Oil Spill Detection using SAR Data Feature Based to Deep Learning Approaches

Abstract

Ocean surface monitoring is one of the important applications to analyze and ensure newlinebalance in the marine ecosystem. Several occasions of intentional and unintentional newlineoil discharges have created discontinuity in marine life s smooth existence and people newlinedwelling in the coastal areas. It is impractical for the human force to timely monitor newlinethe disasters of oil spillage offshore. Due to the increasing distress concerning marine newlinelife protection for substantial development, a project under Indian Space Research newlineOrganization titled Oceanic pollution and other ocean phenomenon monitoring using newlinefeature extraction from multi-polarized SAR data is developed with one of its newlinedominant application as oil spill detection. This thesis aims to assist human forces newlinein coastal relief zones to get timely alerts with on-time oil spill detection. The context newlineof my thesis is to develop optimized feature extraction technique for extracting newlinesignificant features from SAR images for oil spill detection. Optimized feature extraction newlinetechniques are developed to classify the ocean surface features into five major newlineclasses, namely, oil, look-alike, ship, land, and clean sea. The ship provides ancillary newlineinformation to confirm whether the detected dark region is oil or look-alike. For our newlineexperimentation we have considered SAR data from various sensors such as ALOS, newlineRISAT-1, and SENTINAL-1. newlineBoth Level-1 and Level-2 SAR data are researched to boil down to the applicability newlineof the type of data that suits the best for ocean surface monitoring, especially oil spill newlinedetection. Starting from the preprocessing of Level-2 data including speckle filtering newlineand landmasking, the experiments were done using Otsu, Hysterisis 3D, Modified newlineOtsu for dark spot detection. A modified version of Otsu was proposed were the seed newlinevalues were automatically calculated. The classification was done using ANN. The newlinework further progressed exploring the polarimertic SAR data. Various polarimetric newlinefeatures and decomposition were analysed using PolSARpro and later t