I need to get 2 perspectives on a fast moving object for 3D mapping. I’m hoping for ~100FPS. And would using OpenCV be a possible bottleneck for this task?
Should be fine for the camera software stack but can’t promise if for the OpenCV post processing.
this also depends-on your sensor capability, you may parse your sensor output frame-rate as below.
$ v4l2-ctl -d /dev/video0 --list-formats-ext