Detection of Drivers Drowsiness on Four-Wheeled Vehicles using the Haar Cascade Algorithm and Eye Aspect Ratio
One of the most common types of threats to four-wheeled vehicle drivers is microsleep. Microsleep is a condition in which a person's loss of attention or consciousness due to a state of fatigue or drowsiness. In general, microsleep lasts for a short duration, about a fraction of a second to a f...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Fakultas Ilmu Komputer UMI
2025-04-01
|
| Series: | Ilkom Jurnal Ilmiah |
| Subjects: | |
| Online Access: | https://jurnal.fikom.umi.ac.id/index.php/ILKOM/article/view/2362 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | One of the most common types of threats to four-wheeled vehicle drivers is microsleep. Microsleep is a condition in which a person's loss of attention or consciousness due to a state of fatigue or drowsiness. In general, microsleep lasts for a short duration, about a fraction of a second to a full 10 seconds. One way to modify the driver's sleepy condition is to form a drowsiness detection system through the extraction of facial feature points. The extraction of facial feature points refers to 68 predictor landmarks with detection in the eyes and facial movements of the driver in the form of poses with the determination of the angle threshold of changes in the position of the face while driving which indicates a state of drowsiness. This study implements the use of the Haar Cascade Classifier algorithm in detecting the drowsiness of four-wheeled vehicle drivers and the Eye Aspect Ratio of the points that form the eyes using Euclidean Distance. In detecting the eye index on the face predictor landmarks uses the dlib python library to detect objects, face detection, and face landmark detection. This study also uses the Face Detector library to create a face detector object and a Landmark Predictor. The test results showed that the detection system was 98.33% accurate with the condition of facial features that could still be identified by the system even though the difference in face distance with the webcam acquisition tool was far away. This detection system is also able to detect driver drowsiness with an average time duration of less than 5 seconds with a distance of up to 50 meters. The system detects drowsiness quickly with a notification in the form of a warning in the form of an alarm sound, which is very important in order to reduce the number of accidents due to drowsiness. |
|---|---|
| ISSN: | 2087-1716 2548-7779 |