Use of Gis Methodology for Online Urban Traffic Monitoring

Main goal of traffic management is the optimisation of traffic flow. This aim is based on possession of precise traffic describing data and parameters. Two dimensional traffic data acquisition becomes a research emphasis of German Aerospace Center (DLR)s Institute of Transport Research to improve these data basis. Data sources are ground based like floating car data as well as airborne remote sensing data and the developed procedures are appropriate to support traffic management effectively. A preliminary real time system for airborne traffic monitoring (ATMS) was already successful proved to collect traffic data on demand. The whole chain of the system extends from the sensor technology including a stabilized platform and image processing, via the data transmission to the ground to a traffic center for the further processing of images in which the information will be refined with prognosis tools. An open interface ensures a multi-facetted utilization of the information by multiple user groups. The data processing is based on a commercially available digital map. Practical tests showed that the quality of traffic flow data extraction will increase significantly by improving the digital map basis. The aim of the work presented here is to use heterogeneous image data, especially high resolution satellite data (Quickbird, Ikonos) to extend the a priori knowledge for implementation into the monitoring system. In parallel the image data from ATMS can be evaluated in focus of problems of description and assessment of the urban space. Consequently a GIS platform seems to be an adequate tool to implement, process, analyse and visualise the actual geodata. Special GIS data products are available for simulations or analysis and merge of sociodemographic and socioeconomic structure data (people and pixels),and the methodology can be adopted for for real-time applications like ATMS. Additionally the GIS System offers the opportunity to implement and manage the historical traffic flow data generated and derived from all data collecting systems (terrestrial, airborne and spaceborne sensors). In front of this background the paper will present the analysis of existing Satellite and ATMS data for the region of Berlin. As a first step the application of a vegetation detection algorithm will be shown. This results in an information layer concerning structures of trees at or near street areas and will eliminate an essential disturbance factor at real time airborne data collection and will therefore enhance the validity of the derived data. Since at the moment the digital map will not account for parking areas, high resolution remote sensing data will be analysed to extract accordant information layer for future applications. The implementation of the extended a priory knowledge into both the ATMS as well as the GIS will be presented and the improvements will be demonstrated.