Skip to content

Computer Vision

WebCoder49 edited this page Sep 16, 2023 · 2 revisions

👁 Computer Vision

In the autonomous period, computer vision, or identifying and locating April Tags / game pieces using a webcam, can be a complex but valuable method of play. Here are some suggestions both from us and FIRST.

  • Attach the webcam using this link. The "Sample Op Modes" are at the link below.
  • Debug your Computer Vision OpModes on the Driver Station:

image

  • Under any circumstances, start with the sample OpModes. They provide quick ways of identifying April Tags and locating game pieces (with the TensorFlow Object Detection library).
  • Use TensorFlow Object Detection and April Tags at first, then if you really want a challenge and have time after this, try training your own TensorFlow Object Detection model using the FIRST Machine Learning Toolchain on the team's custom Team Prop. For FTC-ML to work, you will need a login - ask the lead teacher, Mr. S.
  • TensorFlow Object Detection on game pieces hasn't worked too well in the past, but April Tags are quite easy to detect.

Clone this wiki locally