| Title | Technology Development and Application for Dynamic Traffic Information (III) -- The assessment of applying image detection technologies to CCTV systems |
| Dept | Transportation Technology and Information Division |
| Year | 2010 |
| Month | |
| Price | |
| Summary | CCTV systems have been widely constructed through traffic control departments and spread in domestic arteries to monitor and manage traffic conditions. Under limited budgets and manpower, the integration of vision-based vehicle detectors and CCTV systems is beneficial for collecting traffic information automatically. In this project, the CCTV test images are captured from the traffic control centers in Taipei City, Taipei County, and Taoyuan County, and testified the detection results through the vehicle detection system developed in the previous project. In order to analyze the factors that may influence the accuracy rate of vehicle detection, including image quality, frame rate, altitude and angle of the installed CCTV camera, period, weather, and traffic flow, a statistical technique called analysis of variance (ANOVA) is applied to analyze the correlation of the factors. The experimental results have shown that image quality, frame rate, and illumination have a great influence on the accuracy rate of vehicle detection. Additionally, a dual-camera detection system combined with a license plate tracking system is proposed to identify the vehicles which violate traffic regulations. According to the experimental results, the dual camera system employed to enforce the laws is verified as feasible. Furthermore, the projects have shown that the level of automated traffic management is promoted to the level of automated traffic law enforcement. |
| Post date | 2011/04/01 |
| Hashtags |
View count:
110