For the same topic , I’m using an NVIDIA JETSON TX2 and INTEL REAL sens camra D435 , I can estimate the depth. I’m planning to work on path planning and obstacles avoidance , I’m using RRT* path planner, But I don’t know how to exploit the camera information to make the path planning. I have been working on path planning without camera, so I give the obstacle position (X,Y,Z)…(Xn,Yn,Zn) to the path planner, : “I modeled the obstacle as a parallelepiped shape”: then the RRT* path planner generates the appropriate path without hitting the obstacles.
-The question is can I use the camera to locate the obstacle position? If yes, what position can the camera provide me? Because in the case of paralleleppiped I have the coordinates of all the obstacle segments?