Screenshot from 2024-03-04 21-05-01환경
ROS version : noetic
Ubuntu version : 20.04
RGB-D Camera : Realsense D455
Installation
//realsense2_camera:
sudo apt-get install ros-$ROS_DISTRO-realsense2-camera
//imu_filter_madgwick:
sudo apt-get install ros-noetic-imu-filter-madgwick
//rtabmap_ros:
sudo apt-get install ros-noetic-rtabmap*
//robot_localization:
sudo apt-get install ros-noetic-robot-localizationRosbag
//Save all topic
rosbag record -O my_bagfile_1.bag /camera/aligned_depth_to_color/camera_info camera/aligned_depth_to_color/image_raw /camera/color/camera_info /camera/color/image_raw /camera/imu /camera/imu_info /tf_static /rtabmap/cloud_mapRoslaunch
roslaunch realsense2_camera opensource_tracking.launchSetting
The pointcloud and a bunch of arrows, or axes marks, will appear on screen. These axes represent all the different coordinate systems involved. For clarity you could remove most of them.
From the Displays Panel:
TF -> Frames, and then leave out as marked only map and camera_link. The first represents the world coordinate system and the second, the camera.
You may want to watch the on-line video as well:
From the Displays panel:
Image->Image Topic: rewrite to /camera/color/image_raw
Topic check
rtabmap에 intput으로 들어갈 rtabmap/cloud_map 토픽이 publish 되고 있는지 확인한다
rostpic echo /rtabmap/cloud_map1. 카메라를 들고 맵핑하려는 장소로 이동한다
2. 맵핑이 어느정도 완료되면 rosbag record 를 하고 있는 터미널로 가서 Ctrl + C를 눌러 종료한다
PCL visualization
1. PCL viewer
sudo apt-get install pcl-tools
rosrun pcl_ros pointcloud_to_pcd input:=/rtabmap/cloud_map
2. rtabmap viewer
roslaunch rtabmap_ros rtabmap.launch
rosbag play hyeon.bag
Result
다음과 같이 .pcd 형태로 저장된 것을 알 수 있다
Visualization
pcl_viewer 1709049377663825.pcd
[후기]
RGB-D 카메라 V-SLAM 튜토리얼이 있어 세팅해봤다
핸드헬드 타입을 만들어서 하려고 했으나, SBC 메모리 부족으로 SSD를 구매하기 전까지 당분간 데스크탑으로 연결해서
사용해봐야겠다
<Reference>
https://github.com/IntelRealSense/realsense-ros/tree/ros1-legacy
https://github.com/IntelRealSense/realsense-ros/wiki/SLAM-with-D435i