Real Time Driver Drowsiness Detection Based on Convolution Neural Network

Main Article Content

M.A. Ahmed, Harith A. Hussein, Mohammed Basim Omar, Qabas A Hameed


Fatigue, in reality, is a severe threat on the road since it affects a driver's ability to react and process information. Fatigue is the primary cause of the rising frequency of road accidents. The power of these algorithms to accommodate variance in human face and lightning conditions is one of their biggest hurdles. We want to implement an advanced processing system that will dramatically minimize traffic accidents. We may use this method to determine face traits in drivers, such as the proportion of time their eyes are closed. We offer an effective and non-intrusive approach for detecting driver weariness in this paper using eye extraction. A webcam is used to observe the driver in this system continuously. Haar cascade classifiers are used to recognize the driver's face and eye. Eye pictures are collected and sent into a custom-built Convolutional Neural Network to determine if the left and right eyes are closed. The eye closure score is derived based on the classification. An alert will sound if the driver is judged asleep.

Article Details