Pose estimation from less than six non coplanar points

Pose estimation has many applications in robotics, such as robot localization using a vision sensor, or position-based visual servoing. In this paper, a simple but an efficient pose estimation method from minimal point pairs is given. The method we propose is based on exhaustive optimization of only two most critical parameters. It allows to solve the local minima problem. The both model based and free model pose estimation are treated. Finally, simulation results are given to validate the proposed method