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
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.
|Publication size in sheets||0.5|
|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 Classification||; ; ; ; ; ; ;|
|Score||= 5.0, 30-01-2020, MonographChapterAuthor|
|Publication indicators||= 0; = 0; : 2016 = 0.552|
* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.