitz constant
Journal of Optimization Theory and Applications
Generalization of the Lambertian model and implications for machine vision
International Journal of Computer Vision
Recovering the shape of polyhedra using line-drawing analysis and complex reflectance models
Computer Vision and Image Understanding - Special issue on physics-based modeling and reasoning in computer vision
IEEE Transactions on Pattern Analysis and Machine Intelligence
Overcoming Superstrictness in Line Drawing Interpretation
IEEE Transactions on Pattern Analysis and Machine Intelligence
MULTIMEDIA '06 Proceedings of the 14th annual ACM international conference on Multimedia
Precise 3d reconstruction from a single image
ACCV'12 Proceedings of the 11th Asian conference on Computer Vision - Volume Part IV
| Hi-index | 0.14 |
We propose a new method for recovering the 3D shape of a polyhedral object from its single 2D image using the shading information contained in the image and the prior information on the object. In a strict sense, we cannot recover the shape of a polyhedron from an incorrect line drawing, even if it is practically almost correct. In order to overcome this problem, we propose a flexible face positioning method that can permit inconsistencies in the recovered shape that arise from vertex-position errors contained in incorrect line drawings. Also, we propose to use prior information about the horizontality and verticality of special faces and the convex and concave properties of the edges in order to attain good solutions and present a method of formulating such prior information as physical constraints. The shape-from-shading method is formulated as a minimization problem of a nonlinear cost function with the nonlinear constraints and its solution is searched by a global optimization algorithm. In the experiments with a synthetic image and three kinds of real images, shapes that are similar to those of the actual objects were recovered in all cases. As a result, the proposed method has proven to be effective in the shape recovery of simple-shape polyhedral objects.