-
Notifications
You must be signed in to change notification settings - Fork 1.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Are there any demo of D534i+SLAM with ROS2? #3046
Comments
Hi @PANQingrui SLAM on ROS2 can be difficult to implement. RealSense ROS2 users who have attempted it have typically used a combination of slam_toolbox and depthimage_to_laserscan. The link for slam_toolbox is below. You can select a branch of the tool for a specific ROS2 version. As you are using 22.04, Humble or Iron would likely be the best choice. https://github.com/SteveMacenski/slam_toolbox The ROS2 branch of depthimage_to_laserscan is here: https://github.com/ros-perception/depthimage_to_laserscan/tree/ros2 Alternatively, you could consider using ORB-SLAM with ROS2 and running it in Stereo mode with your D435i stereo depth camera. It has a branch for Humble. |
I'm running into this same issue. For traceability, I was able to get your D435i example (from 2019) running in a Docker container. Needs a little clean-up, but here's the container code:
And the accompanying compose file:
When I switched a few things to use Humble, I ran into a lot of missing topics (they would show in ros2 topic list, but a ros2 topic hz would indicate that they weren't actually publishing). Having a pre-baked solution would be great to have for long-term support of this product. Is it a hardware or driver limitation for getting it to ROS2? I know TF was clearly missing when I was fiddling around in rviz2. |
Thanks so much @Forsyth-Creations for sharing your Docker tracking solution! There is an official ROS2 RealSense SLAM guide that uses depthimage_to_laserscan at the link below, though it is old now and was designed for the obsolete https://intel.github.io/robot_devkit_doc/pages/rs_slam.html https://github.com/IntelRealSense/realsense-ros/tree/ros2-legacy The guide uses both a 400 Series camera and the RealSense T265 Tracking Camera, so if you do not have a T265 then you would also have to skip the T265 parts of the instructions and use the rs_launch.py launch file instead of the guide's suggested Another possible option is the ROS2 wrapper for the Kimera 'semantic SLAM' tool. https://github.com/MIT-SPARK/Kimera-VIO-ROS2 https://www.youtube.com/watch?v=Zjevg5wQTdI For a more modern ROS2 SLAM solution that is compatible with RealSense and is an out of the box solution, you could look at SLAMcore. https://www.imveurope.com/press-releases/slamcore-teams-realsense-provide-out-box-slam-algorithms |
Hi @PANQingrui and @Forsyth-Creations Do either of you require further assistance with this case, please? Thanks! |
Hi Marty, thank you for your suggestions. I haven't tried your suggestions yet, but I just checked this issue "#2992", and it seems to be similar to my needs. I will try ORB-SLAM first, and thank you for your help again! I will update here if further assistance is needed. |
You are very welcome, @PANQingrui - good luck! |
Hi @PANQingrui Do you require further assistance with this case, please? Thanks! |
Case closed due to no further comments received. |
I have checked related issues and it seems that "https://github.com/IntelRealSense/realsense-ros/wiki/SLAM-with-D435i" only support "Kinetic" and "Melodic". Since 5 years passed, I would like to know if there are any alternative applications, to build a SLAM with D435i on Unbuntu 22.04.
Thank you for your help with this matter!
The text was updated successfully, but these errors were encountered: