| Popis: |
There are four main factors causing the accident on the highway: the first is the sleepy driver, the second is the condition of the vehicle used, the third is the road factor, the fourth is the environmental factor. The human eye in a normal physical state has a different size, so that when a person who has large eyes when drowsy then the large opening of his eyes will be different from the size of the eyes of someone who has small eye size. Therefor in this research the system was designed by taking a unique eye reference per person. Preliminary data retrieval is taken when the driver starts driving a vehicle assuming the initial condition of the driver is not drowsy. This initial data will be used as a reference for threshold decision making in the process of determining drowsiness in subsequent experiments. In this research using the EAR value of each driver from the initialization process, and the EAR value is also used as a baseline for stating a driver’s eyes are closed or open based on the EAR threshold. If the EAR value (from driving process) is below the threshold for several consecutive times (with a reference of 1.5 seconds) the driver will be detected drowsy by the system. If detected drowsiness, it will be given a response in the form of motor vibration and sound. The drowsiness detection system has been designed using the proposed EAR threshold method from each ratio of eyes each individual driver. From this research get the EAR driver is 0.03-0.69. EARmin for each driver is in the range: 0.03-0.18 and EARmax for each driver in the range: 0.35-0.69. The minimum value of the EAR threshold is 0.20 and the maximum value of the EAR threshold is 0.41. From the four times the drowsiness detection system testing, the system successfully to detect drowsiness from each driver. |
