手持激光雷达在“森林遥感光谱样地”外业质量检查中的应用与研究
The application and research of handheld LiDAR in the field quality inspection of"forest spectral plots
[目的]本研究的核心目标在于验证手持激光雷达(LiDAR)技术在森林光谱样地的质量检查中的应用效果，并探讨其在提升森林调查精度方面的潜力。森林资源的精确调查对于理解森林现状、分布和生态变化至关重要。特别是在大兴安岭这一中国林业的重要支柱区域，采用先进技术以提高林业调查的效率和准确性显得尤为迫切。[方法]在内蒙古大兴安岭林区实施的《自然资源光谱信息研究暨其在调查监测中的创新应用项目》框架下，我们设计并布设了172块森林光谱样地，以精确获取不同林型的遥感光谱信息。项目外业质量检查中使用了SLAM100 激光雷达扫描仪，沿样地边界及内部进行扫描，收集点云数据。数据采集之后，通过重采样、去噪、地面点分类和高程归一化等一系列预处理步骤，提取了单木胸径信息。进一步结合大兴安岭地区的一元材积表，对比分析了两种不同数据源的胸径数据和蓄积量。[结果]研究结果揭示，激光雷达点云分析得到的胸径平均值与现地调查结果具有极高的一致性，偏差为 2.34%，且相关性指数达到了0.98.在树高测量方面，尽管两组数据间相关性指数为-0.04，表明缺乏相关性，但激光雷达技术在树高数据质量检查方面显示出优越性，均方根误差(RMSE)为 3.87米。蓄积量分析中，激光雷达预测的蓄积量与现地调查结果相比，偏差为 5.56%，且相关性指数为 0.97，显示了良好的相关性。[结论]本研究成功证明了手持激光雷达技术在森林光谱样地外业质量检查中的有效性，尤其在提升树高测量精度方面表现突出。研究不仅提升了外业质量管理水平，而且培养了能够适应新技术的基础人才。展望未来，我们计划普及激光雷达点云建模方法，进一步提升调查精度，并探索无人机、智能软件、虚拟现实(VR)技术以及人工智能大模型等新兴技术的应用，以推动林业资源管理向自动化、智能化发展。
[Objective]The core objective of this study is to validate the application effectiveness of handheld LiDAR technology in the quality inspection of forest spectral plots and to explore its potential to enhance the precision of forest surveys.Accurate forest resource surveys are crucial for understanding the current status,distribution,and ecological changes of forests.Particularly in the Greater Khingan Range,an important pillar of China's forestry,the adoption of advanced technology to improve the efficiency and accuracy of forestry surveys is particularly urgent.[Method]Under the framework of the"Natural Resource Spectral Information Research and Its Innovative Application in Survey Monitoring Project"implemented in the Inner Mongolia Greater Khingan Range forest area,we designed and set up 172 forest spectral plots to accurately acquire spectral information of different forest types.The SLAM100 LiDAR scanner was used in the field quality inspection of the project,scanning along the boundaries and interiors of the plots to collect point cloud data.After data collection,a series of preprocessing steps including resampling,denoising,ground point classification,and elevation normalization were conducted to extract individual tree diameter at breast height information.Further,combined with the monomeric volume table of the Greater Khingan Range area,we compared and analyzed the diameter at breast height data and stock volume from two different data sources.[Result]The study results reveal that the average diameter at breast height obtained from LiDAR point cloud analysis is highly consistent with the results of on-site surveys,with a deviation of 2.34%,and a correlation index reaching 0.98.In terms of tree height measurement,although the correlation index between the two sets of data is-0.04,indicating a lack of correlation,LiDAR technology has shown superiority in the quality check of tree height data,with a root mean square error(RMSE)of 3.87 meters.In the stock volume analysis,the deviation between the predicted stock volume by LiDAR and the on-site survey results is 5.56%,and the correlation index is 0.97,showing a good correlation.[Conclusion]This study successfully proves the effectiveness of handheld LiDAR technology in the field quality inspection of forest spectral plots,especially in improving the accuracy of tree height measurement.The research has not only improved the level of field quality management but also cultivated basic talents who can adapt to new technologies.Looking forward,we plan to popularize LiDAR point cloud modeling methods to further enhance survey accuracy and explore the application of emerging technologies such as unmanned aerial vehicles,intelligent software,virtual reality(VR)technology,and artificial intelligence large models,to promote the development of forestry resource management towards automation and intelligence.
于家强;刘微;
内蒙古风天林草调查咨询责任有限公司 内蒙古自治区 牙克石 022150;内蒙古风天林草调查咨询责任有限公司 内蒙古自治区 牙克石 022150;
Q958.15;S154.5
Handheld LiDAR Forest Survey Spectral Quality Management Correlation Analysis
1392-1401 / 10
评论