Split olive image dataset v2

The dataset contains 86,790 YOLO-format annotations of olive fruits on 315 original UAV images collected from a single olive grove in Tinj, Croatia from May to November 2025. The close range images acquired using a DJI Mavic 3M UAV represent predominantly three olive cultivars: Olea europaea L. cv. Oblica Olea europaea L. cv. Leccino Olea europaea L. cv. Istarska bjelica The high-resolution original images annotated using standard YOLO object-detection format (class_id x_center y_center width height), labeling olive fruits in a single class. The zip archive contains 315 .jpg image and .txt label files separated into images/ and labels/ directories. File naming convention The filename encodes information about image acquisition, e.g.:250722_0857_S_IB_101_002 Component Example Description Date and Time 250722_0857 Acquisition date and time in YYMMDD_HHMM format Weather S Weather during the acquisition; S for sunny, O for cloudy Cultivar IB Cultivar of the olive tree in the image; IB = Istarska bjelica, LE = Leccino, OB = Oblica, etc. Tree 101 Internal individual tree ID Image 002 Internal UAV image sequence number