This project was completed over three months as part of a subject requirement at Swinburne University of Technology.

Driver fatigue causes up to 20% of fatal road accidents in Victoria, resulting in over $27.1 billion in losses across Australia. New regulations mandate safety features in new car models. However, the expanding used car market—driven by the cost-of-living crisis—often lacks these features due to vehicle age and lenient regulations.

The Fatigue Tracking System, based on MATLAB Computer Vision, was created to demonstrate the capability of a single 720p RGB camera and the benefits of machine learning algorithms. While other implementations were explored, such as IR-based cameras similar to Microsoft's Kinect and Apple's TrueDepth camera, the RGB implementation ensures accessibility for all at a low cost—because safety shouldn't be a luxury. The system tracks potential fatigue through a series of calculations based on the ratio of the eyes and mouth, mitigating issues related to distance.

The low-cost hardware has the potential for seamless integration into car dashboard units, reducing the number of parts needed and enhancing overall maintenance while providing an increased level of safety.