REAL-TIME DROWSINESS DIAGNOSTIC SYSTEM USING OPENCV ALGORITHM

Abstract

Drowsiness or fatigue is a major cause of road accidents and has a significant impact on road safety. There are many fatal accidents that can be avoided if drowsy drivers are warned early. There are a variety of sleep detection systems available that monitor drivers' drowsiness while driving and alert drivers if they are not focused on driving. Appropriate features can be extracted from facial expressions such as yawning, eye closing, and head movement to determine the level of sleepiness. The biological condition of the drivers' body, as well as the behavior of the vehicle, are analyzed to determine if the driver is drowsy. It presents a comprehensive analysis of the available mechanisms for the driver's drowsiness and presents a detailed analysis of the most commonly used classification strategies in this regard. We divide existing strategies into three categories: behaviors, physical, and strategies based on life parameters. Second, the supervised learning methods used for sleep apnea are being reviewed. Third, the pros and cons and comparative research of different approaches is discussed. In addition, the research frameworks are detailed in the diagrams for better understanding. Because of the dangers posed by road fatigue, researchers have developed various mechanisms to detect driver drowsiness and each procedure has its own benefits and limitations. In order to make an important review of Drowsiness Detection Techniques (DDT) and appropriate classification methods, we created a search engine unit to gather relevant information. We keep our search focused on publishing reputable journals and conferences. We have developed a multi-stage selection process and testing process.