Eye Aspect Ratio based on Histogram Oriented Gradient and Linear Support Vector Machine to Microsleep Detection

Traffic accidents are caused by several factors, especially due to driver fatigue. To minimize human negligence, developing a prototype microsleep detection system trigger an alarm is necessary. This research uses Histogram Oriented Gradient and Support Vector Machine methods detect objects. The programming language used Python, using the dataset from Idlib face landmark as marker in point area, which will then be calculated based on eye aspect ratio. implemented video input captured webcam real-time. output of this buzzer alert driver. In study, test results were obtained well, carried out with 2 scenarios distance 40cm - 100cm testing light levels 33 lux 226 lux. From these results, accuracy at 88% each.