The number of road incidents have increased tremendously over past decade. The current methods to detect any incident are good but not efficient enough to detect the incidents on time which leads to higher travel times and inefficient use of transportation network. The aim of this project is to develop a robust and fast incident detection on a road. In the method explored, live images and videos are captured from the traffic cameras situated on various roads and highways. The images are used to detect congestion on the road and the videos are used to detect any stalled vehicle at the side of the road. The software also captures the time the incident started as well as the time it ended. A copy of the images and videos of detected incidents is stored separately for further analysis.