Authors: Fu Li([email protected]),Shaowu Yang([email protected])
Current version: 1.0.0
CORB-SLAM(collaborative ORB-SLAM) is a centralized multi-robot visual SLAM system based on ORB_SLAM2[1-2]. The ORB-SLAM2 is a versatile and accurate visual SLAM method that has been popularly applied in single robot applications. However, this method cannot provide support to multi-robot cooperation in environmental mapping. The CORB-SLAM system consists of multiple ORB_SLAM2 clients for local mapping and a central server for global map fusion. Specifically, we extend each of the ORB_SLAM2 clients with a memory managing module that organizes the local map and communicates with the central server. In the central server, we detect the overlaps of multiple local maps by the DBoW method, and fuse these maps by utilizing the PnP method and global optimization through bundle adjustment.
[1] F. Li, S. Yang, X. Yi, X. Yang. CORB-SLAM: a Collaborative Visual SLAM System for Multiple Robots. submitted.
- ubuntu 14.04
- ROS indigo
- ORBSLAM2 1.0.0
- boost
Inatall ROS indigo according to the instructions in ROS wiki.
Our CORB-SLAM system is build on the foundation of [ORB_SLAM2(https://github.com/raulmur/ORB_SLAM2). You should follow the instructions provided by ORB_SLAM2 build its dependencies. We do not list here.
We use boost library to serialize and deserialize the data. We can install boost library using the following instruction in terminal.
sudo apt-get instal libboost-dev
The CORB-SLAM runs as the ROS package. and the CORB-SLAM src directory should be the ROS package directory, catkin_src/. We use catkin tool to organize the CORB-SLAM packages, corbslam_client and corbslam_servcer.
Terminal in the catkin_src/ directory.
catkin_make
catkin_install
roscore
rosrun corbslam_servcer PATH_TO_VOCABULARY/ORBvoc.txt PATH_TO_CONFIGURATION/KITTIX.yaml
- run KITTI datasets
Download the dataset (grayscale images) from http://www.cvlibs.net/datasets/kitti/eval_odometry.php
Execute the following command. Change KITTIX.yaml
to KITTI00-02.yaml, KITTI03.yaml or KITTI04-12.yaml for sequence 0 to 2, 3, and 4 to 12 respectively. Change PATH_TO_DATASET_FOLDER
to the uncompressed dataset folder. Change SEQUENCE_NUMBER
to 00, 01, 02,.., 11.
rosrun corbslam_client_stereo_kitti corbslam_client_stereo_kitti Vocabulary/ORBvoc.txt Examples/Stereo/KITTIX.yaml PATH_TO_DATASET_FOLDER CLIENT_ID
[1] Mur-Artal R, Montiel J M M, Tardos J D. ORB-SLAM: a versatile and accurate monocular SLAM system[J]. IEEE Transactions on Robotics, 2015, 31(5): 1147-1163.
[2] Mur-Artal R, Tardos J D. ORB-SLAM2: an Open-Source SLAM System for Monocular, Stereo and RGB-D Cameras[J]. arXiv preprint arXiv:1610.06475, 2016.
CORB-SLAM is released under a GPLv3 license.
For a closed-source version of M2SLAM for commercial purposes, please contact the authors.