A Novel Omni-Directional Vision Sensing Technique for Traffic Surveillance

During the past decade, numerous research projects have been carried out in traffic surveillance. A variety of algorithms and techniques have been developed, primarily for observing vehicles from stationary rectilinear cameras mounted near roadways. Various applications include extracting traffic information such as average traffic speed, flow, and density. However, less research has been carried out in observing lane-level vehicle activity operating around specific vehicles in the traffic stream. In this research, we describe a unique orthogonal omni-directional vision system (OODVS) that has been developed to observe lane-level activity surrounding a vehicle, as well as the surrounding roadway geometry. This vision system uses a special catadioptric mirror providing a 360 degree orthogonal view of the environment. It is different from other catadioptric mirror-based omnidirectional vision systems in that it directly provides an orthogonal image without the need of warping a polar-coordinate based image to a perspective view. Based on this unique OODVS, a roadway traffic surveillance system has been designed and implemented. It consists of three major parts: 1) lane-level surrounding vehicle detection; 2) roadway lane information detection; and 3) localized traffic parameter estimation. Further, relationships can be established between the traffic parameters obtained from the probe vehicle with the embedded infrastructure sensors for complementary data fusion. Combined with a global positioning system (GPS) receiver that provides approximately 2-3 meters spatial resolution, this traffic surveillance system can be applied in several lane-level traffic applications such as detailed roadway geometry acquisition, and lane-level navigation tasks.