SLAM realization using a stereo camera (Intel R200 RealSense equipped on intel AERO)

1. Flashing the Intel AERO [1][2]

Key:
(1) Erase Yocto and use all space for Ubuntu
(2) Press "shift" + "ESC" for accessing to the GRUB screen (if keyboard and mouse are frozen...)

2. Install RealSense SDK [2]

3. Install ROS [3]

4. Install RTAB-Map [4]

5. Install ORBSLAM2
(1) Pangolin
(2) OpenCV
(3) Eigen3

6. Make catkin workspace

▪ mkdir -p ~/catkin_ws/src
▪ cd ~/catkin_ws/src
▪ catkin_init_workspace
▪ cd ~/catkin_ws/
▪ catkin_make

7. Install ORBSLAM2 & ROS Nodes [4]

Key: Copy necessary library to correct directory. (Solution for installing ROS Nodes) [5]

To do list:
(1) Control the intel AERO screen remotely. (like VNC)
(2) Completely understand the principle of ORBSLAM2
(3) Send SLAM information to flight controller (mavros)

8. Get camera parameters and save them in a .yaml file [4]

9. Run ORBSLAM2

roscd realsense_camera
roslaunch realsense_camera r200_nodelet_rgbd.launch &
rosrun ORB_SLAM2 RGBD {path}/ORB_SLAM2/Vocabulary/ORBvoc.txt {path}/RealSense.yaml /camera/depth_registered/image_raw:=/camera/depth/image_raw


Reference:
[1] https://github.com/intel-aero/meta-intel-aero/wiki/02-Initial-setup
[2] https://github.com/intel-aero/meta-intel-aero/wiki/90-(References)-OS-user-Installation
[3] https://github.com/intel-aero/meta-intel-aero/wiki/05-Autonomous-drone-programming-with-ROS
[4] simultaneous localization and mapping on a quadcopter
[5] ubuntu16.04编译ORBSLAM2问题解决

留言

這個網誌中的熱門文章

Tuing PID parameters in QGroundcontrol (2)

Useful PX4 Parameters

Burn linux image to eMMC storage on Banana Pi M3