matlabbe

> You mentioned in other thread as well as in paper, that Map assembling is computationally heavy process ... (and more precisely) when there is a loop closure and graph...

The number of neighbor links are roughly the same number of nodes added to the graph. I don't now what `Rtabmap/DetectionRate` you are using and actual processing time per node,...

@samarth-eic Yes, `Rtabmap/DetectionRate` is the update rate I referred. I corrected the typo in my previous post. Like what I thought, the processing time is high, so it cannot process...

> "the rate is limited either by Rtabmap/DetectionRate or actual maximum rate rtabmap can process based on the machine", Like this is the reason why we kept it at 30...

The normal segmentation is based on that experiment: http://wiki.ros.org/rtabmap_ros/Tutorials/StereoOutdoorNavigation#Global_costmap : > Since the ground is not even, the ground is segmented by normal filtering: all points with normal in the...

The first file seems not available anymore. For the nano database, disabling `Grid/NoiseFilteringRadius` or `Grid/NoiseFilteringMinNeighbors` can make the grid maps denser. > We hardly get 4 to 5 FPS by...

We publish diagnostics about input and output of main nodes (rtabmap, odometry nodes), that could be used to investigate to know where the difference is coming from: ``` ros2 run...

Like the warning says, can you check the frame rate of these topics? ``` ros2 topic hz /camera/color/image_raw ros2 topic hz /camera/aligned_depth_to_color/image_raw ros2 topic hz /camera/color/camera_info ```

The images should be rectified. There is an example here: https://github.com/introlab/rtabmap_ros/blob/ros2/rtabmap_demos/launch/isaac/isaac_sim_vslam_demo.launch.py

I don't have example with 3D LiDAR on isaac simulator. Last time I tried it I was not satisfied with how they simulated a 3D LiDAR with 3 or 4...