ACCIDENT DETECTION AND NOTIFICATION SYSTEM USING DEEP LEARNING TECHNIQUE
DOI:
https://doi.org/10.14456/lsej.2023.26Keywords:
Road accident detection system, Road accident notification system, Deep learningAbstract
This study aimed to develop an accident detection and notification system for road accidents related to vulnerable groups on the road, such as motorcyclists, who account for 80% of all road fatalities in Thailand. YOLOv5 is used to develop a system where objects in the image can be detected, whether a person or a motorcycle involved in an accident. The comparison of accident detection results obtained with different YOLOv5 models led to the selection of the most suitable model. Then, the notification system was developed in Python language along with LINE Notify API (Line Notify Application Programming Interface) for sending images and notifications to groups of people through the Line application when the system detects a road accident. The results show that YOLOv5x has the best performance in accident detection with 93.21% compared to the results of other models. Moreover, the developed system was 100 percent successful in sending images along with a short message indicating the number of motorcycles and the number of people involved in the accident. This intelligent accident detection and notification system can detect accidents immediately by alerting rescue workers, police officers or other parties. The information about road accidents, including images or messages, can help evaluate the situation and promptly prepare the rescue team and the necessary equipment, leading to immediate assistance to accident victims.
References
Guendalina C, Sardar J, Kenneth MG. A literature survey of recent advances in chatbot. Information 2022;13(1):41.
IBM Cloud Education. What is artificial intelligence (AI). 2020. Available at: https://www.ibm.com/cloud/learn/what-is-artificial-intelligence. Accessed March 1, 2022.
Morera A, Sanchez A, Moreno BA, Sappa DA, and Velez J F. SSD vs. YOLO for Detection of outdoor urban advertising panels under multiple variabilities. Sensors 2020;20(16):4587.
Rendi N, Mohammad I. Comparative analysis of deep learning models for vehicle detection. Journal of Systems Engineering and Information Technology 2022;1(1):27-32.
Simon K. Digital thailand. 2023. Available at: https://datareportal.com/reports/digital-2023-thailand. Accessed March 3, 2023.
Tangsiri. Announcing the number of users reaching 50 million, the 10th anniversary of LINE Thailand. 2021. Available at: https://brandinside.asia/line-thailand-10-yrs. Accessed March 2, 2022.
Thai Road Safety Culture. Road Accident Statistics in Thailand. 2023. Available at: https://www.thairsc. com/data-compare. Accessed March 2, 2023.
Thanadkit C. 10 years of Line Thailand. 2021. Available at: https://thestandard.co/line-thailand-user-and-new-features. Accessed March 1, 2022.
Thanaphong J. Road accidents "the silent danger" that are still dangerous and challenging management. Journal of Emergency Medicine of Thailand 2021;1(1):71-76.
Upesh N, Hossein E. Comparing YOLOv3, YOLOv4 and YOLOv5 for autonomous landing spot detection in Faulty UAVs. Sensors 2022;22(2):1-15.
Wang Z, Wu Y, Yang L, Thirunavukarasu A, Evison C, Zhao Y. Fast personal protective equipment detection for real construction sites using deep learning approaches. Sensors 2021;21(10):1-22.
Winston C, Tejas S. Exploring low-light object detection techniques. 2021.Available at: https://arxiv.org/ pdf/2107.14382.pdf. Accessed March 12, 2022.
World Health Organization. Road traffic injuries. 2022. Available at: https://www.who.int/news-room/fact-sheets/detail/road-traffic-injuries. Accessed March 1, 2022.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2023 Life Sciences and Environment Journal
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Each article is copyrighted © by its author(s) and is published under license from the author(s).
0-5526-7000 Ext. 7214
