Are there any RealSense camera tutorials for NVBLOX (C++)?

Robotics & Edge Computing Robotics - Isaac Isaac ROS
1

The tutorial I used:

https://nvidia-isaac.github.io/nvblox/pages/core_library_run_an_example.html

My machine configuration is:

  1. Jetson Orin; Ubuntu 22.04; Architecture: AArch64 (ARM 64); R36-4.3 (Jetpack 6.2); NVIDIA driver version: 540.4.0; CUDA version: 12.6; NVblox C++ Docker version installed

  2. Ubuntu 22.04; RTX 5090; Driver Version: 580.95.05; CUDA Version: 13.0; Using isacc_ros_nvblox-release:3.2 Docker version;

My questions:

  1. I see RealSense mapping tutorials only in nvblox_pytorch, but not in the nvblox (C++) version. How can I use RealSense cameras for mapping in the C++ version of nvblox?

  2. The demo uses color, depth, and pose data. How can I collect this data separately?

  3. The maps built by nvblox are in mesh.ply format. How can I put this into isaac_ros for vslam + nvblox + nav2 navigation?

2

Hi @fhqjwyk ,

The documentation you linked to is for the standalone Nvblox core library, but it sounds like you want to use the Isaac ROS Nvblox package. Isaac ROS Nvblox is a ROS 2 node that wraps around the Nvblox core library.

You can find the documentation for Isaac ROS Nvblox here.

We have a tutorial for using RealSense with Isaac ROS Nvblox here.

As you can see in the API details, the Isaac ROS Nvblox node publishes all the color, depth, and pose information on separate topics for your consumption.

3

The demo uses color, depth, and pose data. How can I collect this data separately? If I only have the nvblox（c++） environment