文献类型：期刊文献

中文题名：树木三维建模与可视化模拟技术进展与应用

英文题名：Progress and Applications of 3D Modeling and Visualization Simulation Technology for Trees

作者：崔泽宇[1,2] 张怀清[1,2] 刘洋[1,2] 张京[1,2] 杨廷栋[1,2] 傅汝饶[1,2,3]

第一作者：崔泽宇

机构：[1]中国林业科学研究院资源信息研究所,北京100091;[2]国家林业和草原科学数据中心,北京100091;[3]中南林业科技大学,长沙410004

年份：2025

卷号：61

期号：6

起止页码：1-12

中文期刊名：林业科学

外文期刊名：Scientia Silvae Sinicae

收录：;北大核心:【北大核心2023】；

基金：国家重点研发计划课题(2023YFF1303604);国家自然科学基金项目(32271877,32071681)。

语种：中文

中文关键词：树木三维建模;生成式人工智能;多源数据融合;数字孪生;元宇宙

外文关键词：tree 3D modeling;generative artificial intelligence;multi-source data fusion;digital twin;metaverse

分类号：S757

摘要：树木三维建模与可视化模拟技术是研究森林结构与生长动态规律的重要手段,也是开展森林精准经营管理的关键技术和基础。本文从基于一维文本规则、二维图像、三维激光雷达点云以及多源数据融合等不同维度,系统分析了不同类型树木三维建模技术,并梳理了树木形态结构、生长、多态性可视化模拟等方面的进展。结合树木三维建模技术的发展趋势,重点讨论了生成式人工智能GAI、多源数据融合、大模型等技术在树木三维建模中的融合应用前景,论述了树木三维建模个性化、多样性、大规模、智能化发展方向,并梳理了在GAI发展背景下树木三维建模的研究思路,提出以多源数据融合为基础、GAI为驱动的智能化三维建模框架。分析了数字孪生、元宇宙等技术体系对树木三维模型的应用推广,剖析了其在树木育种培育、古树名木保护、景观设计、森林经营、生态监测灾害预警以及生态保护与修复等树木全生命周期应用场景中的特征与需求。通过树木三维模型模拟,能够有效突破时空限制,辅助智能决策,为森林精准、高效经营管理以及生态系统服务修复等提供关键技术支撑,并为树木培育、监测、保护等提供全新的研究思路与智能化的手段,助力林业全业务场景的跨模态数据协同处理与机制探索。
The 3D modeling and visualization simulation technology for trees is essential tools for investigating forest structure and growth dynamics,and a key technology and foundation for carrying out precision forest management.This paper systematically reviews various tree 3D modeling techniques from different dimensions,including rule-based one-dimensional text descriptions,two-dimensional imagery,three-dimensional LiDAR point clouds,and multi-source data fusion approaches,and also summarizes the progress in the visualization and simulation of tree morphology,growth,and polymorphism.With an emphasis on current trends in 3D tree modeling,this paper focuses on the integration application prospects of generative artificial intelligence(GAI),multi-source data fusion,and large model technologies.It discusses future directions in personalized,diverse,large-scale,and intelligent 3D modeling of trees.In the context of rapid GAI development,a novel framework for intelligent 3D tree modeling is proposed,grounded in multi-source data fusion and powered by GAI.The paper further examines the enabling role of emerging technologies such as digital twins and the metaverse in expanding the applications of tree 3D models.It analyzes their functional requirements and potential across the entire lifecycle of trees,including breeding and cultivation,heritage tree conservation,landscape design,forest operations,ecological monitoring and disaster early warning,as well as ecosystem protection and restoration.Through 3D model simulations of trees,spatial and temporal constraints can be effectively overcome to support intelligent decision-making.These technologies offer critical support for precise and efficient forest management and ecosystem service restoration,provide novel research paradigms and intelligent approaches for tree breeding,monitoring,and conservation,and enable cross-modal data integration and mechanism discovery across the entire spectrum of forestry applications.