应用地基激光雷达数据提取毛竹单木参数

以毛竹（Phylotachys eduis)为研究对象，利用地基漱光雷达（TIS)多站扫描状取样地毛付林完整点云数据，根据毛竹笔直的干形以及密集的林分环境，提出一种新的茎干识刷方法。该方法首先裁取只舍竹秆的水平条带，然后对水平条带进行体元化，利用垂直连续性来识别毛什茎干，并比较分析圆合和不同高度圆柱合对毛竹胸径提取的差异。结果表明：在密集的林分环境下，毛付茎干检测率达91.93%;固拟合估测牌径的R为0.86,相对误差为0.8cm;圆柱合估测胸径最高的R为0.91,相对误差为0.4cm.因此，新的茎干识别方法提高了单株毛竹识刷率及胸径估测精度，为竹林调查工作提供技术支持。

Taking moso bamboo (Phyllostachys edulis, corrected from the original Phylotachys eduis) as the research subject, complete point cloud data of moso bamboo forests in multi-station scanning plots was collected using terrestrial laser radar (TIS). In view of the straight stem shape and dense stand environment of moso bamboo, a novel stem detection method was proposed. The method first extracts horizontal strips that only contain bamboo stems, then voxelizes these horizontal strips, and identifies moso bamboo stems by utilizing vertical continuity. We compared and analyzed the differences in diameter at breast height (DBH) extraction of moso bamboo between circle fitting and cylinder fitting at different heights. The results demonstrated that under dense forest stand conditions, the detection rate of moso bamboo stems reached 91.93%; the R value of circle fitting for DBH estimation was 0.86, with a relative error of 0.8 cm; the highest R value of cylinder fitting for DBH estimation was 0.91, with a relative error of 0.4 cm. Consequently, the novel stem detection method enhances the detection rate of individual moso bamboo trees and the accuracy of DBH estimation, providing technical support for bamboo forest inventory operations.

搜集汇总

背景与挑战

背景概述

该数据集聚焦于利用地基激光雷达技术提取毛竹单木参数，通过多站扫描获取点云数据，并开发了一种新的茎干识别方法。该方法在密集林分环境下实现了91.93%的茎干检测率，显著提升了胸径估测精度，最高R值达0.91，为竹林资源调查提供了有效技术支持。