The global dimension of urbanization requires adequate approaches to monitor and analyze the extent of urban areas and the environmental conditions therein. This thesis investigated the potential of airborne hyperspectral data to provide information on impervious urban areas that is needed for an integrated analysis of such coupled natural and human systems. For this purpose all processing steps from raw data to the final map products were performed. Two processing steps that were lacking optimization with regard to the challenges typical for urban areas have been advanced, i.e. the normalization of brightness gradients and the land cover classification approach. Impacts of the mandatory geocoding and the frequently applied image segmentation on map accuracy were investigated and discussed against the background of workflow optimization.
The metropolitan area of Berlin proved to be a useful study site for this investigation, because of three reasons. First, a variety of urban structure types were covered. Second, a large amount of field data were collected parallel to the remote sensing data acquisition, e.g. land cover maps and spectroscopic field measurements. Third, abundant additional data were available for validation of land cover maps and impervious surface estimates. Thus, comparison of image processing products to various independent data sets was possible at each stage of the processing chain.
The four research questions that were stated in Chapter I are addressed individually before main conclusions are drawn:
(1) Can brightness gradients in airborne hyperspectral data from urban areas be eliminated using an empirical normalization approach that requires no additional field measurements?
In Chapter II, the existence of surface type specific brightness gradients was shown. The curvature of these view-angle dependent gradients was explained based on the directional properties of corresponding surfaces. The two empirical approaches that were suggested for the normalization of this phenomenon, i.e. the class-wise and the weighted class-wise method, both eliminated the brightness gradients. Based on reference surfaces from additional HyMap data the superiority of the suggested approaches to the traditional global normalization approach was demonstrated.
The class-wise normalization approach constitutes an empirical solution that is well suited for the spectral and spatial heterogeneity of urban environments. It does not require information from spectral libraries or the parameterization of complex models. Extending the approach by assigning weighted correction factors offers the possibility to handle surface types that are characterized by smooth transition. The approaches may thus also be useful in natural ecosystems where gradual changes in surface types are typical. Examples for such situations include forest type transitions along altitudinal gradients, arid- and semi-arid regions, or agricultural areas at early growth stages.
(2) Do support vector machines bear the potential to directly use the full hyperspectral information for the successful delineation of urban land cover classes such as built-up and non built-up impervious surfaces without separate feature extraction or the previous definition of spectral sub-classes?
In Chapter III, SVM were used to classify a large and heterogeneous HyMap image into five land cover classes. All classes, including those characterized by high spectral heterogeneity and multimodal distributions (compare Fig. I-2), were separated without a previous definition of spectral sub-classes. The classification was performed on the original spectral bands and yielded high overall and class-specific accuracies for all urban structure types. Results underline the high potential of SVM for complex classification problems without previous feature extraction or selection.
Thus, results from Chapter III show that a sequential processing workflow similar to the one suggested by Kuo and Landgrebe (2004) (Fig. I-3) is not necessary. This is further stressed by the results from additional tests which went beyond the scope of Chapter III:
- transforming the HyMap image into PCs did not improve classification results. Results achieved on the first 20 PCs, for example, were inferior to those achieved on 114 original spectral bands and the feature extraction decreased classification accuracy.
- a sequential classification approach with hierarchically organized SVM did not lead to higher accuracy. Such approaches that split the complex multiclass problem into more simple sub-problems proved useful for the work with artificial neural networks (Udelhoven et al., 2000). For SVM, however, such a strategy did not improve classification accuracies. These findings are accordance with those from Melgani and Bruzzone (2004) and prove that SVM classifications are very accurate without complex classification setups.
- results from the SVM classification did not decrease when applied to the image data without normalized brightness gradients. Therefore, similar classification results might be achieved with a processing workflow even simpler than the one in this work. Nevertheless, the normalization of the brightness gradients is mandatory when spectral libraries or image data from different acquisition times or locations are integrated into the classification.
SVM can thus be recommended for complex problems due to their high classification accuracy and for their ability to achieve optimal results with a simple and intuitive setup. This way, they fulfill the requirements Richards (2005) mentions for future classification methods.
(3) How accurate can land cover and impervious surface coverage be mapped from hyperspectral images and what are the main sources of inaccuracy?
In Chapter V, the influence of different potential sources of inaccuracies on the land cover map and the derived map on impervious surface coverage were assessed. The accuracies of the two maps were evaluated prior to and after geocoding based on various reference data. This way, the decrease in accuracy could be linked either to individual processing steps or to phenomena that generally limit remotely sensed information.
Results show that hyperspectral data provides the spectral information needed to differentiate the major elements of urban heterogeneity according to Cadenasso et al. (2007) and to estimate impervious surface coverage based on this land cover information. Problems with mixed pixels and spectrally ambiguous surfaces during SVM classification of the HyMap data were overall relatively little, but contributed to the overall error. Additional relevant error sources could be identified, though: the geocoding generally caused inaccuracies, as was seen in nadir regions; the additional impact of the missing information on building heights was shown by the increasing error of the corresponding class at larger view-angles; the view-angle independent assignment of streets to vegetation proved the influence of tree crowns obscuring surfaces underneath. This phenomenon leads to a general negative offset in the estimation of impervious surface coverage.
The consequences of individual errors and the overall error for subsequent analyses depend on the spatial scale and the scope of analysis. For example, information on the spectrally well recognized but systematically displaced buildings can be used for a block-wise analysis of the spatial patterns of building. However, severe problems will be caused by such spatial offsets when data from additional sources is combined with the remotely sensed information on individual buildings. In general, the occlusion of surfaces behind buildings and the overestimation of vegetation due to tree crowns – two problems that are not exclusive to hyperspectral data – must be considered as the most critical error source.
(4) To what extent is the efficiency of the processing workflow in terms of processing times and accuracy influenced by alternative processing sequences and the introduction of data compression by image segmentation?
The influence of image segmentation and the processing workflow were discussed at several points in this work. In terms of accuracy, image segmentation did not prove a useful processing step for the classification of the HyMap data from Berlin: the comparison of pixel- and segment-based results in Chapter III showed that no ideal single aggregation level can be identified. Segment-based processing should therefore not be generally preferred over pixel-based approaches. Moreover, concepts are required which enable a simple integration of the positive but scale-dependent influences from multiple levels for the classification of heterogeneous areas. The multi-level approach introduced in Chapter III has been extended for agricultural areas by Waske and van der Linden (2008); tests on the HyMap data from Berlin will follow.
Despite the lower accuracy of classifications of single segment levels, the issue of data size reduction in such approaches is worth keeping in mind when setting up the workflow. Processing power of modern computers increases constantly. However, so do data volumes due to simultaneously increasing resolutions. Operational processing of very large data sets is often difficult without using data compression. Spectral compression, such as feature extraction by PC transformation, was shown to decrease accuracies when using SVM (compare research question 2). Image segmentation on the other hand appears to be a useful alternative for spatial compression, due to the preservation of hyperspectral characteristics, its high compression factor and little decrease in classification accuracy. The discussion of Chapter IV is round-up by the multiple assessments in Chapter V which were also applied to the segmented data at average segment size 13.1 pixels. The error sources identified on pixel-based data in Chapter V affected the segment-compressed data to the same extent. Thus, no additional drawbacks of segment-compressed analysis can be reported and the approach appears worthwhile when data size becomes very large and processing times during classification might be reduced by a factor of 70, for example.
The position of the geocoding in the processing workflow has to be seen in a similar context. Moving this processing step to the end of the workflow can be more time effective while being only slightly less accurate. When additional data sources are required during processing, it might even be useful to convert these data into raw image geometry – if possible. The suggestion by Schläpfer et al. (2007) to customize processing workflows according to the requirements of final end-user products is therefore supported by findings from this work.
The operational use of remote sensing approaches for monitoring the dynamic modifications of the Earth's surface by humans requires accurate and reliable data products. In order to generate such products, raw data with a high content of useful information are needed. The methods used to process these data have to make best use of the contained information while being simple in their approach, generally applicable, and capable of dealing with large data sets. The presented work addresses these requirements in several ways and investigates the general strengths and limitations of airborne hyperspectral data for mapping impervious land cover in urban areas. In this context it appeared useful to concentrate on five basic land cover classes on a relatively large and heterogeneous data set instead of optimizing a more differentiated classification scheme on a subset of the data.
Results from this work confirm on the one hand that urban areas are challenging for remote sensing approaches and in parts require special processing steps. On the other, they show that the accuracies of land cover maps and products derived from those depend not solely on the image classification but on several processing steps and on the general limitations of remote sensing.
The spectral information of the HyMap data allowed delineating the spectrally similar built-up and non built-up impervious surfaces at high accuracy. This differentiation is essential for many urban analyses (Cadenasso et al., 2007). Maps on impervious surface coverage can well be derived from such land cover information and inaccuracies of such maps do not relate to the spectral characteristics of the data. Thus, the additional spectral information content of hyperspectral data compared to multispectral data is valuable for urban applications and the usefulness for further development of more operational hyperspectral systems is underlined.
The high spectral classification accuracy can also be attributed to the strength of the SVM classifier. SVM have previously been compared to other classifiers by several authors (e.g.Huang et al., 2002;Pal and Mather, 2006). Similar comparisons based on the HyMap data from Berlin showed that SVM outperformed the traditional maximum likelihood classifier (Fu et al., 1969) and advanced decision tree classifications such as Random Forests (Breiman, 2001). All supervised classifiers depend to a great extent on the selection of training data. However, the good results in this work were achieved with a simple and time-saving training data collection scheme and SVM were shown to make best use of all provided data. Due to their good performance in the complex urban environment, SVM can be expected to be well suited for most other applications in a variety of environments. In the same way SVM allowed for a simple classification setup in this work, they can be expected to simplify other challenging classification problems, such as change detections where complex classes are frequent and training data for traditional classifiers is notoriously difficult to collect (e.g.Kuemmerle et al., 2008).
The limitations of the work with airborne hyperspectral data in urban areas are not mainly caused by the spectral characteristics but rather by the wide FOV or general problems of remote sensing: displaced buildings and occluded surfaces always exist when data is acquired at large view-angles; the surface underneath tree crowns is invisible for any optical sensor. This causes in parts the drawbacks of remote sensing data compared to data from field surveys (Miller and Small, 2003), especially for estimating absolute values of impervious surface coverage. However, the areas of most rapid urbanization are often those where no additional data are available. In this case remotely sensed maps are the best solution available. Quantifying errors associated with the final mapping product, as carried out in this thesis, is therefore of great interest and an important prerequisite to help end-users and decision makers in judging the reliability of their data.
Image processing in this work was challenging, due to the focus of the application, the characteristics of airborne hyperspectral data, and the complexity introduced by the heterogeneous urban environment. Thus, the methodological insights derived via mapping impervious areas in Berlin may be of relevance for many more – equally complex or more simplistic – applications. This includes not only applications using hyperspectral data from urban areas but also urban applications with data from other optical sensors or hyperspectral applications of non urban areas. This thesis is therefore an important step towards the broader application of remote sensing in urban areas. More studies of this kind will have to follow which deal with the research questions arising from the conclusions drawn above, studies that for example perform the step from land cover to land use or urban biotopes (Bochow et al., 2007) or that integrate SVM classification into the monitoring of the spatiotemporal growth of megacities.
"Urbanization brings with it both opportunities and challenges" (UN, 2006). The concentration of people is a response to the most dynamic economic activities in urban centers, which leads to various social and economic benefits. Concurrently, urban dwellers enjoy higher quality and more accessible health services. Cities are also at the forefront of political and cultural change. They are places where new ideas and products emerge and from which they spread. Thus, urbanization in less developed countries can be viewed as an indicator of development rather than a phenomenon with mainly negative consequences. It is a measure of globalization. (Sánchez-Rodríguez et al., 2005;UN, 2006)
Against this background, urbanization will continue and so will its major role in altering the ecosystems in cities and their surroundings (Kareiva et al., 2007). A better understanding of urbanization will help to understand its influence on local, regional and global ecosystem. Remote sensing and EO in general are of crucial importance in this context, particularly since urbanization is most dynamic in regions with little spatial information. Thus, in the immediate future the potential role for the application of remote sensing data alone is likely to be greater in cities in less developed regions than in cities in developed countries. Here the integration of remote sensing with other data types is likely to be most fruitful (Miller and Small, 2003).
From a technical perspective, new developments that appear interesting for urban applications can be reported in almost any acquisition domain. Various new sensors have recently become available or will become available in the near future: multispectral very high spatial resolution imagery decreases the mixed pixel problem in urban environments to the greatest extent (Ehlers, 2007); synthetic aperture radar (SAR) data of up to 1 m spatial resolution as acquired by TerraSAR-X (Stangl et al., 2006) will offer new opportunities for radar remote sensing in urban areas, but also lead to new, so far unknown challenges for data processing; the combination of terrain information from laserscanning or SAR with optical data can be expected to become more frequent (Gamba and Houshmand, 2002); and the extension of airborne hyperspectral sensors towards the thermal wavelengths regions will further improve the understanding of urban environments by means of remote sensing (Richter et al., 2005). The combined use of image data from such new sources will help to further improve the challenging applications addressed in this work. Especially the introduction of detailed information on surface elevation will help to delineate buildings, to improve the geocoding and to identify trees. This way the analysis of urban imperviousness can be expected to become more accurate. However, the results from the present work underpin that a detailed assessment of potential errors is essential when new data sources are used or data from different sources are combined.
Richards (2005) discusses the increasing abundance of data from different sources and mentions new approaches for the processing of multisource data as one of the greatest challenges in remote sensing. SVM can be expected to be of great importance for future urban applications. Their algorithmic development is still ongoing (e.g.Bazi and Melgani, 2006;Bruzzone et al., 2006) and they have already proved successfull in combining hyperspectral data and laserscanning information (Koetz et al., 2008) or data from SAR and optical sensors (Waske and van der Linden, 2008) in non-urban environments. Although results achieved by SVM are convincing, traditional classifiers are still widely used and users hesitate to integrate rather recent machine learning developments into their analysis. To broaden the community of users of such more effective approaches, user-oriented implementations are needed that are optimized for the requirements remote sensing applications and minimize the number of parameters to be set (e.g.Janz et al., 2007).
Despite the constant improvement in data quality and new methodological developments, urban remote sensing alone can not advance the knowledge on the impacts of urbanization. It will, however, play an important role in many future approaches that further integrate ecological and social sciences. It is this integration of different disciplines that it is critical to move beyond the existing approaches for studying coupled systems, to develop more comprehensive portfolios, and to build an international research network spanning local, regional, national, and global levels (Liu et al., 2007). Studying the complex process of urbanization requires such integrated approaches.
|© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.|
|DiML DTD Version 4.0||Zertifizierter Dokumentenserver|
der Humboldt-Universität zu Berlin