Skip to Main Content
The most common measurement of usage in road network is vehicular traffic flow, which is defined as the number of vehicles passing a specific location along a road during a unit of time. Flow is often expressed as AADT (the annual average daily traffic). Traditionally, the traffic vehicle flow is measured by using ground-based sensors, most common of which are inductive loop detectors and pneumatic tube detectors. The sampling rate of a detector depends on the application and the operating agency, typically ranging from under one minute to a full day. This paper presents methods for spatial measures and interpretation of roadway network from aerial remotely sensed image. First, this paper present the interpretation of roadway network using co-registration between up-to-date aerial imagery and geospatial database, which is called image-to-GIS co-registration. The up-to-date aerial imagery and their extracted road network features are thought as reference data, and the early GIS data are co-registered onto aerial images after orthorectified in order to interpret the remotely sensed image; (2) the vehicles are detected based on the co-registered data; (3) Based on the co-registered data, and recognized vehicle, the measurement including vehicle density, headway are conducted, further to compute the flow rate; (4) the experimental results are compared with the results from the ground-based detection. Finally, some conclusions are drawn up.