Stereo-Plane-SLAM

Authors: Zhang, Xiaoyu; Wang, Wei; Qi, Xianyu; Liao, Ziwei.

Plane feature is a kind of stable landmark to reduce drift error in SLAM system. It is easy and fast to extract planes from dense point cloud, which is commonly acquired from RGB-D camera or lidar. But for stereo camera, it is hard to compute dense point cloud accurately and efficiently. In this project, we propose a novel method to compute plane parameters using intersecting lines which are extracted from the stereo image. The plane features commonly exist on the surface of man-made objects and structure, which have regular shape and straight edge lines. In 3D space, two intersecting lines can determine such a plane. Thus we extract line segments from both stereo left and right image. By stereo matching, we compute the endpoints and line directions in 3D space, and then the planes from two intersecting lines. We discard those inaccurate plane features in the frame tracking. Adding such plane features in stereo SLAM system reduces the drift error and refines the performance. We test our proposed system on public datasets and demonstrate its robust and accurate estimation results, compared with state-of-the-art SLAM systems.

Related Publications:

For more details, please read our paper Stereo Plane SLAM Based on Intersecting Lines, which has been accepted by IROS 2021.

If you have some questions to discuss, you could contact me by e-mail ([email protected]) for fast reply.

If you use our system in an academic work, please cite the paper.

@INPROCEEDINGS{9635961,  
  author={Zhang, Xiaoyu and Wang, Wei and Qi, Xianyu and Liao, Ziwei},  
  booktitle={2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},   
  title={Stereo Plane SLAM Based on Intersecting Lines},   
  year={2021}, 
  pages={6566-6572},  
  doi={10.1109/IROS51168.2021.9635961}
}

Video

https://www.youtube.com/watch?v=3VWF-JJU9T8

Stereo-Plane-SLAM System

We build our SLAM system based on ORB-SLAM2 Stereo version. For some prerequisites, you could read their page, ORB-SLAM2

We only add PCL library to deal with point cloud, we tested on PCL-1.80. The system only supports Stereo sensor. We have not removed those unrelevent components coming from ORB-SLAM2.