This image dataset: "Sentinel-1 SAR Oil spill image dataset for train, validate, and test deep learning models. Part II", is the second part of the image dataset for train and validate deep learning models. This part contains only the training and validation images for No Oil and Lookalike. the dataset comprises Sentinel-1 SAR images in Sigma0, in decibels (db), along with their ground truth. The images are 2048x2048x2, also the ground truth is 2048x2048; all of them are in TIFF format. The files are organized in the following manner: 01_Train_Val_No_Oil_Images: There are 685 images of oil-free Sentinel-1 SAR Sigma0 images in the database. 01_Train_Val_No_Oil_mask: There are 685 images in this dataset that represent the ground truth of the Sentinel-1 SAR Sigma0 images in db of the oil-free scenes. In these images, the foreground has a value of 1, and the background has a value of 0. Because the focus of this dataset is on oil spills, the ground truth of oil-free images is assigned a value of 0. 01_Train_Val_Lookalike_images: The database contains 685 Sentinel-1 SAR Sigma0 images that correspond to oil-like surfaces, also known as look-alikes. 01_Train_Val_Lookalike_mask: It includes 685 images that correspond to the ground truth of the Sentinel-1 SAR Sigma0 images in the database of the look-alike images, where the foreground has a value of 1 and the background has a value of 0. Since the focus was on oil spills, the value of the ground truth of the look-alike images is only 0. Each corresponding ground truth has the same number as its respective image. For instance, the image of an oil spill has a corresponding number of 0001, as well as its ground truth. The complete dataset consists of three parts: Sentinel-1 SAR Oil spill image dataset for train, validate, and test deep learning models. Part I. (10.5281/zenodo.8346860) Sentinel-1 SAR Oil spill image dataset for train, validate, and test deep learning models. Part II. (10.5281/zenodo.8253899) Sentinel-1 SAR Oil spill image dataset for train, validate, and test deep learning models. Part III. (10.5281/zenodo.13761290)

Note that only the Sentinel-1 Sigma0 images in decibels (db) with two polarizations (VV, VH) and dimensions of 2048x2048x2 are georeferenced. The masks or ground truth of each of these images are not georeferenced because they were treated as matrices for the purpose of training, validation, and testing of deep learning models.