审稿中...

Abstract 摘要

We present ORB-SLAM2 a complete SLAM system for monocular, stereo and RGB-D cameras, including map reuse, loop closing and relocalization capabilities.

ORB-SLAM2是基于单目，双目和RGB-D相机的一套完整的SLAM方案。它能够实现地图重用，回环检测和重新定位的功能。

The system works in real-time in standard CPUs in a wide variety of environments from small hand-held indoors sequences, to drones flying in industrial environments and cars driving around a city.

无论是在室内的小型手持设备，还是到工厂环境的无人机和城市里驾驶的汽车，ORB-SLAM2都能够在标准的CPU上进行实时工作。

Our backend based on Bundle Adjustment with monocular and stereo observations allows for accurate trajectory estimation with metric scale.

ORB-SLAM2在后端上采用的是基于单目和双目的光束法平差优化（BA）的方式，这个方法允许米制比例尺的轨迹精确度评估。

Our system includes a lightweight localization mode that leverages visual odometry tracks for unmapped regions and matches to map points that allow for zero-drift localization.

此外，ORB-SLAM2包含一个轻量级的定位模式，该模式能够在允许零点漂移的条件下，利用视觉里程计来追踪未建图的区域并且匹配特征点。

The evaluation in 29 popular public sequences shows that our method achieves state-of-theart accuracy, being in most cases the most accurate SLAM solution.

我们用29个广泛使用的公共数据测试的结果显示，在大多数情况下，本文方案比此前方案精度更高，

We publish the source code, not only for the benefit of the SLAM community, but with the aim of being an out-ofthe-box SLAM solution for researchers in other fields.

此外，我们开源了ORB-SLAM2源代码，不仅仅是为了整个SLAM领域，同时也希望能够为其他领域研究者提供一套SLAM的解决方案。