AUTOMATED FEATURE EXTRACTION BY COMBINING POLARIMETRIC SAR AND OBJECT-BASED IMAGE ANALYSIS FOR MONITORING OF NATURAL RESOURCE EXPLOITATION

摘要

An automated feature extraction procedure based on therncombination of a pixel-based unsupervised classificationrnof polarimetric synthetic aperture radar data (co-corndual-polarimetric TerraSAR-X) and an object-basedrnpost-classification is presented. The former is based onrnthe entropy/alpha decomposition and the hereon basedrnunsupervised Wishart classification, while the latterrnconsiders in addition feature properties such as shapernand area. The feature extraction procedure is developedrnfor monitoring oil field infrastructure. For developingrncountries, several studies reported a high correlationrnbetween the dependence of oil exports and violent conflicts.rnConsequently, to support problem solving, anrnindependent monitoring of the oil field infrastructure byrnEarth observation is proposed.