Multi Temporal Analysis of Floods and Tsunami Effects: Annotation and Quantitative Analysis

摘要: This paper addresses the problem of multitemporal analysis an available TerraSAR-X data time series covering Sendai region in order to assess flood extent and damages caused by Tohuku-oki tsunami. Over last decade use Earth Observation satellites support disaster emergency relief has considerably grown. In fully exploit highresolution satellite images, a method based on patches (each image is divided into non-overlapping tiles) proposed extract relevant contextual information. The local features each patch act as compact content descriptor. Further on, considering descriptors, next step cluster find similar semantic classes. SVM classifier implements concept query example using content. results include well-defined classes, derived through semiautomatic methods thus developing effective approach multitemporal analysis.