MACHINE LEARNING BASED FEATURE EXTRACTION METHODS FOR ESTIMATION OF DRIVER'S DROWSINESS
Journal of Contemporary Issues in Business and Government,
2021, Volume 27, Issue 3, Pages 1671-1681
AbstractAt present, the number of vehicles in the cities are increasing rapidly and also which may cause accidents, so the prevention of accidents is a major challenge. According to autonomous vehicle technology and development, machine learning methods have been used to detect the driver‟s condition to improve the safety of the passengers and commuters in the road. Apart from the basic characteristics such as age, gender, driving experience, driver‟s previous accident history of the driver, driver's condition can be identified by considering the factors such as driver's facial expressions, eye blinking, head movements, usage of cell phone while driving, alcohol and accident sensor. Recent technologies such as video processing, image processing, and analysis using machine learning algorithms could be used to capture the constant images and videos of the driver to detect the behavior and to calculate the level of drowsiness. For example, a driver driving a long distance may feel tired and can be warned to take rest by giving a warning signal by an audio alarm. Thus, this work uses machine learning-based feature extraction methods for determining the drowsiness level of a driver.
- Article View: 17
- PDF Download: 33