Sampling strategies using panoramic photographs

良好的森林管理需要良好的管理工具和良好的森林清查数据。森林清查的有效设计需要就抽样单位的数目和空间安排以及所使用的抽样单位的类型作出决定。对于加拿大的大多数森林清查，使用了比特利希的角度计数抽样（ACS）。ACS与有效的树木子采样相结合以进行详细的树干测量通常具有成本效益，并且比其他常用采样单元更不容易产生测量偏差和误差。在这个方案中，ACS被应用于360度全景照片。提出了自动检测和测量树木的图像分析算法，并提出了现场采样技术来纠正由于隐藏树木造成的计数偏差。

Good forest management requires good management tools and good forest inventory data. Efficient design of forest inventories requires decisions regarding the number and spatial arrangement of sampling units as well as the type of sampling unit utilized. For most forest inventories in Canada, Bitterlich's angle count sampling (ACS) is used. ACS coupled with efficient subsampling of trees for detailed stem measurements is generally cost-effective and less prone to measurement bias and error than other commonly employed sampling units. In this proposal, ACS is applied to 360-degree panoramic photographs. Image analysis algorithms are proposed that will automate the detection and measurement of trees and field sampling techniques to correct for count biases due to hidden trees are proposed.