Videonystagmography System
A clinical eye-tracking rig built from scratch: camera, optics, signal chain and the software that reads it.
2023 · Final-year engineering project, hardware and software
The problem
Videonystagmography diagnoses balance (vestibular) disorders by measuring involuntary eye movement, but commercial systems are expensive and closed. The project called for a complete VNG system, from the camera hardware through to the signal processing that turns raw eye footage into clinically useful measurements.
What I built
- Real-time pupil and glint detection pipeline with configurable filtering, contour analysis and template matching
- Homography-based gaze calibration mapping eye position to on-screen coordinates
- Clinical test suite: gaze, saccade and smooth-pursuit tests with accuracy measurement
- C++/Qt desktop application with live camera feed, calibration workflow and debug tooling
Built with C++, Qt, Signal Processing, Hardware Design, Medical Devices, Computer Vision.
How it landed
Placed second in the electronic engineering category of the final-year project awards.
Try it
A cut-down version runs in your browser on an ordinary webcam: follow a dot for ten seconds and see your eye’s path drawn against it. Nothing is uploaded.
Open the lab