JEOresearch/EyeTracker
A lightweight and robust Python eye tracker
What it solves
It provides a lightweight and robust way to track the 3D position of the pupil in video or image data, simplifying the process of detecting the pupil ellipse in eye-tracking applications.
How it works
The project uses a Python-based algorithm to process video frames or images. It identifies the pupil by calculating a rotated rectangle that represents the pupil ellipse. The system includes two versions: a standard robust version and a "lite" version optimized for efficiency but requiring better lighting conditions.
Who it’s for
Developers and researchers who need 3D eye tracking capabilities without heavy software overhead, particularly those building custom IR cameras or VR-related hardware.
Highlights
- Lightweight Python implementation
- Supports both robust and efficient "lite" tracking modes
- Compatible with video files, camera captures, or images
- Includes guidance on building low-cost IR camera hardware to support the tracking algorithm
Related
- Project
- Project
- Project
- Project
- Project