Near real time operational oil spill detection service using a classification tree

Abstract

Today the health of ocean is in danger due to over offshore oil exploration and increasing maritime traffic. Operational activities show regular occurrence of accidental and deliberate oil spill over major European shipping route and offshore platform locations. European oil spill detection service, ‘CleanSeaNet’ currently uses manual image interpretation technique in order to report oil spill to its member states. Anticipating regular and large amount of data from ESA’s Sentinal-1 mission (under Copernicus Service), a major focus of research in this area is the development of automated/semi-automated algorithms to distinguish oil spills from ‘look-alikes’ complementing the visual analysis carried out by current operational services. This paper describes the development of an semi-automated approach for oil spill detection from TerraSAR-X images using classification tree in Near Real Time (NRT) environment A total number of 8 feature parameters were extracted from 143 segmented dark-spot (oil spill and ‘look-alike’) representing different characteristic, which are then used to train the proposed Classification tree. An initial evaluation of this methodology has been carried out on a large dataset and reported