Inspiration

Do you want to run faster? Do you want to jump higher? For every aspiring athlete, biomechanics is the secret to improvement in performance. Weightlifting doesn't do it all -- a fast boat with no direction is lost. PROTRACC.CO revolutionizes the field of biomechanical analysis. No longer do you have to pay with an arm and a leg for some equipment, no longer do you have to wait impatiently for the gruesome instructions of some old coach. PROTRACC.CO lets you see where you have gone wrong, and decide for yourself.

What it does

Our project is a web-based platform that allows users to upload videos of themselves performing a movement, from a golf swing to a simple vertical jump. The platform then uses machine learning algorithms to track the user's joints and generate a 3D visualization of their movement. This visualization can be compared to ideal movements, and users can receive feedback on their technique and suggestions for improvement.

How we built it

The backend consisted of Mediapipe, OpenCV, and Flask to triangulate the Cartesian coordinates of the movements. The front end involved Three.js to visualize and interact with the 3D visualizations and React to create the webpages. We also used MongoDB to store user data and video files.

Startup In this market, PROTRACC.CO addresses a critical pain point: access to affordable, accurate, and actionable biomechanics analysis. Current solutions are often limited to high-end sports teams or elite athletes, with prices ranging from $5,000 to $50,000 per year. Meanwhile, individual athletes and fitness enthusiasts are left with limited options, relying on trial-and-error methods or expensive personal coaching. PROTRACC.CO democratizes biomechanics analysis, making it accessible to anyone with a smartphone and an internet connection.

What's next for PROTRACC.CO

Of course, there are other features to consider about the product as well. A sense of depth is difficult with a single view—Humans have two eyes for a reason. That is why the next move for PROTRACC.CO is to integrate a second camera view. In fact, this has already been in production -- by placing two cameras at perpendicular angles of the scene, the cameras can effectively triangulate the coordinates of the subject. From prototyping, we found that having such a feature skyrockets the accuracy of the model: there are now rarely blind spots, and the visualization seems to come to life. Furthermore, we aim to analyze and compare the user's biomechanics with the best athletes to reflect pointers that will work the best for them.

Built With

Share this project:

Updates