Object Detection and Path Generation

[vyang2968]

Purpose

Machine learning and deep learning has made huge strides in all aspects of life. In the world of autonomous vehicle control, object detection has paved the way for vehicles to make decision based on environmental input. We are looking to bring the same thing to FRC. Machine learning has been in its infancy in FRC, only providing bare minimum resources and slow models which are unpractical for application on a robot. By exploring machine learning, there can be great advancement in autonomous control and driver assistance.

Goals

• Train object detection model that is fast on edge devices
• Deploy to edge device and verify its ability and speed
• Write code to send processed information to robot
• Center on detected object with drivetrain
• Generate pathing to balls
• Run full autonomous
• Write documentation for replication