Datasets from: "Deep learning for detecting and characterizing oil and gas well pads in satellite imagery"

This repository contains the following: - training folder: Datasets used to train well pad and storage tank models. Note that annotations_image for each example are with respect to a image downloaded from the Google Earth satellite basemap at zoom level 1600 and EPSG:3857, with sizes 640x640 px and 512x512 px for well pads and storage tanks respectively (see also the image_extent column). annotations_latlon also provides the annotations in coordinate space. We also note that we are unable to redistribute the satellite imagery used to train the models in this study due to data licensing. Samples of satellite images may be made available upon request to the corresponding author. - deployment folder: Deployment detections for well pads and storage tanks across the entire Permian and Denver basins. Datasets contain confidence scores (bbox_score) and coordinate locations (geometry) for each detection, as well as a well pad identifier (wp_id) that indicates which well pad a storage tank belongs to.