Olaf Hellwich

Geocoding SAR interferograms by least squares adjustment

A geocoding model for Synthetic Aperture Radar (SAR) interferometry based on a least-squares adjustment combining interferometric phase, range, Doppler centroid frequency, flight path and control point data is proposed. The complete mathematical framework for the computation of object space coordinates without approximations is presented. It provides a way to an efficient implementation of the algorithm for geocoding the pixels of an interferogram. The method is preferably applicable to spaceborne dual-pass interferometry, and independent of the orbit configuration. An accuracy analysis of object point positioning is conducted and results of geocoding an ERS tandem interferogram are shown.

Extraction of linear objects from interferometric SAR data

A new method for the automated extraction of pipelines and other linear objects from Synthetic Aperture Radar (SAR) scenes is presented. It combines intensity data with coherence data from an interferometric evaluation of a SAR scene pair. The fusion is based on Bayesian statistics and is part of a Markov random field (MRF) model for line extraction. Both intensity and coherence data are evaluated using rotating templates. The different statistical properties of intensity and coherence are taken into account by a multiplicative noise model and an additive noise model respectively. The MRF model introduces prior knowledge about the continuity and the narrowness of lines. Posterior odds resulting from the MRF method are input to a method based on ziplock snakes for linear object extraction. This processing step is controlled interactively which is necessary as fully automatic processing of the given noisy data does not provide sufficiently predictable results. The method is applied to data of the ERS tandem mission.

Extracting line features from synthetic aperture radar (SAR) scenes using a Markov random field model

Due to the speckle effect of coherent imaging the detection of lines in SAR scenes is considerably move difficult than in optical images. A new approach to detect lines in noisy images using a Markov random field (MRF) model and Bayesian classification is proposed. The unobservable object classes of single pixels are assumed to fulfil the Markov condition, i.e. to depend on the object classes of neighboring pixels only. The influence of neighboring line pixels is formulated based on potentials derived from a random walk model. Locally, the image data is evaluated with a rotating template. As SAR intensity data is deteriorated by multiplicative noise, the response of the local line detector is a normalized intensity ratio which results in a constant false alarm rate. The approach integrates intensity, coherence from interferometric processing of a SAR scene pair, and given Geographic Information System (GIS) data.

Fusion of SAR/INSAR Data and Optical Imagery for Landuse Classification

SAR/INSAR data and optical imagery such as high-resolution panchromatic or multispectral data show different information about the imaged objects, and have different advantages and disadvantages when used for object extraction or landuse classification. Multispectral optical image data is largely determined by the type of the material an object consists of. Panchromatic data which is often available with a higher resolution than multispectral data emphasizes geometric detail of the objects, e.g. the complex structure of anthropogenic objects such as road networks. In contrary to this, SAR data contain information about small-scale surface roughness and to a lower degree soil moisture. Height information derived by interferometric processing of SAR data contains large-scale surface roughness. Polarimetric SAR data show geometric surface and material structure. These different types of information are referring to different object qualities and are, therefore, largely uncorrelated which helps to reduce ambiguities in the results of object extraction. The main advantage of passive optical imagery with respect to SAR data is the lack of the speckle effect leading to images with a far better extractability of linear as well as areal objects when systems with the same resolution are compared. A major advantage of SAR is its all-weather capability which allows the acquisition of time series of imagery with exact acquisition dates under any climatic condition. In this paper, these complementary properties of SAR and optical image data are demonstrated and used to improve object extraction and landuse classification results. Übersicht SAR/INSAR-Daten und optische Bilddaten wie hochauflösende panchromatische oder multispektrale Daten enthalten unterschiedliche Information über die abgebildeten Objekte und haben daher spezifische Vorbzw. Nachteile, wenn sie für Objektextraktion oder Landnutzungsklassifikation eingesetzt werden. Multispektrale optische Bilddaten werden überwiegend durch die Eigenschaften der Materialien bestimmt, aus denen die Objekte bestehen. Panchromatische Bilder, die häufig in einer höheren Auflösung als Multispektralbilder verfügbar sind, ermöglichen die Extraktion komplexer Strukturen anthropogener Objekte wie Straßennetze oder Gebäude. Im Gegensatz hierzu enthalten SAR-Daten Information über kleinmaßstäbige Oberflächenrauhigkeit und in reduziertem Umfang Bodenfeuchte. Aus interferometrischen SAR-Daten gewonnene Höhendaten enthalten Information über großmaßstäbige Oberflächenrauhigkeit. Polarimetrische SAR-Daten zeigen die geometrische Oberflächenund Materialstruktur. Diese unterschiedlichen Informationsarten beziehen sich auf verschiedene Objekteigenschaften und sind daher weitgehend unkorreliert, was dazu beiträgt, Vieldeutigkeiten in den Resultaten der Objektextraktion zu reduzieren. Der wesentliche Vorteil optischer Bilddaten im Vergleich zu SAR-Daten ist das Fehlen des Speckle-Effekts, was zu Bildern mit einer bei gleicher Auflösung wesentlich verbesserten Extrahierbarkeit sowohl linearer als auch flächenhafter Objekte führt. Ein wesentlicher Vorteil von SAR ist seine Allwetter-Einsatzfähigkeit, die die Aufnahme von Zeitreihenbilddaten mit exakt eingehaltenen Aufnahmezeitpunkten unter beliebigen klimatischen Bedingungen ermöglicht. In diesem Artikel werden die komplementären Eigenschaften von SARund optischen Bilddaten demonstriert und zur Verbesserung der Ergebnisse von Objektextraktion bzw. Landnutzungsklassifikation verwendet. Für die Dokumentation Datenfusion / Bayessche Netze / SAR / optische Bilddaten / multitemporale Bilddaten / Objekterkennung Frequenz 55(2001) 3-4

Experiences with automatic relative orientation

A report on recent experiences with a new procedure for automatic relative orientation is given. A hierarchical approach provides conjugate points using image pyramids. The implemented method is based on feature extraction by an interest operator, feature matching between the two images, area-based image correlation, a robust least squares bundle adjustment and feature tracking through several levels of the image pyramid. The automatic procedure proved to be successful for various combinations of terrain relief, surface cover, and image scale. The paper presents the results of intensive tests. It was found that only a limited number of control parameters of the algorithm have to be individually adjusted to the terrain specifics. Further investigations need to be conducted to find out whether the project- dependent parameters can be reliably predicted for certain classes of images.

Multisensor data fusion for automated scene interpretation

An approach to the combined extraction of linear as well as two-dimensional objects from multisensor data based on a feature- and object-level fusion of the results is proposed. The data sources are DAIS hyperspectral data, AES-1 SAR data, and high- resolution panchromatic digital orthoimages. Rural test areas consisting of a road network, agricultural fields, and small villages were investigated. The scene interpretation is based on a conceptual model consisting of a semantic net for each of the sensors and a semantic net of the real world objects. The sensor nets and the object net are combined into one network by means of a geometry and material level of network nodes. Road networks are extracted from the panchromatic orthoimage and from selected hyperspectral bands. Based on the knowledge that roads compose networks the extraction results are combined. Two-dimensional, i.e. areal, objects are extracted from hyperspectral data after a principal component transformation. The SAR data is segmented using image intensity and interferometric elevation. The classifications of the hyperspectral and SAR data are combined with the extracted road network using rule- and segment-based methods. In the outlook, comments are given on the trade-off between the improvement of the results using the new method and the increasing costs for data acquisition.

Line extraction from Synthetic Aperture Radar scenes using a Markov random field model

Due to the speckle effect of coherent imaging the detection of liens in SAR scenes is considerably more difficult than in optical images. A new approach to detect lines in noisy images using a Markov random field model and Bayesian classification is proposed. The unobservable object classes of single pixels are assumed to fulfill the Markov condition, i.e. to depend on the object classes of neighboring pixels only. The influence of neighboring line pixels is formulated based on potentials derived from a random walk model. Locally, the image data is evaluated with a rotating template. As SAR intensity data is deteriorated by multiplicative noise, normalized intensity ratio is used as the response of the local line detector resulting in a constant false alarm rate. The new approach integrates SAR intensity and coherence from interferometric processing of a SAR scene pair. Besides maximum a posterior and iterated conditional modes estimation of the object parameters, an implementation of local highest confidence first estimation is used. It is initially applied to the sites which are most probably structures in object space, and is then allowed to progress to regions less promising for line detection depending on the results of previous iterations. In this way processing times are substantially reduced.