Open AccessDrowsy Driver Detection using MSER Feature Detection and Binarization on MATLAB
Author(s) -
Asma Jamesh
Publication year - 2021
Publication title -
aijr proceedings
Language(s) - English
Resource type - Conference proceedings
ISSN - 2582-3922
DOI - 10.21467/proceedings.114.31
Subject(s) - artificial intelligence , thresholding , computer vision , computer science , matlab , feature (linguistics) , rgb color model , haar like features , feature extraction , face detection , quadcopter , pattern recognition (psychology) , image (mathematics) , facial recognition system , engineering , linguistics , philosophy , aerospace engineering , operating system
Every year, 1.5 lakh people die in road mishaps in India. Among these accidents, 40% are due to ‘Drowsy or Sleep Driving’. According to several statistics, almost all commercial private drivers tend to drive continuously for 10 hours a day. Nearly all road accidents caused due to lack of sleep and drowsiness are highly hazardous and fatal. Drowsy Driver Detection Algorithm acquires real-time video and captures snapshots using an external Webcam. Using the Viola-Jones Algorithm, the face and the eyes of the driver are detected. The original RGB eye image is converted to a Gray image and then into a Binary image. Two techniques, Maximally Stable Extremal Regions Feature Detection and Binarization are deployed to determine the status of the driver. This research paper focuses on the development of a MATLAB algorithm to alert the driver or the co-passenger on-time by plotting the MSER Features and thresholding the acquired real-time images.