Game Piece Identification (Vision)


Just wanted to share a project I’ve been working on. Its goal is to identify objects (this year’s game pieces) off of a camera feed.


Example Video:


Update: I’m working on the same thing but for cubes

We are also planning to implement game piece identification. We plan to train a YOLOv5 based detector.


YOLOv5 looks cool. Do you plan on running your vision code on the RIO or on a separate processor.

Video using cubes:

We have a Seeed Studio Grove Vision AI unit. We need time and tons of pictures of the game elements.

1 Like

Thank you for sharing. The seed camera looks very promising. I wonder if I can apply something similar to my code.