Automatic SAR simulation technique for object identification in complex urban scenarios

Most of the existing SAR simulators provide simulated images only for visual interpretation. This paper presents a new approach for supporting the automatic interpretation of meter resolution SAR images in complex urban scenarios. To this end, an automatic system for generating and geocoding of simulated images has been developed, based on the SAR simulator RaySAR and digital surface models as geometric information for urban scenes. Various simulated images and different layers (layover, shadow, double bounce, and ground) are generated for the whole scene as well as for individual buildings, with consideration of neighborhood influences. The proposed approach is tested on an area in central Munich with a LiDAR digital surface model as the input. The simulated images are compared with a TerraSAR-X spotlight image and show good geocoding accuracy, reasonable mask layers, and precise individual building layover contours.