Abstract

Laser scanning technique is an important area of the optical and laser technology, which makes the access of 3D individual tree information becomes available. In order to deal with the biomass and structure estimation of the urban forest, many algorithms have been developed for 3D point clouds to extract individual tree information, including tree counts, tree locations, branching structure and tree heights. However, due to the fact that the urban forest environment is complex, i.e. tree stems are non-vertical, tree crowns are overlapped and tree branches are in different structures, the existing methods are far from being desired in terms of the identification accuracy and robustness. The goal of this paper is to present a novel tree mapping algorithm that provides both tree stems and main branches, i.e. main nonphotosynthetic components, for inadequately identifying branches information. This work is based on an iterative clustering method to group point clouds and uses a growing strategy to merge tree branches and trunks with the help of the Euclidean distance and elevation difference information. The experiment dataset contains different types of roadside trees collected by the mobile laser scanning technique. Results show that the correctness and completeness of the proposed method are 95.2 and 88.5%, respectively, in the clustering of trees’ main nonphotosynthetic components, which presents a promising approach for street trees identification.

中文翻译：

---

从移动激光扫描数据中识别街树的主要非光合作用成分

摘要

激光扫描技术是光学和激光技术的重要领域，这使得3D单个树信息的访问变得可用。为了处理城市森林的生物量和结构估计，已经开发了许多算法来处理3D点云，以提取单个树木信息，包括树木数量，树木位置，分支结构和树木高度。然而，由于城市森林环境复杂，即树茎不垂直，树冠重叠且树枝处于不同结构，这一事实在识别精度和鲁棒性方面远非期望。本文的目的是提出一种新颖的树图映射算法，该算法既提供树的茎又提供主要的分支，即主要的非光合作用成分，用于识别分支机构信息的信息不足。这项工作基于迭代聚类方法对点云进行分组，并在欧几里得距离和高程差信息的帮助下使用了一种增长策略来合并树枝和树干。实验数据集包含通过移动激光扫描技术收集的不同类型的路边树木。结果表明，在树木主要非光合成分的聚类中，该方法的正确性和完整性分别为95.2和88.5％，为行道树的鉴定提供了一种有希望的方法。这项工作基于迭代聚类方法对点云进行分组，并在欧几里得距离和高程差信息的帮助下使用一种增长策略来合并树枝和树干。实验数据集包含通过移动激光扫描技术收集的不同类型的路边树木。结果表明，在树木主要非光合成分的聚类中，该方法的正确性和完整性分别为95.2和88.5％，为行道树的鉴定提供了一种有希望的方法。这项工作基于迭代聚类方法对点云进行分组，并在欧几里得距离和高程差信息的帮助下使用了一种增长策略来合并树枝和树干。实验数据集包含通过移动激光扫描技术收集的不同类型的路边树木。结果表明，在树木主要非光合成分的聚类中，该方法的正确性和完整性分别为95.2和88.5％，为行道树的鉴定提供了一种有希望的方法。