Posturemon

Build healthy habits by taking care of your own virtual pet, where your habits affect pets mood and growth!

Comment

Inspiration

We notice many people spend hours in front of a computer studying or coding. This is something we personally experience and have led to bad posture, neck pain, and fatigue. I personally spent many hours researching on how to fix my bad posture and notice that I always forget and has not been effective. We wanted to build something fun and engaging that helps users with this problem. Introducing our virtual companion that encourages better posture through AI-powered monitoring and gamified feedback! That’s how Posturemon was born... your posture buddy that grows stronger as you sit straighter!

What it does

Posturemon uses your webcam to analyze your sitting posture in real time. It: • Detects whether you’re maintaining good posture using MediaPipe Pose • Tracks your distance and alignment to the camera • Logs posture data to a local database for daily and weekly stats • Provides live feedback (“GOOD POSTURE” / “SLOUCHING”) • Encourages consistency with gamification elements, a digital pet that evolves as your posture improves

How we built it

• Frontend: Next.js (React), Tailwind CSS, and shadcn/ui for a clean interface
• Backend: Flask + OpenCV + MediaPipe Pose for real-time vision processing
• Database: SQLite for local data logging and analytics
• Architecture:
• Webcam → MediaPipe Pose landmarks
• Flask API → Posture analysis + database logging
• Frontend dashboard → Stats visualization & calibration control
• Tech stack highlights:
• Python threading for concurrent camera capture
• REST API endpoints for /api/posture/current, /api/stats/today, /api/stats/week
• Cross-platform with CORS for frontend integration

Challenges we ran into

• Handling real time pose tracking efficiently without frame drops
• Designing a non intrusive UI that runs smoothly alongside camera streams
• Ensuring consistent detection across different lighting and camera angles
• Calibrating posture in a way that’s both simple and accurate

Accomplishments that we're proud of

• Built a fully working real time posture detection system
• Integrated database logging and analytics for user insights
• Implemented a smooth API flow between camera and web dashboard
• Created a foundation for gamified health monitoring, opening the door to future AI companions

What we learned

• How to integrate computer vision models (MediaPipe + OpenCV) into a web app
• Building robust Flask APIs for real-time data streaming
• Structuring full-stack projects with frontend and backend communication
• The importance of UX in wellness tools, feedback needs to be encouraging, not overwhelming

What's next for Posturemon

• Introduce a virtual pet evolution system that grows with posture streaks
• Add leaderboards and daily quests for motivation
• Learn users habit to effectively suggest better habits such as procrastination
• Support for multi device syncing and cloud logging
• Eventually release as a desktop or mobile app (Electron / PWA)

Built With

Share this project:

Updates