| With the development of computer technologies, more and more video surveillance systems have been widely applied in production and life. Intelligent Transportation Monitoring System of freeway is one of the important applications, and is the trend of traffic development in the future. Moving vehicle detection based on video is the key to the Intelligent Transportation Monitoring System of freeway.The thesis researches some key technologies about moving vehicle detection of the Intelligent Transportation Monitoring System based on video. It regards traffic scene picture obtained by camera as the research object, and it is on the premise of rotatable camera.Considering the judgment of the camera rotation, the article analyzes the method of road edge extraction and the experimental results based on video sequence images. According to correlation coefficient of these road edge images extracted, the method of the camera rotation judgment is proposed, and the feasibility is verified. The problem is solved that whether the background need to be reconstructed. It is judged automatically by the computer system. In the detection of moving vehicles in static background, three methods, including'optical flow','background difference','differential between adjacent frames', are discussed. Through comparison and analysis, the method of background difference is mainly researched. Considering the background updating, different selective background updating algorithms are analyzed. One adaptive algorithm is adopted through comparison and experimental analysis. On the base of the existing shadow detection and removal algorithms, a little improvement about a shadow detection method is realized. The results show that most of the shadow can be detected and removed with this method. Then, accurate vehicle extraction is realized. At the same time of the research about moving vehicle detection, the flow chart of the moving vehicle detection system is devised and explained simply at last. |
PDF Full Text Request