I am looking for advice on the following topic: Lets say I have 1 GMSL-Camera connected to the DRIVE AGX. I want to capture images and broadcast them to several “Nodes” i.e. functions like lane-detection, object detection etc. I want to run each of these functions separately but processing the same images from the single camera. Which is the best and most efficient way with DRIVEWORKS to do this? Are the image-streamer samples the way to go?
At best, I would want the same functionality like ROS. One node caputures the images from the camera, broadcasts a topic and multiple other nodes subscribe to that topic and apply their functions to it. I tried implementing it with ROS, but it gets very inefficient since you have to use OpenCV etc. and cross-compiling things gets quite messy rather quickly.
I hope someone can give me some advice.