Estimation of traffic parameters in urban areas from satellite images

In this paper an approach for automatic vehicle detection from satelliteimagery is presented. In space-borne images single vehicles havea very limited size and a detection scheme which only focuses onsingle vehicles would deliver ambiguous results. Therefore we extractvehicle queues because their repetitive pattern improves the distinctionfrom similar objects. Additionally information about the locationand approximately size of roads are used to limit the search spaceand thus decrease the number of false detection. Single vehiclescan be determined by analysis of the width and contrast behaviorof the queues. The results show that the analysis of width and contrastinformation extracts vehicles with high reliability, but since onlya certain number of vehicles appear grouped, the overall completenessis not sufficient yet. The used data sets are Quick Bird images coveringthe inner city of Munich (Germany) with a ground sampling distanceof approximately 0.6 m for the panchromatic channel.     «

In this paper an approach for automatic vehicle detection from satelliteimagery is presented. In space-borne images single vehicles havea very limited size and a detection scheme which only focuses onsingle vehicles would deliver ambiguous results. Therefore we extractvehicle queues because their repetitive pattern improves the distinctionfrom similar objects. Additionally information about the locationand approximately size of roads are used to limit the search spaceand thus decrease the number...     »