RESEARCH ON THE KEY TECHNOLOGIES IN COLLABORATIVE VIDEO BASED TRAFFIC FLOW DETECTION FOR SMALL AND MEDIUM SIZE CITIES IN CHINA
暂无分享,去创建一个
Current video based traffic flow detection equipments require a high position of collection and a shelter-free scenario, which has impeded the applications for these technologies in small and medium-sized cities in China. Realizing the facts of these cities, where there are narrower roads, fewer high buildings and more noticeable blocks on roadsides, this paper develops a novel collaborative video-based traffic flow detection approach. In the proposed method, to obtain the precise traffic flow data and overcome the limitations on site for video detection, video detectors monitoring the same road section are placed on different height and different angle with the centerline of the road section. Experimental data is collected on different types of road sections in Hohhot. Key technologies for the proposed method are discussed in-depth based on the statistic and comparative analysis conducted on the field data. It shows that the proposed approach can provide better precision within reasonable costs.
[1] Sharath Pankanti,et al. Appearance models for occlusion handling , 2006, Image Vis. Comput..
[2] Jitendra Malik,et al. Towards realtime visual based tracking in cluttered traffic scenes , 1994, Proceedings of the Intelligent Vehicles '94 Symposium.
[3] Rita Cucchiara,et al. Probabilistic people tracking for occlusion handling , 2004, ICPR 2004.