ABSTRACT:This paper presents a method for the classification of satellite images into multiple predefined land cover classes. The proposed approach results in a fully automatic segmentation and classification of each pixel, using a small amount of training data. Therefore, semantic segmentation techniques are used, which are already successful applied to other computer vision tasks like facade recognition. We explain some simple modifications made to the method for the adaption of remote sensing data. Besides local features, the proposed method also includes contextual properties of multiple classes. Our method is flexible and can be extended for any amount of channels and combinations of those. Furthermore, it is possible to adapt the approach to several scenarios, different image scales, or other earth observation applications, using spatially resolved data. However, the focus of the current work is on high resolution satellite images of urban areas. Experiments on a QuickBird-image and LiDAR data of the city of Rostock show the flexibility of the method. A significant better accuracy can be achieved using contextual features.
This paper focuses on the description and demonstration of a simple, but effective object-based image analysis (OBIA) approach to extract urban land cover information from high spatial resolution (HSR) multi-spectral and light detection and ranging (LiDAR) data. Particular emphasis is put on the evaluation of the proposed method with regard to its generalization capabilities across varying situations. For this purpose, the experimental setup of this work includes three urban study areas featuring different physical structures, four sets of HSR optical and LiDAR input data, as well as statistical measures to enable the assessment of classification accuracies and methodological transferability. The results of this study highlight the great potential of the developed approach for accurate, robust and large-area mapping of urban environments. User's and producer's accuracies observed for all maps are almost consistently above 80%, in many cases even above 90%. Only few larger class-specific errors occur mainly due to the simple assumptions on which the method is based. The presented feature extraction workflow can therefore be used as a template or starting point in the framework of future urban land cover mapping efforts.Index Terms-Accuracy, data fusion, land cover, multi-sensor, object-based image analysis (OBIA), transferability, urban.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.