Authors : Natramizh Saravanan; Vigneshwaran S.; Syed Roshan Abbas Kazmi; Dr. V. Saminadan

Volume/Issue : Volume 11 - 2026, Issue 4 - April

Google Scholar : https://tinyurl.com/mwv4s4ca

Scribd : https://tinyurl.com/3ryr2u6h

DOI : https://doi.org/10.38124/ijisrt/26apr1846

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : The rapid growth in vehicle usage due to increasing population has led to a significant rise in road accidents, making it a serious global concern. According to the World Health Organization (WHO), road accidents are among the leading causes of death worldwide, claiming millions of lives each year. Factors such as reckless driving, violation of traffic rules, increased congestion in urban areas, and distractions like mobile phone usage contribute heavily to these incidents. Additionally, delayed emergency response and lack of immediate medical assistance are major reasons for increased fatalities, highlighting the need for a faster and more efficient accident response system. To address this issue, an automated accident detection and rescue system based on deep learning is proposed. The system operates in two phases: the first phase involves detecting accidents using image preprocessing and a Convolutional Neural Network (CNN) with the ResNet50 algorithm, trained on a custom dataset created from online video sources due to limited dataset availability. In the second phase, once an accident is detected, an alert message is automatically sent to emergency services to initiate rescue operations. This approach eliminates the need for human intervention and ensures quicker response times, thereby improving the chances of saving lives. The use of ResNet50 enhances detection accuracy, making the system more reliable compared to traditional methods.

Keywords : Accident Detection, Deep Learning, Convolutional Neural Network (CNN), ResNet-50, Computer Vision, Traffic Surveillance, Automated Alert System, Transfer Learning, Internet of Things (IoT).

References :

1. Hadi Ghahremannezhad, Hang Shi, Chengjun Liu, "Real-Time Accident Detection in Traffic Surveillance Using Deep Learning“, IEEE International Conference on Imaging Systems and Techniques, ICIST 2022.
2. M. P. Rathod, D. P. Gadhiya, "Real-Time Accident Detection and Alert System using Deep Learning on Edge Devices",IEEE International Conference on Inventive Systems and Control(ICISC), 2020.
3. Bulbula Kumeda, Zhang Fengli, Ariyo Oluwasanmi,"Vehicle Accident And Traffic Classification Using Deep Convolutional Neural Networks",16th IEEE International Computer Conference on Wavelet Active Media Technology and Information Processing,ICCWAMTIP ,2020.
4. Lakshmy S, Renjith Gopan,et al, "Vehicle Accident Detection and Prevention using IoT and Deep Learning",IEEE International Conference on Signal Processing, Informatics, Communication and Energy Systems (SPICES),2022.
5. Nikhlesh Pathik, Rajeev Kumar Gupta, Yatendra Sahu,"AI Enabled Accident Detection and Alert System Using IoT and Deep Learning for Smart Cities", MDPI journal on sustainability,2022.
6. Mohammed Imran Basheer Ahmed, Rim Zaghdoud, Mohammed Salih Ahmed ,"A Real-Time Computer Vision Based Approach to Detection and Classification of Traffic Incidents", MDPI Journal on Big data and Cognitive Computing, 2023.
7. Renu, Durgesh Kumar Yadav , Iftisham Anjum ,Ankita, "Accident Detection using Deep Learning: A Brief Survey", International Journal of Electronics Communication and Computer Engineering ,Vol 11, Issue 3, ISSN : 2249–071X, 2020.
8. Ghosh, S., Sunny, S.J. and Roney, R ”Accident detection using Convolutional Neural Networks”, IEEE International Conference on Data Science and Communication (Icon DSC) pp. 1-6, 2019.