Oil Spill Detection using Synthetic Aperture Radar

Oil spills are a global issue because they generate environmental difficulties as well as financial losses. Automatic detection is critical in making quick decisions to minimize harm. Deep learning is a strong tool in this setting, with successful applications in many fields. On a wide scale, we provide a deep learning approach for classifying oil spills in synthetic aperture radar (SAR) images as well as for detecting them. Using a properly constructed neural network model for photo segmentation trained on a large dataset, we obtain state-of-the-art performance in oil spill detection, yielding results comparable to those provided by human operators. We also provide a unique classification problem in the context of oil spill detection in SAR. Using deep learning-based CNN with LBP and GLCM model, this study suggests a novel method for detecting oil spills. Here collected SAR image dataset that is preprocessed the input images like resize, Contrast Enhancement, filtering and image conversion. After apply image segmentation, the image is segmented by means of the thresholding and Morphological algorithm. The proposed CNN with LBP + GLCM mode gets 97.56% accuracy, 97.11% sensitivity and 99.2% specificity of this model that used SAR image dataset.