In modern traffic surveillance, computer vision methods have commonly been utiliezed to detect vehicles because of the rich information content contained in an image. And detection and tracking of moving vehicle in traffic environment is one of the most important components in intelligent transportation system (ITS). The adaptive background modeling method was used to eliminate the negative effects from moving vehicle and rebuild the background images. The moving vehicles were segmented with difference images between background and current images. To suppress noise caused by segmentation and improve robust performance of vehicle detection, a template with 3-by-3 window was utilized to decrease isolated noise points around vehicle contours. Then, the morphological filtering, including erosion and dilation operation, was also applied, which minimizes the influence of the discontinuous block noise. Finally, to reduce the searching scope of vehicle detected and improve the calculation speed, Kalman filter model was performed to track motive vehicles. The experimental results verifid the effectiveness and real-time of algorithm.
School of Civil Engineering and Transportation, South China University of Technology, 510640 Guangdong, China;Xinjiang Technical Institute of Physics and Chemistry, Chinese Academy of Sciences, Urumqi 830011, China
You, Feng,Zhang, Ronghui,Wen, Huiying,et al. An algorithm for moving vehicle detection and tracking based on adaptive background modeling[J]. Journal of Theoretical and Applied Information Technology,2012,45(2):480-485.