Beat Quarantine with a Raspberry Pi Boxing Workout Tracker
This game uses computer vision to count punches.
Putting on a pair of gloves and throwing some punches can be a great way to work out indoors with a minimal amount of equipment, and if you need a little motivation to help you get going, check out this gamified system by Cisco. His element14 build tracks punches via an Intel RealSense camera using Python 3 and the OpenCV image processing library.
Cisco walks through the project's software setup in the video below, including installation on a Raspberry Pi 4 and an a overview of the blob/frame detection employed to sense when punches are thrown. To make things a little more interesting, a countdown timer challenges you to complete a certain number of jabs over a short period. He also adds a few MP3 sounds to provide encouragement, while an official Raspberry Pi 7” touchscreen is used for visual feedback.
The camera is mounted to the ceiling with a 3D-printed bracket. The Pi can run on a portable power supply for easy movement about the room, or wherever a quick workout is needed. Notably, the RealSense camera used for glove tracking is capable of depth detection, as well as pure visual information. This wasn’t incorporated here, but could potentially be implemented for even more advanced tracking applications.