Road Incident Detection and Alert System for Immediate Assistance

In this paper, we have developed a road accident detection system using Image Processing techniques. The performance of three different models-CNN, Yolov8, and Dino DETR-was tested to determine which model provides the highest accuracy in detecting road accidents. The results of the experiments showed that the Dino DETR model outperformed both CNN and Yolov8, achieving an accuracy rate of 96%. This accuracy is approximately 7% higher than that of the CNN model and Yolov8, making it the most effective approach for detecting road accidents in this study. The findings of this study highlight the potential of using the Dino DETR model in road accident detection systems. With a detection accuracy of 96%, this model demonstrates significant promise for real-world applications, where quick and accurate accident detection is crucial for timely assistance and response. The successful implementation of such systems could significantly improve road safety, reduce accident-related fatalities and injuries, and enhance the efficiency of emergency response services. This research lays the groundwork for future developments in automated traffic monitoring systems that can assist in preventing road accidents and minimizing their impact.