The main objective of this project is to detect driver’s drowsiness and alert the driver which is an important precautionary measure in order to avoid accidents. Here two different algorithms based on Convolution Neural Network (CNN) were applied and the results were compared respectively. “Highway Hypnosis” is a serious issue to be addressed while driving especially on highways. Drivers who travel on highways continuously for more than 3 hours must be aware of this serious problem. If there is proper knowledge of it, fatalities would be drastically reduced. In this project, a dedicated detection coupled with an alarm system is provided to alert the driver in case of drowsiness. CNN is used since it is very effective in analyzing images and videos. In this project, a live video feed is used to detect drowsiness by suitable algorithms.

Background: Drowsiness is one of the underlying causes of driving accidents, which contribute, to many road fatalities annually. Although numerous methods have been developed to detect the level of drowsiness, tech-niques based on image processing are quicker and more accurate in comparison with the other methods. The aim of this study was to use image-processing techniques to detect the levels of drowsiness in a driving simula-tor. Methods: This study was conducted on five suburban drivers using a driving simulator based on virtual reality laboratory of Khaje-Nasir Toosi University of Technology in 2015 Tehran, Iran. The facial expressions, as well as location of the eyes, were detected by Violla-Jones algorithm. Criteria for detecting drivers’ levels of drowsi-ness by eyes tracking included eye blink duration blink frequency and PERCLOS that was used to confirm the results. Results: Eye closure duration and blink frequency have a direct ratio of drivers’ levels of drowsiness. The mean of squares of errors for data trained by the network and data into the network for testing, were 0. 0623 and 0. 0700, respectively. Meanwhile, the percentage of accuracy of detecting system was 93. Conclusion: The results showed several dynamic changes of the eyes during the periods of drowsiness. The present study proposes a fast and accurate method for detecting the levels of drivers’ drowsiness by considering the dynamic changes of the eyes.

Background: We determined the levels of subjective and observer drowsiness and facial dynamics changes. Methods: This experimental study was done in the virtual reality laboratory of Khaje-Nasir Toosi University of Tech-nology in 2015. Facial dynamics changes like changes in eyes, mouth and eyebrows were surveyed on twenty-five driv-ers in 2015byKSS (Karolinska Sleepiness Scale) and ORD (Observer Rating of Drowsiness). ANOVA Repeated Measure and MANOVA Repeated Measure tests were used for data analysis. Also, neural network and Viola-Jones were used to detect facial characteristics. PERCLOS (Percentage of Eye Closure), blink frequency and blink duration were inspected for eyes parameters. The size of open mouth during drowsiness was inspected for mouth parameter. During the inspection of eyebrow, the number 50 denoted eyebrow in normal position. For eyebrows above the nor-mal position, a range of 50 to 55 was specified; in addition, 45-50 was found as the specified range for eyebrows under normal position. Results: Descriptive statistics of the dynamic changes in mouth and eyes illustrated that during the driving process, the level of sleepiness increased as well as changes of eyes and mouth. However, statistical findings during car driving revealed that dynamic changes in eyebrows had clear expression with a constant trend. Similar studies on data ob-tained from KSS and ORD showed that both of these parameters simultaneously increased as well as the level of drowsiness. In addition, a significant relationship existed between facial expression and drowsiness. Conclusion: This research would be an effective and efficient tool for timely alarming and detecting the drowsiness quickly and precisely.