The Concept for High-Resolution ISAR-based Traffic Surveillance System

The paper presents a concept of classifying objects moving on a predefined or estimated trajectory based on their ISAR images. The main idea is to use model-based techniques with 3D CAD models of ground targets and modeled movement along an approximately known trajectory. An example of such a setup is an analysis of road traffic, where the aim is to classify different vehicles moving along the road, and the road layout constrains the trajectory partially. Still, the exact movement path is not known a priori. Thus, either the trajectory knowledge has to be refined from the acquired data, or the model needs to be robust enough to the trajectory variation. The concept presentation is backed by the results of simulations making use of a realistic SAR/ISAR signal simulator.