Overview
Blink tracking for
healthier screen time.
Eye Guardian is a webcam-based desktop app that monitors eye comfort during long screen sessions. It tracks blink patterns, estimates dry-eye risk, and gives feedback for students, gamers, employees, and general users.
The refreshed version replaces the older hackathon pipeline with a cleaner MediaPipe Face Landmarker flow, adaptive blink thresholds, a polished Tkinter interface, and session logs saved for review.
2nd
REVA Hackathon Prize
4
User Modes
3
Sensitivity Levels
CSV
Session Logging
Problem → Approach
→ Results
01 — Problem
The Challenge
Long screen sessions can reduce natural blinking and make eye strain harder to notice until discomfort is already high. The project turns webcam signals into a lightweight awareness tool for personal screen-health habits.
02 — Approach
How It Was Built
A Tkinter desktop interface runs a threaded OpenCV webcam loop, feeds frames into a MediaPipe Face Landmarker detector, computes blink count, blink rate, eye openness, dry-eye risk, and keeps the UI responsive during timed sessions.
03 — Results
What It Achieved
The app gives immediate session feedback, supports manual stop or timer completion, and records comfort logs. It is clearly framed as wellness guidance, not a medical device, because lighting, glare, glasses, and camera quality can affect detection.
Screenshots &
Visuals
Add screenshot-1.png
01
Add screenshot-2.png
02
Add screenshot-3.png
03
Tech
Stack
Python
Language
Tkinter
Desktop UI
OpenCV
Camera Pipeline
MediaPipe
Face Landmarks
NumPy
Signals
Threading
Responsive UI
CSV Logs
Session History
MediaPipe Model
Vision Asset