No abstract

Abstract-In this paper, we describe a method for solving the Euclidean reconstruction problem with a perspective camera model by incrementally performing Euclidean reconstruction with either a weak or a paraperspective camera model. With respect to other methods that compute shape and motion from a sequence of images with a calibrated camera, this method converges in a few iterations, is computationally efficient, and solves for the sign (reversal) ambiguity. We give a detailed account of the method, analyze its convergence, and test it with both synthetic and real data.

This paper presents a method for estimating the position and orientation of a camera with respect to a known 3-D object from line correspondences. The main idea of the method is to estimate a pose with either a weak perspective or a paraperspective camera model and to improve this pose iteratively. At convergence the result is compatible with a perspective camera model. This iterative improvement of a linear (affine) camera model has already been used for points but has never been extended to lines. Known methods which compute pose from line correspondences deal with a set of nonlinear equations which are solved either in closed-form or using minimization techniques. These methods have to deal with multiple solutions. In contrast our method starts with a solution which is very close to the true solution and converges in very few iterations (typically three to five iterations). The rank analysis of the linear system to be solved at each iteration allows us to characterize geometric configurations which defeat the algorithm.

Abstract. In this paper we describe a method to perform Euclidean reconstruction with a perspective camera model. It incrementally performs reconstruction with a paraperspective camera in order to converge towards a perspective model. With respect to other methods that compute shape and motion from a sequence of images with a calibrated perspective camera, this method converges in a few iterations, is computationally efficient, and does not suffer from the non linear nature of the problem. Moreover, the behaviour of the algorithm may be simply explained and analysed, which is an advantage over classical non linear optimization approaches. With respect to 3-D reconstruction using an approximated camera model, our method solves for the sign (reversal) ambiguity in a very simple way and provides much more accurate reconstruction results.