南京林业大学学报(自然科学版) ›› 2018, Vol. 42 ›› Issue (01): 127-132.doi: 10.3969/j.issn.1000-2006.201701008

• 研究论文 • 上一篇    下一篇

基于移动二维激光扫描的单木三维绿量测定

李秋洁,郑加强,周宏平,束义平,徐 波   

  1. 南京林业大学机械电子工程学院,江苏 南京 210037
  • 出版日期:2018-03-30 发布日期:2018-03-30
  • 基金资助:
    基金项目:国家林业局“948”项目(2015-4-56); 江苏省产学研联合创新资金-前瞻性联合研究项目(BY2014006-02); 江苏省基础研究计划青年基金项目(BK20170930); 国家自然科学基金面上项目(61473156) 第一作者:李秋洁(liqiujie_1@163.com),副教授,博士。

Three-dimensional green biomass measurement for individual tree using mobile two-dimensional laser scanning

LI Qiujie, ZHENG Jiaqiang, ZHOU Hongping, SHU Yiping, XU Bo   

  1. College of Mechanical and Electronic Engineering, Nanjing Forestry University, Nanjing 210037, China
  • Online:2018-03-30 Published:2018-03-30

摘要: 【目的】单木三维绿量的有效测定为森林资源调查、森林生态环境效益评价、林业精准管理提供有用信息,对现代林业发展具有重要意义。为提高单木三维绿量测量效率与测量精度,提出一种基于移动二维激光扫描的低成本单木三维绿量测定方法。【方法】采用二维激光扫描仪对单木进行多角度直线扫描,获取树冠三维形貌信息; 融合不同角度的采集数据,建立激光扫描线的向量参数方程; 建立树冠体素模型,将待测空间离散化为体素,体素尺寸与扫描系统分辨率一致,以未被扫描线穿透的体素集合表达树冠体积,求取单木三维绿量。【结果】实验选取6株单木,体素分辨率为0.022 m,6株单木三维绿量平均测算时间为105.94 s,平均帧处理时间为0.039 s。【结论】基于移动二维激光扫描的单木三维绿量测定,无需人工筛选点云,可自动处理数据,测量效率高,改进后的体素模型充分利用测量数据,能真实反映树冠内部体积及空隙。

Abstract: 【Objective】The effective measurement of three-dimensional green biomass for individual trees can provide useful information for forest resource survey, evaluating the environment benefit for forest ecosystem, and precision management of forests, which are great significance to the development of modern forestry. This paper proposed a low-cost measurement method for three-dimensional green biomass of individual tree using mobile two-dimensional laser scanning. 【Method】Firstly, the individual tree to be measured was scanned in straight lines from multiple angles by the mobile two-dimensional laser to capture three-dimensional morphology of the crown. Thereafter, the coordinate system and vector parameterization equation of moving trajectory were created to fuse the data from different angles. An automatic trunk recognition method based on line segment detection was proposed to derive two important parameters of vector parameterization equation of moving trajectory. Subsequently, vector parameterization equations of all the scanning lines were figured out. The space to be measured was discretized as a collection of voxels, whose size was consistent with scanning system resolution. The crown volume was regarded as the non-penetrating space of the laser scanning line. Finally, the voxels, which were penetrated by the scanning lines, were removed,and the remainder was the crown volume. 【Result】Six trees were selected and the results of the experiments showed that average time consumed for measuring individual trees was 105.94 s, and the average frame processing time was 0.039 s, with 0.022 m voxel resolution. The computational complexity of the proposed method increased linearly with the number of scan lines. The proposed method automatically processed the data without filtering the point cloud manually. 【Conclusion】The improved voxel model made full use of the measured data, which can truly reflect the internal volume and void of the canopy.

中图分类号: