Soccer Formation Analysis With Computer Vision

Inspiration

We are all Soccer fans

What it does

It annotates an ariel image of a soccer match to show the positions of players in a team

How we built it

We used an hugging face computer vision model which runs on YoloV8 to detect players, then used a K-means test to classify players into teams based on color jersey, then return an annotated image. The user interacts with a website where they upload an image, then our model works in the backend

Challenges we ran into

Finding an Algorithm to classify players into teams by color Getting the frontend and backend to integrate properly and display annotated images

Accomplishments that we're proud of

Creating a full stack project Creating a project that uses computer vision

What we learned

how to create a functional frontend and backend How utilize and enhance computer vision APIs in a real-world project

What's next for Soccer Formation Analysis With Computer Vision

Detect boundaries of field to get player position data relative to the field Extend our single-image tracking to track players over time in videos

Built With

• css
• javascript
• jsx
• python
• typescript