Evaluation of Traffic Monitoring based on Spatio-Temporal Co-Registration of SAR Data and Optical Image Sequences

Abstract

This article presents an approach for spatio-temporal co-registration of dynamic objects in Synthetic Aperture Radar (SAR) and optical imagery. Background of this work is the performance evaluation of vehicle detection and velocity estimation from SAR images when comparing it with reference data derived from aerial image sequences. In terms of detection rates for individual vehicles, the results of evaluation show the challenges of traffic monitoring with SAR. However, it is also shown that general traffic flow parameters can be derived with high quality.