Automated and accurate traffic surveillance system
Mehmet Kaplan, Muhittin Gökmen · 2010
In this work, a traffic surveillance system, which contains background subtraction, occlusion handling and tracking steps, is designed in order to extract traffic parameters from low resolution video sequences. In every step, utilized algorithms are improved and more accurate results than existing approaches are obtained. Additionally, a new automatic Region of Interest detection approach is proposed to make the system fully automated.