CSIRO Sentinel-1 SAR image dataset of oil- and non-oil features for machine learning ( Deep Learning )

What this collection is:\nA curated, binary-classified image dataset of grayscale (1 band) 400 x 400-pixel size, or image chips, in a JPEG format extracted from processed Sentinel-1 Synthetic Aperture Radar (SAR) satellite scenes acquired over various regions of the world, and featuring clear open ocean chips, look-alikes (wind or biogenic features) and oil slick chips.\n\nThis binary dataset contains chips labelled as:\n- "0" for chips not containing any oil features (look-alikes or clean seas) \n- "1" for those containing oil features. \n\nThis binary dataset is imbalanced, and biased towards "0" labelled chips (i.e., no oil features), which correspond to 66% of the dataset.\nChips containing oil features, labelled "1", correspond to 34% of the dataset.\n\nWhy:\nThis dataset can be used for training, validation and/or testing of machine learning, including deep learning, algorithms for the detection of oil features in SAR imagery. Directly applicable for algorithm development for the European Space Agency Sentinel-1 SAR mission (https://sentinel.esa.int/web/sentinel/missions/sentinel-1 ), it may be suitable for the development of detection algorithms for other SAR satellite sensors.\n\nOverview of this dataset:\nTotal number of chips (both classes) is N=5,630\nClass \t 0\t 1\nTotal\t\t3,725\t1,905\n\nFurther information and description is found in the ReadMe file provided (ReadMe_Sentinel1_SAR_OilNoOil_20221215.txt)