Pose estimation of 3D rigid object based on tentative point correspondence

Abstract

Estimating pose parameters of a 3D rigid object based on 2D monocular image is a fundamental problem in computer vision. State-of-art methods usually assume certain feature correspondences exist a priori between the input 2D image and object's 3D model. This presumption makes the problem more algebraically tractable. However, when there is no feature correspondence available a priori, how to estimate the pose of a general 3D object with monocular vision is still an open problem. In this paper, a new contour-based method is proposed, which features solving both the pose estimation problem and the feature correspondence problem simultaneously and iteratively. Experiment results show that this new method has fast convergence speed and good convergence radius. © 2012 Springer-Verlag Berlin Heidelberg.