Need Guidance: Pick-and-Place Application with OAK-D Camera and Isaac ROS

Robotics & Edge Computing Robotics - Isaac Isaac ROS
1

Hello,

I am new to Isaac ROS. I have an OAK-D 3D camera and an Elephant robot, and I want to implement a pick-and-place application by detecting objects with the camera.

Could someone please guide me step by step on how to achieve this?

Thank you!

2

Hi @hemaalatha.nk

If you are working with an OAK-D camera you can start from our Isaac ROS container:

  1. Install on the Isaac ROS container the OAK-D camera drivers GitHub - luxonis/depthai-ros: Official ROS Driver for DepthAI Sensors.
  2. Follow the documentation for Isaac ROS Image Segmentation — isaac_ros_docs documentation
  3. Remap the ros2 topic to the OAK-D camera publisher

At the end of the development you can build your own docker image following the guide:
https://nvidia-isaac-ros.github.io/v/release-3.1/concepts/docker_devenv/index.html#development-environment

Best,
Raffaello