a review of non-intrusive methods used to determine the driver drowsiness

driver drowsiness has been one of the main causes of on-road crashes which can lead to death, physical injuries and impose significant costs on the societies. the development of non-intrusive methods to be able to detect driver drowsiness in the early stages of drowsiness is great of importance. this is an educational review and its purpose is to provide recent achievements about non-intrusive techniques used to detect driver drowsiness. recently published related articles were searched in the scientific databases such as web of sciences, scoupous, pubmed and google scholar. by studying the articles and extracting the important information, non-intrusive drowsiness detection methods has classified in three distinct categories: vehicle based measures, behavioral measures and non-intrusive physiologic methods. each of mentioned categories has its own advantages and limitations. vehicle based methods are strongly influenced by the road geometry, whether condition and lighting. by tracking the facial expression of the driver, drowsiness can be detected. the main limitation of this method is lighting, because the cameras do not function well at night. however, physiological parameters such as electroencephalography are more reliable than vehicle and behavioral measures, the intrusive nature of these methods limit their applications. in summary, combination of the mentioned method can reliably detect the drowsiness of the driver and further studies about the efficiency of the mentioned techniques in real environments are required.