:::
Summary of IOT Publications
| Title | Application of Road Traffic Flow Analysis and Accident Detection |
|---|---|
| Dept | Transportation Safety Division |
| Year | 2019 |
| Month | 11 |
| Price | 350 |
| Summary | As traffic accidents are random, sudden, hard-to-foreccast events, an important task of the government is to formulate the strategies to reduce the number of traffic-related injuries and fatalities. For this reason, this issue is an urgent issue for the government. Analytical data from traffic safety prevention and improvement form the basis for this study. After analysis of the data and images and the rapifd development of Drivrless Driving and Unmanned Aerial Vehicle (UAV) and computer vision, it is possible to collect detailed traffic data of road intersections with complex traffic flow, which can record wide-area images and has advantages over the roadside CCTV cameras. For the demonstration of the characteristics of high-risk accident areas efficiently, four representative intersections in Taoyuan city are chosen for the data collection. We recorded the videos (1080p/4K) the intersections with a drone, and the videos were then analyzed with several algorithms including video stabilization, vehicle detection, vehicle tracking, vehicle classification. The positions, tracking, vehicle types, volume, speeds, etc. of vehicles in the traffic stream are computed. In a systemartic and integrated manner, a diagnostic and analytical tool was developed, which provides useful information such as vehicle tracking, traffic conflict hot zone with TTC and PET, and traffic characteristics aligning with the traffic signal plan of these intersections. This study uses basic traffic data obtained from UAV footage of intersections to diagnose traffic conflict issues and to estimate conflict risks. The results can be useful to the traffic management agencies and professionals as a reference tool for the improvement of traffic safety. |
| Post date | 2019-12-05 |
Count Views:625