Simulation-based recognition of airplane signatures from TerraSAR-X data

Abstract:
Recognition of medium scaled objects like airplanes with current high resolution spaceborne SAR image data is, in most cases, impossible due to restrictions in available spatial resolution and signal-to-noise ratio. Furthermore, radar specific imaging effects like layover, multi-path propagation or speckle noise make signature analysis a very challenging task. Recently, a SAR simulator has been developed to assist image operators in signature analysis and recognition. Additionally, TerraSAR-X initiated a new staring spotlight imaging mode improving spatial resolution significantly. This paper will show the benefits of applying simulation-based recognition of airplanes on this imaging mode in comparison to the existing image products.