School of Land and Resources Engineering, Kunming University of Science and Technology, Kunming 650093, China.
Plication Engineering Research Center of Spatial Information Surveying and Mapping Technology in Plateauand Mountainous Areas Set by Universities in Yunnan Province, Kunming 650093, China.
Sensors (Basel). 2022 Aug 24;22(17):6388. doi: 10.3390/s22176388.
This work aimed to detect the vegetation coverage and evaluate the benefits of afforestation and ecological protection. Unmanned aerial vehicle (UAV) aerial survey was adopted to obtain the images of tailings area at Ma'anshan near the Dianchi Lake estuary, so as to construct a high-resolution Digital Orthophoto Map (DOM) and high-density Dense Image Matching (DIM) point cloud. Firstly, the optimal scale was selected for segmentation by considering the terrain. Secondly, the visible-band difference vegetation index (VDVI) of the classified vegetation information of the tail mining area was determined from the index gray histogram, ground class error analysis, and the qualitative and quantitative analysis of the bimodal index. Then, the vegetation information was extracted by combining the random forest (RF) classification algorithm. Finally, the extracted two-dimensional (2D) vegetation information was mapped to the three-dimensional (3D) point cloud, and the redundant data was eliminated. Fractional vegetation cover (FVC) was counted in the way of surface to point and human-machine combination. The experimental results showed that the vegetation information extracted from the 2D image was mapped to the 3D point cloud in the form of surface to point, and the redundant bare ground information was eliminated. The statistical FVC was 36.06%. The field survey suggested that the vegetation information in the turf dam area adjacent to the open phosphate deposit accumulation area research area was sparse. Relevant measures should be taken in the subsequent mining to avoid ecological damage caused by expanded phosphate mining. In general, applying UAV measurement technology and related 2D and 3D products to detect the vegetation coverage in an open phosphate mine area was of practical significance and unique technical advantages.
本研究旨在检测植被覆盖情况，并评估造林和生态保护的效益。采用无人机 (UAV) 航空测量获取滇池口附近马鞍山尾矿区的图像，构建高分辨率数字正射影像图 (DOM) 和高密度密集影像匹配 (DIM) 点云。首先，考虑地形选择最优的分割尺度。其次，从指数灰度直方图、地面类误差分析和双峰指数的定性和定量分析中，确定尾矿开采区分类植被信息的可见波段差异植被指数 (VDVI)。然后，结合随机森林 (RF) 分类算法提取植被信息。最后，将提取的二维 (2D) 植被信息映射到三维 (3D) 点云中，并消除冗余数据。采用人机结合的方式，对二维植被信息进行点到面的统计，计算植被的分数植被覆盖 (FVC)。实验结果表明，从二维图像中提取的植被信息以点到面的形式映射到三维点云中，消除了冗余的裸地信息。统计的 FVC 为 36.06%。实地调查表明，研究区附近草坪堤坝区的植被稀疏。在后续开采中应采取相关措施，避免因扩大磷矿开采而造成生态破坏。总之，将无人机测量技术及相关二维和三维产品应用于露天磷矿区的植被覆盖检测具有重要的实际意义和独特的技术优势。
