HIRo Perfection Game

Fall 2021

Description

Project to have the Olin HIRo robotic arms play the children's game Perfection. The robot picks up individual shapes using a gripper, then classifies and finds the orientation of the piece. The pieces are detected using an Intersection-Over-Union algorithm demonstrated by the video below. We also attempted to classify shapes using Fourier Descriptors but found the IOU method to be less computationally intensive for this application. Finally, the robot follows a path to place each piece according to the detected piece.

https://olinrobotics.github.io/hiro/

Tools Used: ROS, Python, OpenCV, MoveIt!

Team

Position: Lead

# of Members: 3

Full Demo

Vision Demo