Hands of Steel (Team 33)

GamePlay Image

Inspiration

Currently, Rutansh Pathak has had some trouble as the Irvington High School goalie. In pursuit of improvement, he needed a fun and innovative way to keep practicing his skills.

What it does

This game uses the MediaPipe package in Python to detect landmarks on the hand to detect its motion and position in real time. Using this data, we transposed this data to gloves on a screen which will track and stop the ball.

How we built it

We installed the packages and then used them to design a hand tracking algorithm. The MediaPipe package ran in conjunction with the OpenCV package and it was particularly difficult to use as a video format, or rather live stream format. Though the landmarks were able to show up properly, the landmark's coordinates were particularly difficult to work with. We were able to create a rectangular bounding box that served as a colliderect for the gloves in the game.

Challenges we ran into

We ran into a handful of challenges. To name a few, the packages were incompatible with our version of Python, the package had confusing documentation, the rectangle's coordinates had a depth, the rectangle had inverted movement and pygame was struggling to run in conjunction with mediapipe.

Accomplishments that we're proud of

We had some amazing breakthroughs, including everytime we got past a hurdle in our project. Getting past the rectangular bounding box hurdle was one of our best achievements. Overall, this project was a great achievement and learning experience.

What we learned

We learned how hand tracking and tracking systems in general work. We learned how to apply them to a fun app that has real life applicability. We also learned how to make a harder project in Python.