Estimating Tree Height and Diameter at Breast Height (DBH) from Digital Surface Models and Orthophotos Obtained with an Unmanned Aerial System for a Japanese Cypress (Chamaecyparis obtusa) Forest
Estimating Tree Height and Diameter at Breast Height (DBH) from Digital Surface Models and Orthophotos Obtained with an Unmanned Aerial System for a Japanese Cypress (Chamaecyparis obtusa) Forest
复制标题
DOI：
10.3390/rs10010013
复制
发表时间：
2018-01-01
期刊：
影响因子：
5
通讯作者：
Kosugi, Yoshiko
中科院分区：
文献类型：
--
作者：
Iizuka, Kotaro;Yonehara, Taichiro;Kosugi, Yoshiko
准确测量生物物理参数的方法是对各种森林应用进行定量评价的关键组成部分。这些参数的常规原位测量花费时间和费用，在具有异质微观形貌的位置处遇到困难。为了在这种情况下获得精确的生物物理数据，我们在日本山区的柏树林中部署了无人驾驶航空系统（UAS）多旋翼无人机。采用运动恢复结构（SfM）方法，从航空照片中构建森林（树）结构的三维（3D）模型。根据3D模型估计树高，并与原位地面数据进行比较。我们还分析了生物物理参数，胸径（DBH），与冠层宽度和面积从正射影像测量的个别树木之间的关系。尽管在高度密集的森林地区的地面曝光的限制，树高估计的精度均方根误差= 1.712米的观测树高范围从16到24米。胸径在11 ~ 58 cm范围内与冠幅和冠面积呈极显著正相关（R-2 = 0.7786和R-2 = 0.7923）。估算森林参数的结果表明，基于无人机的遥感方法可以用来准确地分析森林结构的空间范围。
Methods for accurately measuring biophysical parameters are a key component for quantitative evaluation regarding to various forest applications. Conventional in situ measurements of these parameters take time and expense, encountering difficultness at locations with heterogeneous microtopography. To obtain precise biophysical data in such situations, we deployed an unmanned aerial system (UAS) multirotor drone in a cypress forest in a mountainous area of Japan. The structure from motion (SfM) method was used to construct a three-dimensional (3D) model of the forest (tree) structures from aerial photos. Tree height was estimated from the 3D model and compared to in situ ground data. We also analyzed the relationships between a biophysical parameter, diameter at breast height (DBH), of individual trees with canopy width and area measured from orthorectified images. Despite the constraints of ground exposure in a highly dense forest area, tree height was estimated at an accuracy of root mean square error = 1.712 m for observed tree heights ranging from 16 to 24 m. DBH was highly correlated with canopy width (R-2 = 0.7786) and canopy area (R-2 = 0.7923), where DBH ranged from 11 to 58 cm. The results of estimating forest parameters indicate that drone-based remote-sensing methods can be utilized to accurately analyze the spatial extent of forest structures.