MLB SPINAL ROTATION TRACKING
A computer vision Python application to measure the angle of spinal rotation for professional baseball players
A major league baseball team needed to track spinal rotation for all of its players as a measurement of mobility and physical health. They asked our team of three to accurately track rotation and display the maximum rotation to the left and right achieved in each test. We decided to approach this using computer vision to track the players' movement, as it would be a minimal change to their current process and would provide accurate, real-time data. I created sketches like the one here to convey our ideas to the client and get feedback before we started fabrication.
As the lead on this project, I designed the system concept to incorporate a ceiling mounted camera, a Raspberry Pi, and a handheld touchscreen device to live stream the video and angle calculations. With the help of a mechanical engineering student and a computer science student, we designed each component of the system, using Fusion360 and GitHub to keep each other updated on our tasks. One of my primary focuses was the touchscreen device --interfacing it with our Raspberry Pi and designing an ergonomic enclosure out of laser cut ABS.
Another task of mine was to use the Raspberry Pi processing to take two points obtained from our computer vision code to calculate and output the current and max angles. I learned Python and Raspberry Pi basics in order to write this code, run our program off of the Raspberry Pi, and assist with and debug our computer vision code to optimally track the points we needed. This project is currently in the evaluation and integration phase, and will soon be shipped out to our client for use in their facilities.