Ulrich Michel

Combined Edge Segment Texture Analysis for the Detection of Damaged Buildings in Crisis Areas

This paper describes the results of a new combined method that consists of a cooperative approach of several different algorithms for automated change detection. These methods are based on isotropic frequency filtering, spectral and texture analysis, and segmentation. For the frequency analysis, different band pass filters are applied to identify the relevant frequency information for change detection. After transforming the multitemporal images using a fast Fourier transform and applying the most suitable band pass filter to extract changed structures, we apply an edge detection algorithm in the spatial domain. For the texture analysis, we calculate the parameters energy and homogeneity for the multitemporal datasets. Then a principal component analysis is applied to the new multispectral texture images and subtracted to get the texture change information. This method can be combined with spectral information and prior segmentation of the image data as well as with morphological operations for a final binary change result. A rule-based combination of the change algorithms is applied to calculate the probability of change for a particular location. This Combined Edge Segment Texture (CEST) method was tested with high-resolution remote-sensing images of the crisis area in Darfur (Sudan). Our results were compared with several standard algorithms for automated change detection, such as image difference, image ratio, principal component analysis, multivariate alteration detection (MAD) and post classification change detection. CEST showed superior accuracy compared to standard methods.

Per-field crop classification in irrigated agricultural regions in middle Asia using random forest and support vector machine ensemble

Accurate crop identification and crop area estimation are important for studies on irrigated agricultural systems, yield and water demand modeling, and agrarian policy development. In this study a novel combination of Random Forest (RF) and Support Vector Machine (SVM) classifiers is presented that (i) enhances crop classification accuracy and (ii) provides spatial information on map uncertainty. The methodology was implemented over four distinct irrigated sites in Middle Asia using RapidEye time series data. The RF feature importance statistics was used as feature-selection strategy for the SVM to assess possible negative effects on classification accuracy caused by an oversized feature space. The results of the individual RF and SVM classifications were combined with rules based on posterior classification probability and estimates of classification probability entropy. SVM classification performance was increased by feature selection through RF. Further experimental results indicate that the hybrid classifier improves overall classification accuracy in comparison to the single classifiers as well as user´s and producer´s accuracy.

Object-based detection of destroyed buildings based on remotely sensed data and GIS

The paper describes an object-based method to detect destroyed buildings as a consequence of an earthquake. The investigation is based on the analysis of remotely sensed raster and vector-based data. The methodology includes three main steps: generation of features defining the states of buildings, classification of building state and data import in GIS. This paper concentrates on the first step of the three, the generation of features. The appropriately selected features are indispensable for the following successful classification. The described methodology is applied to remotely sensed images of areas that had been subject to an earthquake. Our preliminary results confirm the potential of the proposed approach for detection of the building state. The change detection methodology has been developed solely with Open Source Software. GRASS GIS is involved for vector and raster data processing and presentation. Programming languages Python and Bash are used to develop new GRASS-modules.

Object-based detection of LUCC with special regard to agricultural abandonment on Tenerife (Canary Islands)

The island Tenerife has always been used for intensive agriculture, whereby the natural landscape was continuously altered. Especially mountainous areas with suitable climate conditions have been drastically transformed for agricultural use by building of large terraces to get flat surfaces. In recent decades political and economic developments lead to a transformation process (especially inducted by an expansive tourism), which caused concentration- and intensificationtendencies of agricultural land use as well as agricultural set-aside and rural exodus. In order to get information about the land use and land cover (LULC) patterns and especially the agricultural dynamics on Tenerife, a multi-scale, knowledge-based classification procedure for recent RapidEye data was developed. Furthermore, a second detection technique was generated, which allows an exact identification of the total ever utilised agricultural area on Tenerife, also containing older agricultural fallow land or agricultural set-aside with a higher level of natural succession (under the assumption that long-term fallow areas can be detected mainly together with old agricultural terraces and its specific linear texture). These areas can hardly be acquired in the used satellite imagery. The method consists of an automatic texture-oriented detection and area-wide extraction of linear agricultural structures (plough furrows and field boundaries of arable land, utilised and non-utilised agricultural terraces) in current orthophotos of Tenerife. Through the detection of recent agricultural land use in the satellite imagery and total ever utilised agricultural area in the orthophotos, it is possible to define the total non-active agricultural land as well as hot spots of agricultural decrease.

Object-based rapid change detection for disaster management

Rapid change detection is used in cases of natural hazards and disasters. This analysis lead to quick information about areas of damage. In certain cases the lack of information after catastrophe events is obstructing supporting measures within disaster management. Earthquakes, tsunamis, civil war, volcanic eruption, droughts and floods have much in common: people are directly affected, landscapes and buildings are destroyed. In every case geospatial data is necessary to gain knowledge as basement for decision support. Where to go first? Which infrastructure is usable? How much area is affected? These are essential questions which need to be answered before appropriate, eligible help can be established. This study presents an innovative strategy to retrieve post event information by use of an object-based change detection approach. Within a transferable framework, the developed algorithms can be implemented for a set of remote sensing data among different investigation areas. Several case studies are the base for the retrieved results. Within a coarse dividing into statistical parts and the segmentation in meaningful objects, the framework is able to deal with different types of change. By means of an elaborated normalized temporal change index (NTCI) panchromatic datasets are used to extract areas which are destroyed, areas which were not affected and in addition areas which are developing new for cases where rebuilding has already started. The results of the study are also feasible for monitoring urban growth.

Integrated data processing of remotely sensed and vector data for building change detection

In recent years natural disasters have had an increasing impact leading to tremendous economic and human losses. Remote sensing technologies are being used more often for rapid detection and visualization of changes in the affected areas, providing essential information for damage assessment, planning and coordination of recovery activities. This study presents a GIS-based approach for the detection of damaged buildings. The methodology is based on the integrated analysis of vector data containing information about the original urban layout and remotely sensed images obtained after a catastrophic event. For the classification of building integrity a new ‘Detected Part of Contour’ (DPC) feature was developed. The DPC feature defines a part of the building contour that can be detected in the related remotely sensed image. It reaches maximum value (100%) if the investigated building contour is intact. Next, several features based on the analysis of textural information of the remotely sensed image are considered. Finally, a binary classification of building conditions concludes the change detection analysis. The proposed method was applied to the 2010 earthquake in Qinghai (China). The results indicate that a GIS-based analysis can markedly improve the accuracy of change detection analysis. The proposed methodology has been developed solely within the Open Source Software environment (GRASS GIS, Python, Orange). The employment of Open Source Software provides the way for an innovative, flexible and costeffective implementation of change detection operations.

Detektion von zerstörten Gebäuden in Krisengebieten aus pan-chromatischen Daten

Das Ziel dieses Artikels ist die Analyse von Veranderungen in Gebieten, in denen sich Katastrophen mit plotzlichen Anderungen an Gebauden und der Infrastruktur ereignet haben. Standardverfahren der Veranderungsanalyse fuhren zu keinem zufriedenstellenden Ergebnis, daher wurde ein neues Verfahren entwickelt. Die in diesem Artikel dargestellte Methode erlaubt eine schnelle Detektion und Visualisierung von Veranderungen in Krisen- und Katastrophengebieten. Dies ist eine wichtige Voraussetzung fur die Planung und Koordination von Hilfskrafteinsatzen. Die vorgeschlagene Methode basiert auf Frequenzanalysen, Segmentierung und Texturmerkmalen. Sie kombiniert die unterschiedlichen Ansatze in einem Verfahren mittels eines Entscheidungsbaumes. Im Vergleich mit funf Standardverfahren zeigte dieser neue Ansatz die besten Resultate.

Development of a satellite-based multi-scale land use classification system for land and water management in Uzbekistan and Kazakhstan

Satellite remote sensing is an invaluable tool to assess the status and changes of irrigated agricultural systems. Agricultural sites are among the most heterogeneous sites at the landscape level: spatial pattern of agricultural fields, within-field heterogeneity, crop phenology and crop management practices vary significantly. Highly dynamic objects (crops and crop rotations) result in large temporal variability of surface spatial heterogeneity. Technological advances have opened the possibility to monitor agricultural sites combining satellite images with both high spatial resolution and high revisit frequency, which could overcome these constraints. Yet depending on the field sizes and crop phenology of the agricultural system observed, requisites in terms of the instrument´s spatial resolution and optimal timing of crop observation will be different. The overall goal is to quantitatively define region specific satellite observation support requirements in order to perform land use classification at the field basis. The main aspect studied here is the influence of spatial resolution on the accuracy of land use classification over a variety of different irrigated agricultural landscapes. This will guide in identifying an appropriate spatial resolution and input parameters for classification. The study will be performed over distinct locations in irrigated agro-ecosystems in Central Asia, where reliable information on agricultural crops and crop rotations is needed for sustainable land and water management.

Fusion of remote sensing, DGPS, total station and GPR data for the3-D mapping of Himalayan cryosphere: application and future potential (Conference Presentation)

Himalayan Glaciers are a major source of fresh water, and in places are the critical source of water for drinking, irrigation and hydroelectric power generation for South Asia. Modern water resource management requires understanding the volume of source available for a robust planning and development for the future.Glaciers are highly sensitive indicators to any climate change. Contemporary size and volume are critical factors for timely evaluation/assessment for both near-term and long-term changes in both temperature and precipitation and the cryosphere thereupon. Glacier area and surface morphology can be readily mapped from both satellites imagery and aerial photographs. With the help of remotely sensed data and GIS analysis, glacier surface areas have been mapped with mean spatial resolution of 10 meter. The surface mapping with such resolution is neither able to access the exact volume nor determine sensitivity of glacier to the recent climate. Most satellite maps are only two-dimensional mapping of crysophere, but three-dimensional maps of glacier hydrological systems are necessary for volumetric assessment and long-term planning. Therefore, the integration of survey instrument such as DGPS, Total Station with millimeter accuracy, and simultaneous simulation with Ground Penetrating Radar (GPR) surveys over the mapped conduit systems will help in accessing glacier mass accurately and define subsurface geometry for the 3-D modeling of glaciers for better understanding and estimation of water fresh resource. Further, the 3-D mapping of cryosphere will help us to access accurate the volume of fresh water in the Himalayan cryosphere, along with contemporary dynamics. We show an integrated approach to assess and quantify the Himalayan cryosphere by integrating such techniques for a better management and understanding of the Himalayan cryosphere to climatic parameters and management of future water requirements. However, it is an established fact that glaciers show varied sensitivity to climate over the time and space, both in growing and receding.Therefore, monitoring this fresh water resource is most essential for an agrarian country like India where demand for irrigation is great in the Great Plains. Present study will represent the seamless integration of field based 3-D Total station mapping of glacier snout of Gangotri Glacier, GPR Profiling of Glacier at selective locations, Volume estimation and annual change in the volume and its integration with the MODIS LST data in a way to access the present glacier sensitivity to climatic variability, as well as help model future scenario more accurately for robust management of this finite water resource.

GIS4schools: custom-made GIS-applications for educational use

From a didactic point of view the procurement and the application of modern geographical methods and functions become more and more important. Although the integration of GIS in the classroom is repeatedly demanded, inter alia in Baden-Württemberg, Germany, the number of GIS users is small in comparison to other European countries or the USA. Possible reasons for this could, for instance, lie in the lack of GIS and computer knowledge of the teachers themselves and the subsequent extensive training effort in Desktop-GIS [1]. Today you have the technological possibilities to provide the broad public with geoinformation and geotechnology: Web technologies offer access to web-based, mobile and local applications through simple gateways. The objective of the project “GIS4schools” is to generate a service-based infrastructure, which can be operated via mobile clients as well as via Desktop-GIS or a Browser. Due to the easy availability of the services the focus is in particular on students. This circumstance is a novelty through which a differentiated approach to the implementation of GIS in schools is established. Accordingly, the pilot nature of this project becomes apparent as well as its greater importance beyond its actual content especially for the sector of media development at colleges of education. The continuity from Web-GIS to Desktop-GIS is innovative: The goal is to create an adapted multi-level solution which allows both, an easy introduction if desired or a detailed analysis – either to be achieved with a focus especially on students and their cooperation among one another.