Video sequence analysis using local binary patterns for the detection of driver fatigue symptoms

Andrzej Majkowski , Marcin Kołodziej , Dariusz Sawicki , Paweł Tarnowski , Remigiusz Rak , Adam Kukiełka

Abstract

Fatigue is one of the important causes of car accidents. Analysis of falling asleep while driving helps to explain many of the most tragic events. To minimize the tragic consequences of falling asleep at the wheel, solutions were developed that allow early detection of fatigue symptoms. The article presents a system for the detection of symptoms of driver fatigue, based on an analysis of an image recorded by a camera. Two symptoms, that is slow blinking and yawning, are detected. To detect the symptoms of fatigue, cascade classifiers based on Local Binary Patterns were implemented. The classifiers were trained with the use of the OpenCV library. The system was tested on a collection of movies taken from the YawDD database. Conducted tests confirmed the correctness of the developed method. The impact of external factors, that could affect the effectiveness of the solution, was also analyzed. The system is able to correctly detect fatigue symptoms with an average accuracy of 99%. This result is comparable to the best published solutions.

Author Andrzej Majkowski (FoEE / ITEEMIS)
Andrzej Majkowski,,
- The Institute of the Theory of Electrical Engineering, Measurement and Information Systems
, Marcin Kołodziej (FoEE / ITEEMIS)
Marcin Kołodziej,,
- The Institute of the Theory of Electrical Engineering, Measurement and Information Systems
, Dariusz Sawicki (FoEE / ITEEMIS)
Dariusz Sawicki,,
- The Institute of the Theory of Electrical Engineering, Measurement and Information Systems
, Paweł Tarnowski (FoEE / ITEEMIS)
Paweł Tarnowski,,
- The Institute of the Theory of Electrical Engineering, Measurement and Information Systems
, Remigiusz Rak (FoEE / ITEEMIS)
Remigiusz Rak,,
- The Institute of the Theory of Electrical Engineering, Measurement and Information Systems
, Adam Kukiełka (WUT)
Adam Kukiełka,,
- Warsaw University of Technology
Pages48-57
Publication size in sheets0.5
Book Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Lecture Notes in Computational Vision and Biomechanics, 2019, ISBN 9783030209148, 48-57 p.
ASJC Classification1702 Artificial Intelligence; 1706 Computer Science Applications; 1707 Computer Vision and Pattern Recognition; 2210 Mechanical Engineering; 2204 Biomedical Engineering; 1711 Signal Processing; 1700 General Computer Science; 2614 Theoretical Computer Science
DOIDOI:10.1007/978-3-030-20915-5_5
Languageen angielski
Score (nominal)5
Score sourcepublisherList
ScoreMinisterial score = 5.0, 07-01-2020, MonographChapterAuthor
Publication indicators Scopus Citations = 0; WoS Citations = 0; Scopus SNIP (Source Normalised Impact per Paper): 2016 = 0.552
Citation count*
Cite
Share Share

Get link to the record


* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.
Back
Confirmation
Are you sure?