Uwe Soergel

Building Recognition From Multi-Aspect High-Resolution InSAR Data in Urban Areas

The improved ground resolution of state-of-the-art synthetic aperture radar (SAR) sensors suggests utilizing SAR data for the analysis of urban areas. The appearance of buildings in SAR or interferometric SAR (InSAR) data is characterized by the consequences of the inherent oblique scene illumination, such as layover, occlusion by radar shadow, and multipath signal propagation. Therefore, particularly in dense built-up areas, building reconstruction is often impossible from a single SAR or InSAR measurement alone. But, the reconstruction quality can be significantly improved by a combined analysis of multi-aspect data. In this paper, two approaches are proposed to detect and reconstruct buildings of different size from multi-aspect high-resolution InSAR data sets. Both approaches focus on the recognition of buildings supported by knowledge-based analysis considering the mentioned SAR-specific effects observed in urban areas. Building features are extracted independently for each direction from the magnitude and phase information of the interferometric data. Initial primitives are segmented and afterward projected from slant-range into the world coordinate system. From the fused set of primitives of both flight directions, building hypotheses are generated. The first approach exploits the frequently observed lines of bright double-bounce scattering, which are used for building reconstruction in residential districts. In the case of larger buildings, such as industrial halls, often additional features of roof and facade elements are visible. Therefore, in a second approach, extended buildings are extracted by grouping primitives of different kinds. The two approaches are demonstrated in an urban environment for an InSAR data set, which has spatial resolution of about 30 cm and was taken from two orthogonal flight directions.

Potential and limits of InSAR data for building reconstruction in built-up areas

The automatic reconstruction of buildings for the generation of city models is of great interest for different tasks. Three-dimensional information can be directly obtained from both, laser (LIDAR) and radar (InSAR) measurements. The features of both sensors are compared. The data acquisition by SAR is described, with emphasis on the special properties of the interferometric SAR principle. A segmentation approach for building reconstruction is proposed. The results show that building reconstruction is possible from InSAR, but the achievable level of detail cannot compete with LIDAR. The main source of limitation is the inherent side-looking scene illumination of SAR, giving rise to disturbing phenomena interfering with often large parts of the scene. Geometric constraints for the location and size of such problem areas are derived. To identify areas of unreliable data in SAR images of a built-up area, corresponding elevation data are analysed. The impact of the phenomena layover, shadow and dominant scattering at building locations is considered. For this task, a hybrid elevation reference is required. The buildings and the surrounding ground are represented as CAD planes. Natural objects like trees and bushes remain in the raster representation.

Building Detection From One Orthophoto and High-Resolution InSAR Data Using Conditional Random Fields

Todays airborne SAR sensors provide geometric resolution in the order well below half a meter. Many features of urban objects become visible in such data. However, layover and occlusion issues inevitably arise in urban areas complicating automated object detection. In order to support interpretation, SAR data may be analyzed using complementary information from maps or optical imagery. In this paper, an approach for building detection in urban areas based on object features extracted from high-resolution interferometric SAR (InSAR) data and one orthophoto is presented. Features describing local evidence as well as context information are used. Buildings are detected by classification of those feature vectors within a Conditional Random Field (CRF) framework. Although as graphical model similar to Markov Random Fields (MRF), CRFs have the advantage of incorporating global context information, of relaxing the conditional independence assumption between features, and of a more general integration of observations. We show that, first, CRFs perform well in comparison to Maximum Likelihood classifiers and MRFs. Second, the combined use of optical and InSAR features may improve detection results.

A Marked Point Process for Modeling Lidar Waveforms

Lidar waveforms are 1-D signals representing a train of echoes caused by reflections at different targets. Modeling these echoes with the appropriate parametric function is useful to retrieve information about the physical characteristics of the targets. This paper presents a new probabilistic model based upon a marked point process which reconstructs the echoes from recorded discrete waveforms as a sequence ofparametric curves. Such an approach allows to fit each mode of a waveform with the most suitable function and to deal with both, symmetric and asymmetric, echoes. The model takes into account a data term, which measures the coherence between the models and the waveforms, and a regularization term, which introduces prior knowledge on the reconstructed signal. The exploration of the associated configuration space is performed by a reversible jump Markov chain Monte Carlo (RJMCMC) sampler coupled with simulated annealing. Experiments with different kinds of lidar signals, especially from urban scenes, show the high potential of the proposed approach. To further demonstrate the advantages of the suggested method, actual laser scans are classified and the results are reported.

Radar Remote Sensing of Urban Areas

This book presents a unique collection of state-of-the-art contributions by international remote sensing experts focussing on methodologies to extract information about urban areas from Synthetic Aperture Radar (SAR) data. SAR is an active remote sensing technique capable to gather data independently from sun light and weather conditions. Emphasizing technical and geometrical issues the potential and limits of SAR are addressed in focussed case studies, for example, the detection of buildings and roads, traffic monitoring, surface deformation monitoring, and urban change. These studies can be sorted into two groups: the mapping of the current urban state and the monitoring of change. The former covers, for instance, methodologies for the detection and reconstruction of individual buildings and road networks; the latter, for example, surface deformation monitoring and urban change. This includes also investigations related to the benefit of SAR Interferometry, which is useful to determine either digital elevation models and surface deformation or the radial velocity of objects (e.g. cars), and the Polarization of the signal that comprises valuable information about the type of soil and object geometry. Furthermore, the features of modern satellite and airborne sensor devices which provide high-spatial resolution of the urban scene are discussed.

Perceptual grouping for automatic detection of man-made structures in high-resolution SAR data

Modern airborne synthetic aperture radar sensors provide high spatial resolution data. Experimental systems have even achieved decimetre resolution. In such data, many features of urban objects can be identified, which are beyond what has been achieved by radar remote sensing before. An example for the new quality of the appearance of urban man-made objects such as buildings in these data is given and interpreted. The fine level of detail opens the opportunity to reconstruct detailed structures of such objects from SAR data with structural pattern recognition techniques. Artificial intelligence concepts such as production systems provide proper means for this purpose. The feasibility of these methods is demonstrated here. Extended building features such as long thin roof edge lines, groups of salient point scatterers, and symmetric configurations are detected using principles from perceptual grouping and Gestalt psychology. These are good continuation, similarity, proximity and symmetry.

Conditional random fields for urban scene classification with full waveform LiDAR data

We propose a context-based classification method for point clouds acquired by full waveform airborne laser scanners. As these devices provide a higher point density and additional information like echo width or type of return, an accurate distinction of several object classes is possible. However, especially in dense urban areas correct labelling is a challenging task. Therefore, we incorporate context knowledge by using Conditional Random Fields. Typical object structures are learned in a training step and improve the results of the point-based classification process. We validate our approach with two real-world datasets and by a comparison to Support Vector Machines and Markov Random Fields.

Feature Extraction and Visualization of Bridges Over Water From High-Resolution InSAR Data and One Orthophoto

Modern airborne SAR sensor systems provide geometric resolution in the order well below half a meter. By SAR interferometry from pairs of such images, DEM of the same grid size can be obtained. In data of this kind, many features of urban objects become visible, which were beyond the scope of radar remote sensing only a few years ago. However, because of the side-looking SAR sensor principle, layover and occlusion issues inevitably arise in undulated terrain or urban areas. Therefore, SAR data are difficult to interpret even for senior human interpreters. Furthermore, the quality of the InSAR DEM may vary significantly depending on the local topography. In order to support interpretation, SAR data are often analyzed using additional complementary information provided by maps or other remote sensing imagery. In this paper, object feature extraction and visualization from high-resolution InSAR data and one orthophoto is discussed for the example of a scene containing several bridges over water. Bridges are key elements of man-made infrastructure. Monitoring of these important connecting parts of the traffic network is vital for applications such as disaster management or in the context of political crisis, for instance, to evacuate inhabitants and to deliver goods and equipment. Aims of the approach are to derive key features of the bridges geometry from the complementary data sources, to determine the water level, smooth the noisy InSAR DEM data, especially at water surfaces, and, finally, to generate an improved 3-D visualization of the scene by overlapping the optical image on the InSAR DEM.

Visibility analysis of man-made objects in SAR images

The increasing resolution of SAR data offers the possibility to utilize this data for a detailed scene interpretation in urban areas. Different SAR specific phenomena like foreshortening, layover, shadow and multipath-propagation burden the interpretation. A high resolution LIDAR DEM of an urban scene is incorporated to investigate the impact of the phenomena on the visibility of man-made objects by a SAR measurement from a given sensor trajectory and orientation. LIDAR data as ground truth is well suited for this task, because it contains elevation information of man-made and natural objects. Incoherent sampling of the DEM simulates shadow and layover areas. By a variation of viewing and aspect angles a large number of such simulations are carried out. From this set of segmentations the n best are determined with respect to the visibility of roads and buildings. Furthermore, the locations of total reflection or double-bounce scattering in the vicinity of buildings are determined.

Extraction of building polygons from SAR images: Grouping and decision-level in the GESTALT system

The GESTALT-system is a stratified architecture for challenging computer vision tasks. This contribution focuses on the 3rd and 4th layer of it - the grouping and decision layers. As example application building recognition from high resolution SAR-data is presented. The 3rd layer contains an assessment driven perceptual grouping process with any-time capability and flexible control. Important grouping principles such as good continuation and symmetry are utilized. A dynamic programming optimization is used in the final decision and post-processing layer to find closed polygons that describe the outlines of buildings. Further post processing includes polygon editing and consistency enforcement.