Grazing Yak Drone Images, data underlying the research of livestock distribution

Dataset OverviewThis dataset focuses on grazing yaks in the eastern Qinghai - Tibet Plateau, aiming to support UAV - based monitoring of livestock distribution. It provides valuable image resources for detecting and analyzing yaks in high - altitude cold alpine meadow grazing systems.Data Collection<strong>Location</strong>: The data was collected in the core distribution area of yaks in the eastern Qinghai - Tibet Plateau. This region is a representative cold alpine meadow grazing system with an average elevation of 3,840 meters (ranging from 3,705 to 4,080 meters) and a vertical relief of 375 meters.<strong>Time</strong>: Data collection took place during the winter of 2024.<strong>Equipment and Parameters</strong>: A DJI Mavic 3M drone was used to capture 722 aerial images. The flight parameters include: flight altitude of 500 m, cruise speed of 15 m/s, forward overlap of 80%, side overlap of 70%, image dimensions of 5280×3956 pixels, and ground resolution of 23 cm. The orthomosaic stitching of the study area was accomplished via DJI Terra software, resulting in an orthomosaic that covers an area of 9.21 km².Dataset Content<strong>Images</strong>: 6,931 images of 512×512 pixels were selected, covering various areas such as artificial pastures, natural grasslands, residential areas, valleys, and roads.<strong>Labels</strong>: Yak locations in the images were marked using Labelimg.Data Storage Directory Structure<strong>images Folder</strong>: Contains the images of the dataset.<strong>labels Folder</strong>: Holds the labels corresponding to the images.<strong>area Folder</strong>: Includes the orthomosaic of the study area.