A New 2-Correspondences Monocular Vision Navigation Method Under Planar Motion

Abstract

Abstract Vision navigation has non-contact, high-precision and low-cost characteristics compared with other navigation method. This paper proposes a vision navigation method for planar moving platform in known scene based on two 2D-3D correspondences. Taking account of the arbitrary direction of camera and the relative pose between the camera and the platform, we propose a linear solver with unique solution, whose minimal number of requirement correspondences is 2 and can cope with arbitrary number of correspondences bigger than 2 directly. Our method is evaluated on both synthetic data and real-world data. The experimental results demonstrate that our method is high-accuracy and fast for vision navigation of planar moving platform.