Particulate matter and image data for estimation of atmospheric visibility

The dataset contains image captures and particulate matter readings from tractor tracks in agricultural fields, and is intended to develop estimations of atmospheric visibility. The data is divided as follows: A YOLO-formatted object detection task, intended to detect signpost objects in the camera’s field of view, which is used to estimate visibility. The data is pre-split in train and validation sets, and contains a timestamp in the file name. Sensor measurements for each respective image, indexed by timestamp. Note that the sensor measurements are present for all images in both sets, and that some sensor measurements do not have a corresponding image capture.