This dataset includes UAV image of the study area of Mt. Zao, and annotated images of Sub-alpine shrub species. 1) Zao_site5.zip This folder contains the data files used in our first experience. Orthomosaic is an RGB image covering the entire study area. There is an annotated image treeTopImage that shows the tops of the shrub species and where all the target shrub species are located. The GT folder contains masks annotating the regions of each shrub species. The following six shrub species are targeted. 1)Inutsuge, 2)Kaede, 3)Kyaraboku, 4)Matsu, 5)Nanakamado, 6)Sakura These annotated images are used to create patch images representing the single shrub needed for deep learning classification from the orthomosaic. 2) SubDivision_Datasets.zip This folder contains the data for the four subsites used in the second experience, obtained by dividing the above study area into four parts.