Integration of stereo vision system and laser range finder for autonomous obstacle avoidance and map construction

In this paper, we investigate the possibility of integrating a binocular stereo vision system and a laser range finder to construct 3-D map for path planning and navigation. Our proposed system is realized by forming the task as an optimization problem of minimizing the alignment error between scanned local map and selected parts of the developing global map. The problem is then solved using the Simplex method. The computation is further accelerated by reducing the data size by the procedure of Grid Map. To increase the robustness of the searching, multiple initial guesses for the Simplex method are provided. The alignment parameters are used for the data generated by the stereo vision system to be integrated into a 3D map of the environment. Performance of the proposed architecture is verified by experiment results of a real-time mobile vehicle which is also equipped with obstacle avoidance capability.