Hello, I have a robot arm that needs to grasp an object and place it. I have completed ROS Manipulation Course in 5 Days but unfortunately, doing grasping with perception wasn’t included in the course. I have generated an octomap for collision avoidance using depth camera. However, I need to exclude the detected object from the octomap in order to make the arm grasp it. In other words, I want the robot to differentiate between the detected object and the obstacles.
Have a look at this course:
It has manipulation of a cube based on perception among many other things
Thank you but this course is for Enterprise only so I can’t access it. Do you know any other resources?
Hello @moh.nabil ,
Have you reviewed the Grasping Unit of the Manipulation course? Although it is not explained in detail, what you mention it’s actually being done there (grasping with perception). We will include a more detailed explanation of this process in the near future though.
Actually, I have read that section in the grasping unit and my depth camera has a node that publishes the pose of the object but the idea is the robot won’t go to this pose as it’s located in the octomap. For more explanation, I subscribe the depth points to form the octomap so my robot sees the object as an obstacle. What I want to do is, to exclude this object from the octomap so that the robot can go to the pose of the object.