Remote sensing satellite imagery and risk management: image based information extraction

none2The introduction of high and very high resolution multispectral satellite imagery, characterized by ground resolution from one to few meters, has lead to a new perspective in processes of risk estimation, mitigation and management. In particular, the possibility to obtain in a very short time a wide-scene image of areas subjected to a crisis has become useful both for emergency management and for effective damage estimate. Examples of images in the immediate aftermath of natural disasters like earthquakes, hurricanes, floods, fires, tsunami, volcanic eruptions, etc, have been also globally distributed trough information media and web based image systems such as Google Earth. However, a concrete possibility to extract quantitative information from such kind of images is subject to several factors: first of all, accessibility in term of timing in image acquisition and in image delivering to stakeholders, followed by image quality (resolution, absence of clouds, geo-location, informative content, etc.), and finally by the information extraction process. After a review of the problem and of the current situation in terms of available data and applications, the paper focus on an approach for information extraction and quantitative image classification, mainly by object oriented analysis. One of the major objectives is to show the possibility to obtain the outlines and the count of the buildings for an area interested by a severe disaster, saving human time in image interpretation in procedures of risk management and in estimating rebuilding costs; the integration of information coming from existing building databases is also considered. The experiences show the potential of the method and promising results in the classification accuracy.The papers published on this book, realized by scientists and researchers coming from all the world, cover all aspects of risk analysis and hazard mitigation, ranging from specific assessment of risk to mitigation associated with both natural and anthropogenic hazards. The advances in computational methods and models to quantify hazards with high accuracy, combined with GIS systems which can provide spatial distribution of hazards involved, enhances the ability to perform an accurate risk analysis and offers more effective risk management. These developments are important for all areas of human endeavour and have particular relevance to environmental issues where the risks involved are substantial.noneBitelli G.; Gusella L.Bitelli G.; Gusella L