审稿中...

5. Conclusion 结论

We have presented a full SLAM system for monocular, stereo and RGB-D sensors, able to perform relocalization, loop closing and reuse its map in real-time in standard CPUs.

本文呈现了一个基于于单目，双目和RGB-D传感器的完整SLAM框架，在实时和标准的CPU的前提下能够进行重新定位和回环检测，以及地图的重用。

We focus on building globally consistent maps for reliable and long-term localization in a wide range of environments as demonstrated in the experiments.

在实验当中，我们关心的是在大场景中建立可用的地图和长期的定位。

The comparison to the state-of-the-art shows very competitive accuracy of ORBSLAM2, being in most cases the most accurate solution.

与此前的SLAM方案进行对比，在大多数的情况下，ORB-SLAM2展现出一样好的精确程度。

Surprisingly our RGB-D results demonstrate that if the most accurate camera localization is desired,

值得注意的是，我们的RGB-D实现的结果显示，如果相机的定位精度更好，

Bundle Adjustment performs better than direct methods or ICP, with the additional advantage of being less computationally expensive.

那么BA将会比直接法或者ICP的方法更好，将会减少一些额外的计算量。

We have released the source code of our system, with examples and instructions so that it can be easily used by other researchers.

我们开放了我们系统的源码，和一些例子和指导，以便于能够更加方便为其他研究者而使用。

We are aware that it has been already used out-of-the-box in [23].

在文献[23]中已经开始应用到我们的方案。

Future extensions might include, to name some examples, non-overlapping multi-camera, fisheye or omnidirectional cameras support, large scale dense fusion, cooperative mapping or increased motion blur robustness.

未来的方向可能包含，测试更多的序列，多视角相机，鱼眼相机或者其他全相相机的，大场景的稠密重建，以及联合建图或者增加运动模糊的鲁棒性。