UAV Canadian (Quebec) Plantations dataset

该数据集包含了来自种植园区的RGB正射影像、摄影测量数字表面模型（DSMs）以及树木冠层划分和树种标签。数据集被划分为训练集（15,742个图块）、验证集（6,691个图块）和测试集（3,995个图块）。其中包含了基于各站点至少20棵树木的最小计数，分为9个树种类别。整个数据集共有26,428个图块，图像分辨率为5毫米/像素，任务是对树木冠层实例进行分割。

This dataset contains RGB orthophotos, photogrammetric digital surface models (DSMs), as well as tree canopy segmentation annotations and tree species labels collected from plantation areas. It is divided into a training set (15,742 image patches), a validation set (6,691 image patches), and a test set (3,995 image patches). The dataset includes 9 tree species categories, with each category meeting the minimum criterion of at least 20 trees per sampling site. The entire dataset consists of 26,428 image patches in total, with an image resolution of 5 millimeters per pixel, and the targeted task is tree canopy instance segmentation.