Abstract:
Intelligent Transportation system (ITS) takes a vitalpart in global world. It is a novel field that interoperates in different fields of transportation system. Traffic control is one of the important applications of ITS. As the problem of overpopulation worsens, traffic control becomes a grand challenge. Among the main technologies used to implement solutions for this challenge is the Video-based traffic surveillance system.Of late the focus of video-based traffic surveillance has shifted to detect incidents at Highways such as vehicle accidents, traffic congestion, and unexpected traffic blocks.
This project aims to detect accidents automatically by implementing a vision-based Automatic Accident Detection (AAD) Algorithm. After detecting and tracking vehicles from surveillance videos, centroids and bounding boxes are used to determine the occurrence of accident. In case of an accident, it will be reported immediately.
This research discusses the implementation of the proposed AAD algorithm using image processing and computer vision toolboxes of MATLAB. And also the results of offline testing.
As a result, AAD can detect and track vehicles and notify when accident occurs
