基于TLS点云骨架提取的杨树苗木干旱表型特征分析

Analysis of Drought Phenotypic Characteristics of Poplar Seedlings Based on TLS Point Cloud Skeleton Extraction

摘要:

植物干旱胁迫监测的关键在于如何精确定位和识别目标，为此，高效的植物表型提取系统成为必要配备。三维点云信息因其能提供高精度的三维描述，成为这一系统中重要的数据支撑，为植物在干旱环境中的长势监测提供了坚实的技术基础。本文采用地基激光雷达技术采集杨树苗木三维点云数据，并提出了一种结合预分割的L1中值骨架提取算法，实现精细表型提取与干旱特征分析。首先，通过高程分析、半径滤波和颜色指数滤波对原始点云进行去噪预处理；其次，利用改进的DBSCAN算法实现群体点云单木分割，并结合基于贪婪算法的八叉树进行全局搜索以优化分割精度；最终，利用KNN算法与MRF算法对单株点云进行预分割，提升点云数据的空间一致性，降低L1中值算法的计算复杂度，通过得到的骨架点云计算杨树苗木的表型特征。提出引入冠长率和高径比2个新的指标，以揭示杨树苗木在干旱胁迫下通过优化资源分配和减少水分消耗的适应机制。其中，冠长率在CK组和DT组的抗旱性评价中灰色关联度均排名第1，相关系数为-0.85，表明其对水分供应高度敏感，能够全面反映植物的资源利用效率和抗旱能力，是评估杨树苗木干旱适应性的核心指标。通过结合三维点云技术与精细表型分析，为杨树苗木早期干旱胁迫的高效精准监测提供了技术支持，对确定干旱表型指标、优化抗旱性评价体系具有意义。

Abstract:

The key to monitoring plant drought stress lies in how to accurately locate and identify targets, and for this reason, an efficient plant phenotype extraction system has become a necessity. Because of its ability to provide high-precision 3D description, 3D point cloud information has become an important data support in this system, which provides a solid technical foundation for the monitoring of plant growth in arid environments. Ground based lidar technology was used to collect the three-dimensional point cloud data of poplar seedlings, and an L1 median skeleton extraction algorithm combined with pre-segmentation was proposed to realize fine phenotype extraction and drought feature analysis. Firstly, the original point cloud was denoised and preprocessed by elevation analysis, radius filtering and color index filtering. Secondly, the improved DBSCAN algorithm was used to realize the single-tree segmentation of the group point cloud, and the octree based on the greedy algorithm was combined with the global search to optimize the segmentation accuracy. Finally, the KNN algorithm and MRF algorithm were used to pre-segment the point cloud of a single plant, so as to improve the spatial consistency of the point cloud data, reduce the computational complexity of the L1 median algorithm, and calculate the phenotypic characteristics of poplar seedlings through the obtained skeleton point cloud. Two new indexes were introduced to reveal the adaptation mechanism of poplar seedlings under drought stress by optimizing resource allocation and reducing water consumption. Among them, the crown length rate ranked first in the gray correlation degree of drought resistance evaluation in the CK group and DT group, with a correlation coefficient of -0.85, indicating that it was highly sensitive to water supply and could fully reflect the resource use efficiency and drought resistance of plants, which was the core index to evaluate the drought adaptability of poplar seedlings. By combining three-dimensional point cloud technology and fine phenotypic analysis, the research can provide technical support for efficient and accurate monitoring of early drought stress in poplar seedlings, which was of significance for determining drought phenotypic indicators and optimizing the drought resistance evaluation system.