Oge Marques, PI; Borko Furht, Co-PI
Drowsiness causes 7% of all road accidents and mat to blame 18% of fatal collisions.
In this project we are developing a computer vision-based solution to the problem of detecting driver drowsiness based on visual input (images of driver’s face/head) and issuing an alert. We have researched the state-of-the-art in the field, with emphasis on computer vision aspects, such as algorithms and datasets for feature extraction and image classification for related problems
(e.g., open/closed eye classification and head pose estimation).
Our main objective is to implement a mobile prototype (running on a smartphone or comparable device, to be mounted on the dashboard of a car) that is capable of demonstrating how the proposed method achieves its goals in real-time.