高速公路改扩建树木本底信息数智调绘方法与实践

Method and Practice of Digital-Intelligent Surveying and Mapping for Tree Background Information in Expressway Reconstruction and Expansion

摘要:

为解决高速公路改扩建时对传统树木调查效率低、风险高、精度不足等问题，探索高效精准的技术方法，并应用于工程实际。以广深高速公路改扩建项目作为研究对象，采用“无人机+机载LiDAR”调查方法，生成高精度正射影像与三维点云，结合语义分割实现单木提取、树种识别及参数解译，并建立树木本底调查数据库。结果表明：（1）树木数量调查与树种识别平均准确率达97%，2个月内高效完成东莞市、深圳市段10万余株树木调查；（2）基于调查成果优化审批流程，申报数量由10万株核减至6万株，显著提升审批效率；（3）有效降低高速公路涉路作业安全风险，相关许可已获批，为同类工程提供了可复制的实践参考。

Abstract:

In order to solve the problems of low efficiency, high risk and insufficient accuracy of traditional tree surveys in expressway reconstruction and expansion, the efficient and accurate technical methods are explored. Taking the Guangzhou - Shenzhen Expressway Reconstruction and Expansion Project as the research object, the “unmanned aerial vehicle + Airborne LiDAR” survey method is adopted to generate high-precision orthophoto images and 3D point clouds. Combined with semantic segmentation, single-tree extraction, tree species identification and parameter interpretation are realized, and a tree background database is established. The result shows that (1) the average accuracy rate of tree quantity survey and tree species identification reaches 97%, and the survey of over 100,000 trees in Dongguan City and Shenzhen City has been completed efficiently within two months. (2) Based on the survey results, the approval process is optimized, and the declared quantity is reduced from 100,000 to 60,000 trees, which significantly improves the approval efficiency. (3) The relevant permits have been approved to effectively reduce the safety risks of road-related operations on expressways, providing replicable practical references for similar projects.