Efficient vision-based pose determination method of five coplanar points

Abstract

The existing approaches to a solution of object pose fall into two distinct categories: analytical solutions and iterative solutions. In this paper, a two-step pose estimation method based on five coplanar points that combines the advantages of the above two methods is studied. In the first step, after the image coordinates of the points are extracted through image processing, the pose of the object is determined with two affine invariants formed by the five points. In the second step, an iterative method is introduced, and the results of the first step are passed to the current step, being used as initial values of the iterative process to ensure the accuracy and convergence rate of the nonlinear algorithm. The proposed method is then compared with two other existing methods. Experimental results demonstrate that the proposed method has the advantage of lower computational cost. The accuracy, which is the main concern during real-time applications, is also improved.