TY - GEN
T1 - Design and Development of Computer Vision-Based Driver Fatigue Detection and Alert System
AU - Aboagye, Isaac A.
AU - Owusu-Banahene, Wiafe
AU - Amexo, Kevin
AU - Boakye-Yiadom, Kwadwo A.
AU - Sowah, Robert A.
AU - Sowah, Nii Longdon
N1 - Publisher Copyright:
© 2021 IEEE.
PY - 2021
Y1 - 2021
N2 - Vehicle accidents are a common occurrence worldwide, large portions of which are fatigue-related. In this research paper, we proposed the design and development of a system to control fatigue-related accidents. The system comprises a microcontroller, a camera, and a speaker. The microcontroller receives a video stream from the camera and analyses the eyes and mouth of the driver to detect signs of fatigue. The detection of fatigue signs is accomplished using Haar Cascades. Haar cascades are machine learning object detection algorithms. They use Haar features to determine the likelihood of a particular point being part of an object. Boosting algorithms are used to produce a strong prediction out of a combination of 'weak' learners. Cascading classifiers are used to run boosting algorithms on different subsections of the input image received from the camera. The classifiers achieved high accuracy rates in detecting the various facial features with corresponding annotations. The system developed can detect fatigue with high accuracy. This paper recommends integrating computer vision-based fatigue detection and alert system into self-driving cars to automatically switch into autopilot when the driver continuously exhibits signs of fatigue.
AB - Vehicle accidents are a common occurrence worldwide, large portions of which are fatigue-related. In this research paper, we proposed the design and development of a system to control fatigue-related accidents. The system comprises a microcontroller, a camera, and a speaker. The microcontroller receives a video stream from the camera and analyses the eyes and mouth of the driver to detect signs of fatigue. The detection of fatigue signs is accomplished using Haar Cascades. Haar cascades are machine learning object detection algorithms. They use Haar features to determine the likelihood of a particular point being part of an object. Boosting algorithms are used to produce a strong prediction out of a combination of 'weak' learners. Cascading classifiers are used to run boosting algorithms on different subsections of the input image received from the camera. The classifiers achieved high accuracy rates in detecting the various facial features with corresponding annotations. The system developed can detect fatigue with high accuracy. This paper recommends integrating computer vision-based fatigue detection and alert system into self-driving cars to automatically switch into autopilot when the driver continuously exhibits signs of fatigue.
KW - boosting algorithms
KW - computer vision
KW - driver fatigue detection
KW - haar wavelet transforms
KW - mobile application
UR - http://www.scopus.com/inward/record.url?scp=85125415731&partnerID=8YFLogxK
U2 - 10.1109/ICAST52759.2021.9681943
DO - 10.1109/ICAST52759.2021.9681943
M3 - Conference contribution
AN - SCOPUS:85125415731
T3 - IEEE International Conference on Adaptive Science and Technology, ICAST
BT - Proceedings of the 2021 IEEE 8th International Conference on Adaptive Science and Technology, ICAST 2021
PB - IEEE Computer Society
T2 - 8th IEEE International Conference on Adaptive Science and Technology, ICAST 2021
Y2 - 25 November 2021 through 26 November 2021
ER -