Showing papers by "Björn Waske published in 2014"

Can I Trust My One-Class Classification?

Abstract: Contrary to binary and multi-class classifiers, the purpose of a one-class classifier for remote sensing applications is to map only one specific land use/land cover class of interest. Training these classifiers exclusively requires reference data for the class of interest, while training data for other classes is not required. Thus, the acquisition of reference data can be significantly reduced. However, one-class classification is fraught with uncertainty and full automatization is difficult, due to the limited reference information that is available for classifier training. Thus, a user-oriented one-class classification strategy is proposed, which is based among others on the visualization and interpretation of the one-class classifier outcomes during the data processing. Careful interpretation of the diagnostic plots fosters the understanding of the classification outcome, e.g., the class separability and suitability of a particular threshold. In the absence of complete and representative validation data, which is the fact in the context of a real one-class classification application, such information is valuable for evaluation and improving the classification. The potential of the proposed strategy is demonstrated by classifying different crop types with hyperspectral data from Hyperion.

Mapping Land Management Regimes in Western Ukraine Using Optical and SAR Data

Abstract: The global demand for agricultural products is surging due to population growth, more meat-based diets, and the increasing role of bioenergy. Three strategies can increase agricultural production: (1) expanding agriculture into natural ecosystems; (2) intensifying existing farmland; or (3) recultivating abandoned farmland. Because agricultural expansion entails substantial environmental trade-offs, intensification and recultivation are currently gaining increasing attention. Assessing where these strategies may be pursued, however, requires improved spatial information on land use intensity, including where farmland is active and fallow. We developed a framework to integrate optical and radar data in order to advance the mapping of three farmland management regimes: (1) large-scale, mechanized agriculture; (2) small-scale, subsistence agriculture; and (3) fallow or abandoned farmland. We applied this framework to our study area in western Ukraine, a region characterized by marked spatial heterogeneity in management intensity due to the legacies from Soviet land management, the breakdown of the Soviet Union in 1991, and the recent integration of this region into world markets. We mapped land management regimes using a hierarchical, object-based framework. Image segmentation for delineating objects was performed by using the Superpixel Contour algorithm. We then applied Random Forest classification to map land management regimes and validated our map using randomly sampled in-situ data, obtained during an extensive field campaign. Our results showed that farmland management regimes were mapped reliably, resulting in a final map with an overall accuracy of 83.4%. Comparing our land management regimes map with a soil map revealed that most fallow land occurred on soils marginally suited for agriculture, but some areas within our study region contained considerable potential for recultivation. Overall, our study highlights the potential for an improved, more nuanced mapping of agricultural land use by combining imagery of different sensors.

Superpixel-based classification of hyperspectral data using sparse representation and conditional random fields

Abstract: This paper presents a superpixel-based classifier for landcover mapping of hyperspectral image data. The approach relies on the sparse representation of each pixel by a weighted linear combination of the training data. Spatial information is incorporated by using a coarse patch-based neighborhood around each pixel as well as data-adapted superpixels. The classification is done via a hierarchical conditional random field, which utilizes the sparse-representation output and models spatial and hierarchical structures in the hyperspectral image. The experiments show that the proposed approach results in superior accuracies in comparison to sparse-representation based classifiers that solely use a patch-based neighborhood.

Import Vector Machines for Quantitative Analysis of Hyperspectral Data

Abstract: In this letter we explore probabilities derived from an import vector machines (IVM) classifier as quantitative measures of class proportion. We have developed a parameter selection strategy that improves the description of class proportions. This strategy incorporates the use of spectral mixtures, which represent gradual class transitions, into the parameter selection process. In addition, we evaluated the sensitivity of our approach in regard to increasing training uncertainty and signal-to-noise ratio. The approach was tested for binary, two-class problems on hyperspectral in situ measurements. The IVM models generated with our parameter selection strategy achieved similar or even improved classification accuracies compared to parameter selection with the standard IVM classification approach. Furthermore, the respective class probabilities correlated highly with reference class proportions. This new strategy is less affected by the inclusion of random noise and relatively stable against increased training errors.

Synergies from SAR-Optical Data Fusion for LULC Mapping

Abstract: Nowadays Earth Observation (EO) systems play a major role in supporting environmental programs and monitoring compliances, such as the European Global Monitoring for Environment and Security (GMES – Copernicus) program. Copernicus aims on the provision of reliable and current information on our planet and its environmental state in three Earth system domains “Land”, “Atmosphere”, and “Marine”. Moreover, the corresponding services support the management of humanitarian crises, natural disasters and man-made crisis (Aschbacher J, Milagro-Perez MP, Remote Sens Environ 120:3–8, 2012). Products in the Land-domain, comprise accurate and cross-border harmonized information on land cover and land cover change, including information on seasonal and annual changes, the vegetation state and the monitoring of the water cycle. Overall these products will support decision-making and various monitoring applications in context of land use and land cover change, water quality, spatial planning, and global food security (GMES, http://www.gmes.info/pages-principales/services/land-monitoring, Last accessed May 2012, 2012).

Shapelet-based sparse image representation for landcover classification of hyperspectral data

Abstract: This paper presents a novel sparse representation-based classifier for landcover mapping of hyperspectral image data. Each image patch is factorized into segmentation patterns, also called shapelets, and patch-specific spectral features. The combination of both is represented in a patch-specific spatial-spectral dictionary, which is used for a sparse coding procedure for the reconstruction and classification of image patches. Hereby, each image patch is sparsely represented by a linear combination of elements out of the dictionary. The set of shapelets is specifically learned for each image in an unsupervised way in order to capture the image structure. The spectral features are assumed to be the training data. The experiments show that the proposed approach shows superior results in comparison to sparse-representation based classifiers that use no or only limited spatial information and behaves competitive or better than state-of-the-art classifiers utilizing spatial information and kernelized sparse representation-based classifiers.

Pattern Recognition and Classification.

Abstract: Definition The classification of remote sensing images and the corresponding generation of land cover maps are perhaps the most common applications in remote sensing. In general, the aim of a land cover classification is the assignment of each pixel within the imagery to a specific information class (e.g., forest areas). In general, this is performed by methods of machine learning and pattern recognition. Pattern recognition can be defined as a technique to classify data (patterns) based either on a priori knowledge or statistical information extracted from the patterns.