my-first-project

An AI-powered gaze tracking system using MediaPipe and OpenCV to translate iris movement into directional commands for assistive robotics

This is my first major Computer Science project. It uses AI-driven computer vision to track a user's iris and translate eye movement into directional commands (LEFT, RIGHT, UP, DOWN).

The Purpose

Designed as the "Brain" for assistive technology, this software allows individuals with limited mobility to control a robot or a computer interface using only their eyes.

How it Works

1. MediaPipe Face Mesh: Detects 468 facial landmarks.
2. Iris Refinement: Specifically targets Landmark 468 (the pupil center).
3. Gaze Ratio Math: Calculates the relative position of the iris between the eye corners and eyelids.
4. Logic Thresholds: Translates these ratios into commands with a "deadzone" to prevent accidental movement.

Built With

• Python 3.10
• OpenCV (Computer Vision)
• MediaPipe (AI/Machine Learning)
• NumPy (Mathematical calculations