[zucker42]

Teams in the FIRST robotics competition community use the Raspberry Pi for computer vision. There's an open source program called Photonvision that provides an easy way for talented high school students to set up CV. Some of the better teams use Orange Pis or mini PCs for this purpose (and some of the best teams have fully custom solutions), but a Raspberry Pi is a lot easier to set up and is cheaper (even at $60). Also, my understanding is that the architecture of the RPi4 CPU/GPU allows for GPU acceleration of classical CV algorithms without copying (and that feature is integrated into Photonvision), so this hardware upgrade will probably make the Raspberry PI perform on par or better than more expensive options, while being easier to setup.

And two MIPI camera ports is also really nice for this application, because the new advancement in FRC is to use April Tags included with the field for full field localization. For this purpose, I've heard you get better results with global shutter MIPI cameras than USB webcams, and having multiple cameras is good for always seeing an AprilTag.