Rainer Speck

Flood Detection in Urban Areas Using TerraSAR-X

Flooding is a major hazard in both rural and urban areas worldwide, but it is in urban areas that the impacts are most severe. An investigation of the ability of high-resolution TerraSAR-X synthetic aperture radar (SAR) data to detect flooded regions in urban areas is described. The study uses a TerraSAR-X image of a one-in-150-year flood near Tewkesbury, U.K., in 2007, for which contemporaneous aerial photography exists for validation. The German Aerospace Center (DLR) SAR end-to-end simulator (SETES) was used in conjunction with airborne scanning laser altimetry (LiDAR) data to estimate regions of the image in which water would not be visible due to shadow or layover caused by buildings and taller vegetation. A semiautomatic algorithm for the detection of floodwater in urban areas is described, together with its validation using aerial photographs. Of the urban water pixels that are visible to TerraSAR-X, 76% were correctly detected, with an associated false positive rate of 25%. If all the urban water pixels were considered, including those in shadow and layover regions, these figures fell to 58% and 19%, respectively. The algorithm is aimed at producing urban flood extents with which to calibrate and validate urban flood inundation models, and these findings indicate that TerraSAR-X is capable of providing useful data for this purpose.

Efficient and precise processing for squinted spotlight SAR through a modified stolt mapping

Processing of squinted SAR spotlight data is a challenge because of the significant range migration effects of the raw data over the coherent aperture time. Although in theory the ()-algorithm takes care of these aspects, its digital implementation requires a time-consuming interpolation step. Moreover, the limited precision of this interpolation can introduce distortions at the edges of the final image especially for squinted geometries. A wave number domain processing using a modified Stolt mapping will be developed and analyzed to enhance the quality of the final SAR image. Additionally, the proposed algorithm has a decreased computational load compared to the original ()-algorithm. Simulation results will validate the focusing and efficiency performances of the modified wave number domain algorithm.

An end-to-end simulator for high-resolution spaceborne SAR systems

The purpose of this paper is to present an end-to-end simulator for spaceborne high-resolution SAR systems that is capable of simulating realistic raw data and focused images of extended three-dimensional scenes. The simulator is based on precise mathematical modeling of an overall SAR system chain and generates information on the quality of the image data and its suitability to interpret target and background signatures. The principal components of the simulator are: - the generation of an extended scene, including the fully polarimetric scattering behavior of the three-dimensional surfaces, man made objects, and the typical SAR effects like overlay, speckle noise, shadowing;-an accurate SAR sensor simulation (antenna, transmit and receive path);-the generation of the raw data depending on the desired SAR mode (stripmap, spotlight mode);-the image processing and evaluation. The flexible and modular structure allows for adjustment and extension to fulfill different tasks. The most important modules reflecting the basic physical models will be described and simulation results will be demonstrated.

Fast ISAR image generation through localization of persistent scattering centers

Imagery data acquired by recently launched space borne SAR systems demonstrate a very good spatial resolution (e.g. one meter with TerraSAR-X). The designs of such complex systems make it compulsory to do SAR end-to-end simulations to optimize image quality (e.g. spatial and radiometric resolution, ambiguity suppression, dynamic range, etc.). The most complex, critical and challenging modules have to be designed for the generation of SAR raw data and SAR image generation, because the limits of computability and memory requirements are reached very quickly. Moreover, the analysis of SAR images is a demanding task, because of their sensor specific effects. Therefore, a simulation tool is under development to analyze realistic target features and make the scattering processes transparent to the user. With the method presented in this paper, SAR images of complex scattering bodies can be generated in a very efficient way. This is done by directly localizing scattering centers and identifying their persistency along the synthetic aperture. Thus the usual raw data generation and processing steps are dropped. The resulting images show a very good similarity to reality, because scattering centers due to multipath propagation effects are also handled. Furthermore this toolkit makes it possible to visualize the scattering centers and their evolution, by mapping them on the 3D structure of the scattering body. This results in transparency of the whole scattering process, which greatly improves the understanding of the image effects. The paper presents this new approach for the application of inverse SAR (ISAR) and first simulation results.

Applications of simulation techniques for high resolution SAR systems

The interpretation of radar imagery can often be very challenging due to specific imaging effects. Additionally, a fusion with other remote sensing data is not easy to accomplish, because of the imaging radars slant range coordinate system. Simulation techniques are able to provide essential assistance or even practical solutions to such challenges. This paper shows the usage of a developed simulator in practical applications for signature analysis, data fusion and acquisition planning of spaceborne radar systems, but also for parametric end-to-end simulations of a very high resolution ground based radar.

Transforming optical image data into a SAR system's range-based image space

The fusion of image data from different sensor types is an important processing step for many remote sensing applications to maximize information retrieval from a given area of interest. The basic process to fuse image data is to select a common coordinate system and resample the data to this new image space. Usually, this is done by orthorectifying all those different image spaces, which means a transformation of the image’s projection plane to a geographic coordinate system. Unfortunately, the resampling of the slant-range based image space of a space borne synthetic aperture radar (SAR) to such a coordinate system strongly distorts its content and therefore reduces the amount of extractable information. The understanding of the complex signatures, which are already hard to interpret in the original data, even gets worse. To preserve maximum information extraction, this paper shows an approach to transform optical images into the radar image space. This can be accomplished by using an optical image along with a digital elevation model and project it to the same slant-range image plane as the one from the radar image acquisition. This whole process will be shown in detail for practical examples.

Analysis of SAR images by simulation

Accurate simulation tools for the design of space borne synthetic aperture radar systems (SAR) are compulsory for the analysis of the systems capabilities, because ground based experimental tests are in most cases impossible and very costly. Through a simulation process it is possible to analyze the image quality parameters for a given system configuration or evaluating the effects in SAR images when this configuration is changed. A new fast SAR image simulator (SARIS) is currently under development on the basis of an existing toolset called SAR end-to-end simulator (SETES). This image simulator produces SAR images by using the point spread function (PSF) of a focused point target response in contrast to SETESs very expensive raw data generation module. In SARIS the SAR image is produced through a convolution of the PSF with the so-called reflectivity map of the scene. In this paper first simulation results with a prototype of SARIS are given to show effects like motion errors and low peak-to-side-lobe ratios.

Mapping detailed 3D information onto high resolution SAR signatures

Due to challenges in the visual interpretation of radar signatures or in the subsequent information extraction, a fusion with other data sources can be beneficial. The most accurate basis for a fusion of any kind of remote sensing data is the mapping of the acquired 2D image space onto the true 3D geometry of the scenery. In the case of radar images this is a challenging task because the coordinate system is based on the measured range which causes ambiguous regions due to layover effects. This paper describes a method that accurately maps the detailed 3D information of a scene to the slantrange-based coordinate system of imaging radars. Due to this mapping all the contributing geometrical parts of one resolution cell can be determined in 3D space. The proposed method is highly efficient, because computationally expensive operations can be directly performed on graphics card hardware. The described approach builds a perfect basis for sophisticated methods to extract data from multiple complimentary sensors like from radar and optical images, especially because true 3D information from whole cities will be available in the near future. The performance of the developed methods will be demonstrated with high resolution radar data acquired by the space-borne SAR-sensor TerraSAR-X.

Automatic change detection using very high-resolution SAR images and prior knowledge about the scene

Change detection using very high resolution SAR images is an important source of information for reconnaissance applications. Modern SAR sensors are capable of acquiring many images in short periods of time, which creates the need for a reliable automatic change detection method. In this paper, we will describe a new automatic change detection approach that combines very high resolution SAR images with prior knowledge about the imaged scene. In this case, the prior knowledge about the scene will come from vector maps, which can be obtained from a Geographic Information System (GIS). These vector maps will allow us to determine which regions are of interest for the change detection, and what kind of changes/objects can be expected there. The algorithm described in this paper will be applied to a time series of high resolution TerraSAR-X images of a port with military shipyards, and used to automatically detect ship activity and extract information about the detected ships. In this case, the vector maps were obtained from a Geographic Information System (GIS) containing map data from OpenStreetMap

Identification of maritime target objects from high resolution TerraSAR-X data using SAR simulation

In general, interpretation of signatures from synthetic aperture radar (SAR) data is a challenging task even for the expert image analyst. For the most part, this is caused by radar specific imaging effects, e.g. layover, multi-path propagation or speckle noise. Specifically for the application in maritime security, ship signatures exhibit additional defocusing effects due to the ship’s movement even when they are anchored. Focusing on object recognition, the detection of target signatures can be done with a pretty good chance of success, but the identification is often impossible. To assist image analysts in their recognition tasks, a SAR simulation tool has been developed recently. It is very simple to operate, by simulating available 3D model data of ships and test the resulting simulated signatures with their real counterpart from SAR images. This is a very robust way to identify larger vessels out of current one meter resolution space borne SAR data. Nevertheless, for smaller vessels this can be still very challenging, because the resolution is too coarse. Recently, TerraSAR-X initiated a new staring spotlight imaging mode that enhances cross-range resolution significantly and therefore also improves the chance for the identification of smaller vessels. This paper demonstrates the capabilities of the developed simulation tool in assisted target recognition specifically on ship signatures. The improvement of recognition performance will be studied by comparing results for TerraSAR-X sliding spotlight mode and staring spotlight mode data.