Minimizing the object space error for pose estimation: towards the most efficient algorithm

In this paper, we present an efficient branch-and-bound algorithm to globally minimize the object space error for the camera pose estimation. The key idea is to reformulate the pose estimation model using the optimal Lagrangian multipliers. Numerical simulation results show that our algorithm usually terminates in the first iteration and finds an -suboptimal solution. Furthermore, the efficiency of our algorithm is demonstrated by a comprehensive numerical comparison with two well-known heuristics. We also demonstrate the computational power of our algorithm by comparing it with the state-of-the-art global optimization package BARON.