Driver Fatigue Detection Using Image Processing

The quantity of street mishaps that happen every day is rising and greater parts of them are credited to being the driver's deficiency. In many of these cases, a fault in their driving is attributed to fatiguelack of attention, drowsiness or outright dozing off while driving. This work proposes an observing framework that alarms the driver when he capitulates to sleep. The proposed calculation provides the live video feed concentrated on the driver's face and tracks his eye and mouth movements to identify eye closure and also the yawning rates using Haar Cascade classifiers. The driver is determined to be drowsy in two cases. The first is if the driver is found to be sleeping. The second is if the driver is found to be on the verge of sleeping which is determined if the driver yawns continuously. A buzzer is sounded if the driver is sluggish or effectively sleeping. The primary target of this work is to find a proficient methodology for distinguishing whether the driver is distracting from various objectives like drowsiness, etc. Specifically, the proposed method takes input as video using webcam present in the car. Using that it binarizes the picture and identifies whether a person is distracting or not. The proposed method gives an alarm when it detects the driver’s distraction.