Inspiration
In today’s digital age, people often spend long hours sitting at desks, whether for work or for school, and so often unknowingly adopt poor posture. Bad posture can lead to chronic pain, fatigue, reduced productivity, and long-term health issues. Existing solutions, like ergonomic chairs and wearable posture correctors, can be expensive or inconvenient. But we wanted something better.
What it does
This Posture Detection App uses computer vision to analyze posture in real-time and provide instant feedback. By using a webcam, the app detects head tilt, shoulder alignment, and slouching, helping users correct their posture immediately by sending them notifications when they aren't sitting straight. This way, when users start slacking on their posture, they instantly get reminded to correct it, meaning that they can hopefully more easily get rid of bad slouching habits.
How we built it
- We used Flask and OpenCV/Mediapipe in the back end to produce a responsive api.
- Frontend UI uses HTML
Challenges we ran into
Integrating the frontend and backend api was one of our challenges throughout the hackathon. Some of us are unfamiliar with endpoint calling and fortunately, we conquered this obstacle.
Accomplishments that we're proud of
- ✅ Real-time posture monitoring using computer vision (Mediapipe + OpenCV)
- ✅ Instant feedback on posture issues (head tilt, shoulder alignment, slouching)
- ✅ Easy calibration for personalized posture tracking
- ✅ Web-based interface for easy accessibility
What we learned
- Front-end, back-end, and API integration
- How to use Javascript/HTML/CSS to interact with the backend api
- Deep learning posture detection
What's next for Posture-X
- Keep track of user session times to give warnings
- AI learning of user habits
Log in or sign up for Devpost to join the conversation.