Road Map Generation and Extraction of Road Attributes Based on Vehicle Positioning Data

Abstract

Digital road maps are the basis of a lot of road traffic related applications such as vehicle navigation systems, car routing software, simulation tools and infrastructure assessments. Existing technology of producing digital road maps is very expansive and time consuming. This paper describes a new, low cost method to derive basic digital road maps and several street characteristics from vehicle positioning data of a taxi fleet. The data were collected exclusively for fleet disposition purposes using conventional Global Positioning System (GPS) receivers and were not improved with dead reckoning or differntial GPS techniques. the results demonstrate the potential of probe vehicle data analysis for automatic road net generation and attribute extaction possessing a considerable large sample with a good coverage of the region of interest.