Dual-Polarimetric Feature Extraction and Evaluation for Oil Spill Detection: A Near Real Time Perspective

Abstract

Exploitation of polarimetric features for oil spill detection is relatively new and until recently those properties have not been used for operational services. In last decade, a number of semi-automatic and automatic techniques have been proposed in order to differentiate oil spill and look alike spots based on single pol SAR data, however these techniques suffer from high miss-classification rate which is completely undesirable for operational services. In addition to that, small operational spillages from offshore platforms are often ignored as it appears insignificant on traditional ‘ScanSAR’ images. In order to mitigate this situation a major focus of research in this area is the development of automated algorithms based on polarimetric images to distinguish oil spills from ‘look-alikes’. This paper describes the development of a Near Real Time (NRT) oil spill detection processing chain based on dual-polarimetric TerraSAR-X images combining traditional and polarimetric features (e.g. Lexicographic and Pauli Based features). Total number of 35 feature parameters were extracted and evaluated for oil spill and look alike discrimination; including both single polarimetric and dual polarimetric based features from each segmented dark spot. Extracted features are then used for training and validation of a decision tree classifier. Initial performance estimation was carried out for the proposed methodology in order to evaluate its suitability for NRT operational service.