Oil spill detection from SAR image data for remote monitoring of marine pollution using light weight imageJ implementation

Biogenic and Mineral oil spill in the ocean can harm the marine ecosystem significantly. Past research suggests image processing on images generated from Synthetic Aperture Radar (SAR) data are the most suitable way of remote oil spill detection. In this paper, the framework for oil spill detection is explained in three step process for oil spill detection viz., dark spot detection, feature extraction, and classification. The paper focuses on Hysteresis algorithm for segmenting the dark spots and decision tree for the classification of dark spots and look alikes. The major steps in oil spill detection have been implemented in imageJ which gives a light weight solution to existing softwares for the said application.