Object‐oriented classification for urban land cover mapping with ASTER imagery

This document demonstrates the potential of using an object‐oriented approach to map urban land cover. One objective of this work was to test the ability of the object‐oriented classification in the generation of urban land cover maps. Anotehr was to produce an updated land cover map for the city of Beijing from Advanced Spaceborne Thermal Emission and Reflecton Radiometer (ASTER) data, with an evaluation of its accuracy.     

Object‐oriented and textural image classification of the Siberia GBFM radar mosaic combined with MERIS imagery for continental scale land cover mapping

Boreal forests and wetlands play an important role in the climate system, in particular through biosphere–atmosphere flux exchanges. They are an important pool of carbon and their role as sink or source of greenhouse gases is not fully understood. Accurate mapping of the vegetation of Siberia can therefore contribute to a better understanding of these processes at regional scale and of their effects on the climate through regional biosphere modeling. The potential of the combination of radar data with medium‐resolution optical data to obtain regional‐scale land cover mapping is investigated using multi‐spectral imagery from the MERIS sensor at 300 m resolution and a high resolution radar mosaic (pixel spacing of 100 m) covering Western and Eastern Siberia compiled in the framework of the Global Boreal Forest Mapping project. For this purpose, capabilities of oriented‐object image analysis associated to wavelet multi‐resolution techniques are investigated. Results show that wavelet multi‐resolution textures bring relevant additional information for land cover classification. Suggestions are made for the implementation of an object‐based wavelet multi‐resolution texture estimator.     

Bayesian Multi‐net Classifier for classification of remote sensing data

The purpose of this paper is to investigate the applicability of Bayesian Multi‐net Classifier (BMC) to classify remote sensing data. BMC is based on Bayesian Network (BN), which is a graphical model encoding probabilistic relationships among variables of interest. Different from the BNC that has a mere network, a BMC has as many local Bayesian Networks as the predefined classes, which means that the probabilistic relationships among the features can be different for different classes. Classification is done by computing the probability of the class, given the particular instance of the features, and then predicting the class with the highest posterior probability. This method was validated using a Landsat ETM+ image of Beijing acquired on 1 May 2003. Based on the confusion matrix, overall accuracy, Kappa statistic, total normalized probability of misclassification (TNPM), and McNemar's test, classification results of BMC were compared with those of MLC and BNC in the case study. The comparison results show that BMC performs slightly better than MLC and similar to BNC. The local Bayesian Networks of BMC can also lead to a better understanding of the dependencies between bands for different classes.     

A method for classification accuracy evaluation for a precipitation‐oriented classifier

A method of classification accuracy evaluation for a cloud and precipitation classifier applied to geostationary meteorological satellite data is presented. The method has been developed to evaluate the accuracy of a rather precise classification algorithm. The algorithm produces nine classes, four of which involve precipitation. The classes are: (1) clear or insignificant cloud, (2) low thin cloud with no rain, (3) low or middle thin cloud with no rain, (4) low or middle thick cloud with no rain, (5) middle or high cloud with no rain, (6) middle or high cloud with the possibility of rain, (7) middle or high cloud with light–moderate precipitation, (8) middle–high cloud with moderate–heavy precipitation, (9) heavy thunderstorm. The evaluation classifier has been tested for its accuracy (ground truth) using comparison between actual meteorological weather reports and classification results derived from the algorithm applied. For the estimation of classification accuracy, the omission/commission method is applied between the observed and the classification‐produced values. The classifier used has proved to be very reliable for classifying major cloud types and precipitation, tested during the synoptic situation of depression systems approaching the south Balkan Peninsula from the west. In that synoptic situation, different intensities of rainfall as well as heavy thunderstorm were present, and the results are very satisfactory. The method can be used to evaluate classification results produced by algorithms applied to meteorological satellite data, classifying precipitation areas as well as the heaviness of precipitation.     

Envisat multi‐polarized ASAR data for flood mapping

During the August 2002 Elbe river flood, different satellite sensor data were acquired, and especially Envisat Advanced Synthetic Aperture Radar (ASAR) data. The ASAR instrument was activated in Alternating Polarization (AP) and Image (IM) modes, providing high resolution datasets. Thus, the comparison with a quasi‐simultaneous ERS‐2 scene enables the evaluation of the contribution of polarization configurations to flood boundary delineation. This study highlights the increased capabilities of the Envisat ASAR instrument in flood mapping, especially the benefit of combining like‐ and cross‐polarizations for rapid mapping within a crisis context.     

RVM‐based multi‐class classification of remotely sensed data

The relevance vector machine (RVM), a Bayesian extension of the support vector machine (SVM), has considerable potential for the analysis of remotely sensed data. Here, the RVM is introduced and used to derive a multi‐class classification of land cover with an accuracy of 91.25%, a level comparable to that achieved by a suite of popular image classifiers including the SVM. Critically, however, the output of the RVM includes an estimate of the posterior probability of class membership. This output may be used to illustrate the uncertainty of the class allocations on a per‐case basis and help to identify possible routes to further enhance classification accuracy.     

Object‐based classification of residential land use within Accra,Ghana based on QuickBird satellite data

A segmentation and hierarchical classification approach applied to QuickBird multispectral satellite data was implemented, with the goal of delineating residential land use polygons and identifying low and high socio‐economic status of neighbourhoods within Accra, Ghana. Two types of object‐based classification strategies were tested, one based on spatial frequency characteristics of multispectral data, and the other based on proportions of Vegetation–Impervious–Soil sub‐objects. Both approaches yielded residential land‐use maps with similar overall percentage accuracy (75%) and kappa index of agreement (0.62) values, based on test objects from visual interpretation of QuickBird panchromatic imagery.     

An improved SVM method P‐SVM for classification of remotely sensed data

A support vector machine (SVM) is a mathematical tool which is based on the structural risk minimization principle. It tries to find a hyperplane in high dimensional feature space to solve some linearly inseparable problems. SVM has been applied within the remote sensing community to multispectral and hyperspectral imagery analysis. However, the standard SVM faces some technical disadvantages. For instance, the solution of an SVM learning problem is scale sensitive, and the process is time‐consuming. A novel Potential SVM (P‐SVM) algorithm is proposed to overcome the shortcomings of standard SVM and it has shown some improvements. In this letter, the P‐SVM algorithm is introduced into multispectral and high‐spatial resolution remotely sensed data classification, and it is applied to ASTER imagery and ADS40 imagery respectively. Experimental results indicate that the P‐SVM is competitive with the standard SVM algorithm in terms of accuracy of classification of remotely sensed data, and the time needed is less.     

Integration of multitemporal/polarization C‐band SAR data sets for land‐cover classification

This paper investigates the potential of multitemporal/polarization C‐band SAR data for land‐cover classification. Multitemporal Radarsat‐1 data with HH polarization and ENVISAT ASAR data with VV polarization acquired in the Yedang plain, Korea are used for the classification of typical five land‐cover classes in an agricultural area. The presented methodologies consist of two analytical stages: one for feature extraction and the other for classification based on the combination of features. Both a traditional SAR signal property analysis‐based approach and principal‐component analysis (PCA) are applied in the feature extraction stage. Special concerns are in the interpretation of each principal component by using principal‐component loading. The tau model applied as a decision‐level fusion methodology can provide a formal framework in which the posteriori probabilities derived from different sensor data can be combined. From the case study results, the combination of PCA‐based features showed improved classification accuracy for both Radarsat‐1 and ENVISAT ASAR data, as compared with the traditional SAR signal property analysis‐based approach. The integration of PCA‐based features based on multiple polarization (i.e. HH from Radarsat‐1, and both VV and VH from ENVISAT ASAR) and different incidence angles contributed to a significant improvement of discrimination capability for dry fields which could not be properly classified by using only Radarsat‐1 or ENVISAT ASAR data, and thus showed the best classification accuracy. The results of this case study indicate that the use of multiple polarization SAR data with a proper feature extraction stage would improve classification accuracy in multitemporal SAR data classification, although further consideration should be given to the polarization and incidence angle dependency of complex land‐cover classes through more experiments.     

SVM‐based segmentation and classification of remotely sensed data

Support Vector Machines (SVM) is becoming a popular alternative to traditional image classification methods because it makes possible accurate classification from small training samples. Nevertheless, concerns regarding SVM parameterization and computational effort have arisen. This Letter is an evaluation of an automated SVM‐based method for image classification. The method is applied to a land‐cover classification experiment using a hyperspectral dataset. The results suggest that SVM can be parameterized to obtain accurate results while being computationally efficient. However, automation of parameter tuning does not solve all SVM problems. Interestingly, the method produces fuzzy image‐regions whose contextual properties may be potentially useful for improving the image classification process.     

A land‐cover classification for landslide susceptibility mapping by using feature components

Classifying original bands and/or image components may cause unsatisfactory results in fields that have heterogeneous reflectance. In such cases, the demand for accurate land‐use, land‐cover, vegetation, and forestry information may require more specific components. The components should represent peculiar information collected from several inputs for target land covers. In this study, a new technique of land‐cover classification was explored to prepare an input which increases the success of landslide susceptibility mapping in a subtropical region, Asarsuyu Catchment Area (Duzce). Land‐cover mapping is a difficult issue in this area by only carrying out field studies and aerial‐photo interpretations. Moreover, applying different classifications of Landsat Thematic Mapper bands and/or their secondary products does not produce acceptable results. For this reason, vegetation indices, soil/surface moisture indices, topographic wetness index and drainage density were calculated to produce feature representative components for the land‐cover classification process. Results obtained from the proposed technique show that feature representative components significantly improve the conventional classification accuracy from 77% to 89% and the resultant land‐cover map is such a valuable input for landslide susceptibility mapping that it increases the success of the landslide susceptibility map from 63% to 88%.     

Super‐resolution mapping of the waterline from remotely sensed data

Methods for mapping the waterline at a subpixel scale from a soft image classification of remotely sensed data are evaluated. Unlike approaches based on hard classification, these methods allow the waterline to run through rather than between image pixels and so have the potential to derive accurate and realistic representations of the waterline from imagery with relatively large pixels. The most accurate predictions of waterline location were made from a geostatistical approach applied to the output of a soft classification (RMSE = 2.25 m) which satisfied the standards for mapping at 1 : 5000 scale from imagery with a 20 m spatial resolution.     

Object‐based land‐cover classification for the Phoenix metropolitan area: optimization vs. transportability

An object‐based approach was utilized in the development of two urban land‐cover classification schemes on high‐resolution (0.6 m), true‐colour aerial photography of the Phoenix metropolitan area, USA. An initial classification scheme was heavily weighted by standard nearest‐neighbour (SNN) functions generated by samples from each of the classes, which produced an enhanced accuracy (84%). A second classification was developed from the initial classification scheme in which SNN functions were transformed into a fuzzy‐rule set, creating a product transportable to different areas of the same imagery, or for land‐cover change detection with similar imagery. A comprehensive accuracy assessment revealed a slightly lower overall accuracy (79%) for the rule‐based classification. We conclude that the transportable classification scheme is satisfactory for general land‐cover analyses; yet classification accuracy can be enhanced at site‐specific venues with the incorporation of nearest‐neighbour functions using class samples.     

Textural and local spatial statistics for the object‐oriented classification of urban areas using high resolution imagery

Textural and local spatial statistical information is important in the classification of urban areas using very high resolution imagery. This paper describes the utility of textural and local spatial statistics for the improvement of object‐oriented classification for QuickBird imagery. All textural/spatial bands were used as additional bands in the supervised object‐oriented classification. The texture analysis is based on two levels: segmented image objects and moving windows across the whole image. In the texture analysis over image objects, the angular second moment textural feature at a 45° angle showed an improved classification performance with regard to buildings, depicting the patterns of buildings better than any other directions. The texture analysis based on moving windows across the whole image was conducted with various window sizes (from 3×3 to 13×13), and four grey‐level co‐occurrence matrix (GLCM) textural features (homogeneity, contrast, angular second moment, and entropy) were calculated. The contrast feature with the 7×7 window size improved classification up to 6%. One type of local spatial statistics, Moran's I feature with the vertical neighbourhood rule, improved the classification accuracy even further, up to 7%. Comparison of results between spectral and spectral+textural/spatial information indicated that textural and spatial information can be used to improve the object‐oriented classification of urban areas using very high resolution imagery.     

Information fusion approach for the data classification: an example for ERS‐1/2 InSAR data

A huge amount of various remote sensing data have been acquired and archived during recent years. Information extraction from these data is still a challenging task, for example using the data classification. We propose the Bayesian approach to image classification using information fusion from different sources of data. The method of classification is based on the three processing steps: (1) information fission by feature extraction, (2) data and dimensionality reduction by unsupervised clustering, and (3) supervised classification with information fusion. The potential of the classification method is illustrated by the examples on ERS‐1/2 Tandem interferometric synthetic aperture radar data. The continuity of tandem pairs of SAR images is ensured by already started or future missions such as TerraSAR‐X, TanDEM‐X, and COSMO‐SkyMed.     

Design‐code traceability for object‐oriented systems

Traceability is a key issue to ensure consistency among software artifacts of subsequent phases of the development cycle. However, few works have so far addressed the theme of tracing object oriented (OO) design into its implementation and evolving it. This paper presents an approach to checking the compliance of OO design with respect to source code and support its evolution. The process works on design artifacts expressed in the OMT (Object Modeling Technique) notation and accepts C++ source code. It recovers an “as is” design from the code, compares the recovered design with the actual design and helps the user to deal with inconsistencies. The recovery process exploits the edit distance computation and the maximum match algorithm to determine traceability links between design and code. The output is a similarity measure associated to design‐code class pairs, which can be classified as matched and unmatched by means of a maximum likelihood threshold. A graphic display of the design with different green levels associated to different levels of match and red for the unmatched classes is provided as a support to update the design and improve its traceability to the code.     

Sequential masking classification of multi‐temporal Landsat7 ETM+ images for field‐based crop mapping in Karacabey,Turkey

Three Landsat7 ETM+ images acquired in May, July and August during the 2000 crop growing season were used for field‐based mapping of summer crops in Karacabey, Turkey. First, the classification of each image date was performed on a standard per pixel basis. The results of per pixel classification were integrated with digital agricultural field boundaries and a crop type was determined for each field based on the modal class calculated within the field. The classification accuracy was computed by comparing the reference data, field‐by‐field, to each classified image. The individual crop accuracies were examined on each classified data and those crops whose accuracy exceeds a preset threshold level were determined. A sequential masking classification procedure was then performed using the three image dates, excluding after each classification the class properly classified. The final classified data were analysed on a field basis to assign each field a class label. An immediate update of the database was provided by directly entering the results of the analysis into the database. The sequential masking procedure for field‐based crop mapping improved the overall accuracies of the classifications of the July and August images alone by more than 10%.     

Semi‐automated methods for mapping wetlands using Landsat ETM+ and SRTM data

The overarching goal of this study was to develop a comprehensive methodology for mapping natural and human‐made wetlands using fine resolution Landsat enhanced thematic mapper plus (ETM+), space shuttle radar topographic mission digital elevation model (SRTM DEM) data and secondary data. First, automated methods were investigated in order to rapidly delineate wetlands; this involved using: (a) algorithms on SRTM DEM data, (b) thresholds of SRTM‐derived slopes, (c) thresholds of ETM+ spectral indices and wavebands and (d) automated classification techniques using ETM+ data. These algorithms and thresholds using SRTM DEM data either over‐estimated or under‐estimated stream densities (S _d) and stream frequencies (S _f), often generating spurious (non‐existent) streams and/or, at many times, providing glaring inconsistencies in the precise physical location of the streams. The best of the ETM+‐derived indices and wavebands either had low overall mapping accuracies and/or high levels of errors of omissions and/or errors of commissions.

Second, given the failure of automated approaches, semi‐automated approaches were investigated; this involved the: (a) enhancement of images through ratios to highlight wetlands from non‐wetlands, (b) display of enhanced images in red, green, blue (RGB) false colour composites (FCCs) to highlight wetland boundaries, (c) digitizing the enhanced and displayed images to delineate wetlands from non‐wetlands and (d) classification of the delineated wetland areas into various wetland classes. The best FCC RGB displays of ETM+ bands for separating wetlands from other land units were: (a) ETM+4/ETM+7, ETM+4/ETM+3, ETM+4/ETM+2, (b) ETM+4, ETM+3, ETM+5 and (c) ETM+3, ETM+2, ETM+1. In addition, the SRTM slope threshold of less than 1% was very useful in delineating higher‐order wetland boundaries. The wetlands were delineated using the semi‐automated methods with an accuracy of 96% as determined using field‐plot data.

The methodology was evaluated for the Ruhuna river basin in Sri Lanka, which has a diverse landscape ranging from sea shore to hilly areas, low to very steep slopes (0° to 50°), arid to semi‐arid zones and rain fed to irrigated lands. Twenty‐four per cent (145 733 ha) of the total basin area was wetlands as a result of a high proportion of human‐made irrigated areas, mainly under rice cropping. The wetland classes consisted of irrigated areas, lagoons, mangroves, natural vegetation, permanent marshes, salt pans, lagoons, seasonal wetlands and water bodies. The overall accuracies of wetland classes varied between 87% and 94% (K _hat = 0.83 to 0.92) with errors of omission less than 13% and errors of commission less than 1%.     

A comparison of classification algorithms using Landsat-7 and Landsat-8 data for mapping lithology in Canada’s Arctic

To map Arctic lithology in central Victoria Island, Canada, the relative performance of advanced classifiers (Neural Network (NN), Support Vector Machine (SVM), and Random Forest (RF)) were compared to Maximum Likelihood Classifier (MLC) results using Landsat-7 and Landsat-8 imagery. A ten-repetition cross-validation classification approach was applied. Classification performance was evaluated visually and statistically using the global classification accuracy, producer’s and user’s accuracies for each individual lithological/spectral class, and cross-comparison agreement. The advanced classifiers outperformed MLC, especially when training data were not normally distributed. The Landsat-8 classification results were comparable to Landsat-7 using the advanced classifiers but differences were more pronounced when using MLC. Rescaling the Landsat-8 data from 16 bit to 8 bit substantially increased classification accuracy when MLC was applied but had little impact on results from the advanced classifiers.     

DEM‐optical‐radar data integration for palaeohydrological mapping in the northern Darfur,Sudan: implication for groundwater exploration

North‐western Sudan, as a part of the eastern Sahara, is among the driest places on earth. However, the region underwent drastic climatic changes through the alternation of dry and wet conditions in the past. During humid phases, when the rain was plentiful over a prolonged time period, the surface was veined by rivers and dotted by large lakes. The new Shuttle Radar Topography Mission data (SRTM ～90 m) revealed a large endorheic drainage basin, which is centred by a large terminal palaeolake, in the northern Darfur State. The use of GIS methods allowed the delineation of the drainage basin and its associated palaeorivers. The SRTM data along with the Landsat (ETM+) and Radarsat‐1 images corroborate the presence of segments of palaeoshorelines associated with the palaeolake highstands. These constitute a convincing argument of the long‐term existence of a possible pre‐Holocene large water body in the region in the past. The remains of the highest palaeoshoreline have a constant altitude of 573±3 m asl. At its maximum extent, the mega Lake occupied an area of about 30 750 km² (the same size as the Great Bear Lake, Canada's largest lake), which would have contained approximately 2530 km³ of water. This, ancestral lake, which we named the Northern Darfur Megalake (ND Megalake), represents indisputable evidence of the past pluvial conditions in the eastern Sahara. The discovered palaeoshorelines will have significant consequences for improving our knowledge of continental climate change and regional palaeohydorology, and should be taken into consideration in studies of past human habitation in the region. Much of the water carried by the Northern Darfur palaeorivers and the ND Megalake would have percolated into the underlying rocks feeding the Nubian Sandston aquifer. These findings show that the used approach of space‐data integration can help significantly in the groundwater exploration efforts in the Darfur region, where freshwater access is essential for refugee survival, and can be successfully adopted in other parts of Sudan and arid lands in general.