Non-intrusive Driver Drowsiness Detection based on Face and Eye Tracking
Journal Title: International Journal of Advanced Computer Science & Applications - Year 2019, Vol 10, Issue 7
Abstract
The rate of annual road accidents attributed to drowsy driving are significantly high. Due to this, researchers have proposed several methods aimed at detecting drivers’ drowsiness. These methods include subjective, physiological, behavioral, vehicle-based, and hybrid methods. However, recent reports on road safety are still indicating drowsy driving as a major cause of road accidents. This is plausible because the current driver drowsiness detection (DDD) solutions are either intrusive or expensive, thus hindering their ubiquitous nature. This research serves to bridge this gap by providing a test-bed for achieving a non-intrusive and low-cost DDD solution. A behavioral DDD solution is proposed based on tracking the face and eye state of the driver. The aim is to make this research an inception to DDD pervasiveness. To achieve this, National Tsing Hua University (NTHU) Computer Vision Lab’s driver drowsiness detection video dataset was utilized. Several video and image processing operations were performed on the videos so as to detect the drivers’ eye state. From the eye states, three important drowsiness features were extracted: percentage of eyelid closure (PERCLOS), blink frequency (BF), and Maximum Closure Duration (MCD) of the eyes. These features were then fed as inputs into several machine learning models for drowsiness classification. Models from the K-nearest Neighbors (KNN), Support Vector Machine (SVM), Logistic Regression, and Artificial Neural Networks (ANN) machine learning algorithms were experimented. These models were evaluated by calculating their accuracy, sensitivity, specificity, miss rate, and false alarm rate values. Although these five metrics were evaluated, the focus was more on getting optimal accuracies and miss rates. The result shows that the best models were a KNN model when k = 31 and an ANN model that used an Adadelta optimizer with 3 hidden layer network of 3, 27, and 9 neurons respective. The KNN model obtained an accuracy of 72.25% with a miss rate of 16.67%, while the ANN model obtained 71.61% and 14.44% accuracy and miss rate respectively.
Authors and Affiliations
Ameen Aliu Bamidele, Kamilia Kamardin, Nur Syazarin Natasha Abd Aziz, Suriani Mohd Sam, Irfanuddin Shafi Ahmed, Azizul Azizan, Nurul Aini Bani, Hazilah Mad Kaidi
Keywords
Related Articles
- EP ID EP611492
- DOI 10.14569/IJACSA.2019.0100775
- Views 115
- Downloads 0
Mitigation of Cascading Failures with Link Weight Control
Cascading failures are crucial issues for the study of survivability and resilience of our infrastructures and have attracted much interest in complex networks research. In this paper, we study the overload-based cascadi...
Intelligent Irrigation Management System
It is widely known that water resources are decreasing around the world. Rapid urbanization, population growth, industries and the expansion of agriculture are increasing demand for freshwater. In most countries, includi...
Resource Management in Cloud Data Centers
Vast sums of big data is a consequence of the data from different diversity. Conventional data computational frameworks and platforms are incapable to compute complex big data sets and process it at a fast pace. Cloud da...
A Novel Design of Pilot Aided Channel Estimation for MIMO-CDMA System
In order to estimate a fading channel characteristics, a pilot signal is propogated with traffic channel. Fading channel parameter estimation is of paramount importance as it may be utilized to design different equalizat...
A Resource Recommendation Approach based on Co-Working History
Recommending the right resource to execute the next activity of a running process instance is of utmost importance for the overall performance of the business process, as well as the resource and for the whole organizati...