Automated Stand-alone Video-based Microsleep Detection System by using EAR Technique

Abstract

Microsleep while driving is the most event that contributes to the car accident. The proposed automated system deals with the development of system detection based on video of eye opening specifically for microsleep event among drivers. This system is used to help in avoidance of car accident that cause by microsleep among drivers while driving. Mainly, the objective of this proposed system is to detect the microsleep event while driving and to alert drivers because there is nobody can alert the driver with microsleep especially when they are driving alone. The proposed system is developed by using a python programming with image processing technique and Raspberry Pi module to make it applicable as stand-alone device. The EAR algorithm is the most importance in the proposed work as the microsleep detection made through eyes closure. Several conditions of eyes opening and closure, light and darkness, and with or without glasses/sunglasses have been tested. From the result, the proposed system shows that the ability of the proposed automated microsleep detection based on video imagery has showing a promising performance. However, the effectiveness of the eye detection must tolerate with the surrounding condition which also due to the type of camera used. Therefore, the camera is suggested to be able to function with night vision features such as infrared (IR) mode so it can be applied for daytime or night.