A review of optimisation strategies used in simultaneous localisation and mapping

This paper provides a brief review of the different optimisation strategies used in mobile robot simultaneous localisation and mapping(SLAM)problem.The focus is on the optimisation-based SLAM back end.The strategies are classified based on their purposes such as reducing the computational complexity,improving the convergence and improving the robustness.It is clearly pointed out that some approximations are made in some of the methods and there is always a trade-off between the computational complexity and the accuracy of the solution.The local submap joining is a strategy that has been used to address both the computational complexity and the convergence and is a flexible tool to be used in the SLAM back end.Although more research is needed to further improve the SLAM back end,nowadays there are quite a few relatively mature SLAM back end algorithms that can be used by SLAM researchers and users.

Building a dense surface map incrementally from semi-dense point cloud and RGB images

目的:针对仅能通过轻型激光测距仪获取半稠密点云的环境地图构建问题,提出一种构建稠密表面模型的方法。该方法使机器人能够利用所构建的稠密表面模型地图完成定位、导航及目标搜索等任务。创新点:提出一种基于点云分割的点云表面重采样方法及一种基于点云概率模型的表面模型融合方法。对半稠密点云进行保留表面结构特性的重采样来获取观测数据的稠密表面模型。并递增式地将新获得的稠密表面模型融合进已有的稠密表面地图中,从而获得几何一致性较好的环境表面模型地图。实验效果:图6、7展示了基于本文方法所构建的稠密表面模型地图的效果。其几何结构精确且表面纹理清晰。此外,图8、9分别重点展示了表面重采样的作用以及本文提出的重采样方法的效果。图11则展示了本文方法对表面模型动态更新的较好支持。结论:使用本文所提方法,机器人可携带轻便式激光测距仪,获取半稠密点云后再进一步处理和融合得到几何一致性较高、表面精细的稠密表面问题模型地图,更好地实现定位、导航及目标搜索等任务。