2026-10-11T23:03:29Z https://www.scidb.cn/oai

10.57760/sciencedb.39831 2026-06-16T16:09:01Z

2026-06-16 SA-Tree dataset doi:10.57760/sciencedb.39831 en This dataset is a visualization thematic dataset for remote sensing monitoring of individual trees in typical ultra arid areas of Saudi Arabia. The data was collected from China's GF-7 satellite and covers the period from 2023 to 2024. The original panchromatic images have a spatial resolution of over 0.7 meters and cover scenes such as deserts, economic forests, and urban green spaces within Saudi Arabia. The dataset is created using annotation software and Python scripts. Firstly, the polygon vector annotation of over 19000 individual tree boundaries is completed through manual visual interpretation by professional personnel. The structure is complete without any missing parts, and the error mainly comes from shadow interference and sub meter pixel judgment. After multiple rounds of cross validation, the geometric contour accuracy is extremely high.  Individual tree detection; GF-7 satellite; Saudi Arabia; Arid region remote sensing; Object detection Zhang Tao EMBARGO https://creativecommons.org/licenses/by/4.0/ dataset Science Data Bank