kolinguo/orbslam3 repository overview

Container for launching and configuring ORB-SLAM3⁠, the state-of-the-art ORB-feature-based visual-inertial SLAM algorithm on stereo+IMU system.
This docker image has Ubuntu 18.04, CUDA 11.3.1, ROS Melodic, OpenCV 4.4.0, librealsense⁠, realsense-ros⁠, Python 3.6.9 with Jupyter Notebook.

Unfortunately, for RGB-D+IMU systems such as the RealSense D435i, ORB-SLAM3 does not have great performance and can sometimes lose track in featureless areas according to our experiments.

⁠Build & compile codebase

1. ./docker_setup.sh orbslam3 to pull the docker image and build the container.
2. cd slam_algorithms/ORB_SLAM3 && ./build.sh && ./build_ros.sh to compile the packages.

⁠Running

Run roslaunch ORB_SLAM3 rgbd_d435i.launch to launch tracking along with RealSense camera.
The trajectory is written as /maya-slam/slam_algorithms/ORB_SLAM3/Examples_old/ROS/ORB_SLAM3/FrameTrajectory.txt