Design and Evaluation of Multispectral LiDAR for the Recovery of Arboreal Parameters
Design and Evaluation of Multispectral LiDAR for the Recovery of Arboreal Parameters
复制标题
DOI：
10.1109/tgrs.2013.2285942
复制
发表时间：
2014-08-01
影响因子：
8.2
通讯作者：
Buller, Gerald S.
中科院分区：
文献类型：
--
作者：
Wallace, Andrew M.;McCarthy, Aongus;Buller, Gerald S.
多光谱光探测和测距（LiDAR）具有从树木样本中恢复结构和生理数据的潜力，并且当部署在空中或空间平台上时，还可以从森林树冠中恢复数据。在本文中，我们描述了一个原型多光谱激光雷达系统的设计和评估，并演示了通过倾斜针叶树，使其顶端指向观测轴，在“从上面看”的树木样本上进行一系列实验，以测量叶和树皮的面积和丰度剖面。由于所有结构和生理参数的完全恢复都受到四个波长的限制，我们使用实验室测量的叶子和树皮光谱来约束参数反演，采用扩展可逆跳跃马尔可夫链蒙特卡罗算法。然而，我们也在一个单独的实验中展示了多光谱激光雷达如何直接恢复归一化植被指数（NDVI）的剖面，并通过实验室光谱测量进行了验证。我们的工作显示了多光谱激光雷达在恢复结构和生理数据方面的潜力，同时也强调了通过时间相关单光子计数可以实现的精细空间分辨率。
Multispectral light detection and ranging (LiDAR) has the potential to recover structural and physiological data from arboreal samples and, by extension, from forest canopies when deployed on aerial or space platforms. In this paper, we describe the design and evaluation of a prototype multispectral LiDAR system and demonstrate the measurement of leaf and bark area and abundance profiles using a series of experiments on tree samples "viewed from above" by tilting living conifers such that the apex is directed on the viewing axis. As the complete recovery of all structural and physiological parameters is ill posed with a restricted set of four wavelengths, we used leaf and bark spectra measured in the laboratory to constrain parameter inversion by an extended reversible jump Markov chain Monte Carlo algorithm. However, we also show in a separate experiment how the multispectral LiDAR can recover directly a profile of Normalized Difference Vegetation Index (NDVI), which is verified against the laboratory spectral measurements. Our work shows the potential of multispectral LiDAR to recover both structural and physiological data and also highlights the fine spatial resolution that can be achieved with time-correlated single-photon counting.