Is there any way for to do SLAM using only one monocular camera with rtabmap.

[rag0nn]

Hi,
I'm beginner for SLAM. I tried to use rtabmap with monocular data but ı can't handle calculation of /odom data. Rtabmap has 3 nodes to do this: rgbd_odometry, icl_odometry and stereo_odometry, is there any way to use with monocular data these nodes with different parameters? Or is this impossible to do this without depth data.
Thanks!

My Setup:
ROS2 Humble
Ubuntu 22.04

[matlabbe]

We don't provide an odometry node to do mononular visual odometry. You may search for a VIO package/library to do it.

[rag0nn]

Thanks for helps!
I'm using MiDaS to calculate depth images. I provide camera info, rgb image and depth image to rgbd_odometry node but node doesn't send any data. Is there any detailed guide for this? Am i have to run with any other node?

[matlabbe]

Is rgbd_odometry logging something? How long is required to generate the depth image, there could be a sync issue where depth image has significant more delay than the two other topics. The sync_queue_size and/or topic_queue_size could be increased. Are depth images published with same stamp than input image?

Using these two tools on all input topics can give a better idea:

ros2 topic hz ...
tos2 topic delay ...