Classification of Walking Cars and Pedestrians for the use of Automatic Incident Detection as an Effort to Reduce Risk of Accidents on the Highway
Abstract
Traffic problems become very important to minimize the number of accidents. Recorded in 2017, the death toll from accidents reached 703 people. While in 2018 503 people died or fell by 28%. This figure is considered to be the third largest killer, under coronary heart disease and tuberculosis / tuberculosis. Among several causes of accidents such as against the flow of traffic, stops on the road, pedestrians and speeds that are too low compared to other vehicles. Even though the traffic signs are already installed. The low level of awareness of road users will increase the number of accidents. A detection system for potential accidents is needed to reduce the risk and can be used for the investigation process if an accident occurs. The application of a traffic accident prediction system will be a solution to provide a warning of potential accidents. Early detection of incidents is very important to limit consequences such as delays for other road users, lower costs, less time commitment to emergency services, as well as to prevent accidents. Video processing obtained from CCTV installed at intersections, highways, bridges and tunnels will detect pedestrians and oncoming cars automatically. Detection is done by processing each video frame to determine the foreground by the Gaussian mixture models method of each video frame.
Downloads
References
N. Qodar, “Polri: Angka Kecelakaan Lalu Lintas Menurun pada 2018,” Liputan6, 2018. [Online]. Available: https://www.liputan6.com/news/read/3407664/polri-angka-kecelakaan-lalu-lintas-menurun-pada-2018. [Accessed: 23-Aug-2018].
E. T. Pujiastutie, E. D. Antoro, and J. T. Sipil, “Karakteristik kecelakaan dan solusi penanganan untuk mengurangi angka kecelakaan di kota bengkulu,” vol. 6, 2015.
Guiyan Jiang ; Shifeng Niu ; Qi Li ; Ande Chang ; Hui Jiang, “Automated incident detection algorithms for urban expressway,” 2010 2nd Int. Conf. Adv. Comput. Control, 2010.
K. P. ; S. Chivapreecha, “Comparative Study of Threshold Selection for Incident Detection based on California Algorithm,” 2019 16th Int. Conf. Electr. Eng. Comput. Telecommun. Inf. Technol., pp. 911–914, 2019.
K. van Vianen, “Automatic Incident Detection ( AID ),” Pattern Recognit., 2017.
Z. C. ; H. H. ; W. W. ; X. M. ; X. Hu, “Detecting Gathering Incident of Video Surveillance Based on Plane Geometry,” 2010 Int. Conf. Mach. Vis. Human-machine Interface, 2010.
Otilia Popescu ; Sarwar Sha-Mohammad ; Hussein Abdel-Wahab ; Dimitrie C. Popescu ; Samy El-Tawab, “Automatic Incident Detection in Intelligent Transportation Systems Using Aggregation of Traffic Parameters Collected Through V2I Communications,” IEEE Intell. Transp. Syst. Mag., vol. 9, no. 2, 2017.
S. G. Anuradha ; K. Karibasappa ; B. Eswar Reddy, “Morphological change detection system for real time traffic analysis,” 2015 IEEE Int. Conf. Comput. Graph. Vis. Inf. Secur., 2015.
T. Kato, H. Yamamoto, N. Miura, and K. Setoyama, “Application of the Automatic Incident Detection ( AID ) system to the open section on highway in Japan,” 2011.
“Gaussian Mixture Models Explained - Towards Data Science.” [Online]. Available: https://towardsdatascience.com/gaussian-mixture-models-explained-6986aaf5a95. [Accessed: 15-Dec-2019].
