Vision Processing Process - Team5818/vision-2017 GitHub Wiki

In order to maximize efficiency on the Raspberry Pi, the vision server runs processing in a separate process. This process receives raw opencv2 images from the server process, and then runs an algorithm on them to pick out parts of the game. The resulting processed image is then handed back to the server for the camera stream. The positioning data of the parts is written out over serial to the roboRIO to be processed by the main robot code.

There is also a Server class that handles commands over HTTP, rather than the serial communication. It is not currently used by any part of the robot.