MASADA user guide : version 1.3

This user guide accompanies the MASADA tool which is a public tool for the detection of built-up areas from remote sensing data. MASADA stands for Massive Spatial Automatic Data Analytics. It has been developed in the frame of the "Global Human Settlement Layer" (GHSL) project of the European Commission's Joint Research Centre, with the overall objective to support the production of settlement layers at regional scale, by processing high and very high resolution satellite imagery. The tool builds on the Symbolic Machine Learning (SML) classifier; a supervised classification method of remotely sensed data which allows extracting built-up information using a coarse resolution settlement map or a land cover information for learning the classifier. The image classification workflow incorporates radiometric, textural and morphological features as inputs for information extraction. Though being originally developed for built-up areas extraction, the SML classifier is a multi-purpose classifier that can be used for general land cover mapping provided there is an appropriate training data set. The tool supports several types of multispectral optical imagery. It includes ready-to-use workflows for specific sensors, but at the same time, it allows the parametrization and customization of the workflow by the user. Currently it includes predefined workflows for SPOT-5, SPOT-6/7, RapidEye and CBERS-4, but it was also tested with various high and very high resolution sensors like GeoEye-1, WorldView-2/3, Pléiades and Quickbird. The software is designed for the processing of single scenes but also for batch processing of large data sets. This user guide is a comprehensive guide to all aspects of using the MASADA tool. It includes instructions for the set-up of the software, the use of the tool and the manipulation of the data. It presents briefly the basic principles and background information on the different algorithms and modules of the methodology. Some guidelines on the parametrization of the modules are also provided and the inclusion of additional features in the sensor-oriented workflows.