摘要
随着数字林业和精准林业建设的推进,利用三维可视化技术辅助森林管理的需求越发突出,然而3D森林景观建模与建筑物建模相比具有冠层起伏频繁、形状不规则、光线条件复杂、纹理重复度高、像对间存在非刚体性变化等特点,直接采用普通建模方法构建的三维模型效果不理想。本研究实践了一种无人机倾斜摄影系统与背包式三维激光扫描系统相结合的森林景观三维场景建模方法,讨论了数据获取原理、点云配准方法及数据处理流程,并在潭柘寺森林景区进行实验,验证方法的有效性。结果表明,两种点云配准的整体误差小于2.37 mm,能够满足森林景观三维场景建模的精度要求。该方法不仅能省去倾斜摄影的外业控制点采集过程,还实现了两种技术的优势互补,降低了森林景观三维建模的成本,提高了建模效率,对森林精准计测及精细化管理具有一定实践意义。
With the construction of digital forestry and accurate forestry,it is more urgent to use 3D visualization technology to assist forest management,but forest landscape 3D modeling is featured with frequency canopy fluctuation,out of shape,complicated light conditions,high texture repetition,and non-rigid change for image pairs by comparing with the architectural modeling.As a result,the model effect constructed by using a common modeling method is not ideal.In this paper,the author proposed a forest landscape 3D modeling method by combining UAV oblique photography system and knapsack 3D laser scanner system,discussed data acquisition principles,point cloud registering method,and data processing flow,and did an experiment in the forest scenic area of Tanzhe temple to verify the method effectiveness.The experimental results indicated that the overall registering error of two point clouds was less than 2.37 mm,which could meet the precision requirements in forest landscape 3D scene modeling.Such a method not only can save the acquisition process of filed operation control points for oblique photography,but also can complement two technologies’advantages,reduce costs of forest landscape 3D modeling,and improve efficiency,showing some practice significance on accurate measurement and fine management in forests.
作者
曹明兰
李亚东
冯海英
蔡彩
CAO Minglan;LI Yadong;FENG Haiying;CAI Cai(Beijing Polytechnic College,Beijing 100042,China;Beijing Key Laboratory of Precision Forestry of Beijing Forestry University,Beijing 100083,China;Beijing Institute of Surveying and Mapping,Beijing 100038,China)
出处
《中南林业科技大学学报》
CAS
CSCD
北大核心
2019年第12期10-15,33,共7页
Journal of Central South University of Forestry & Technology
基金
北京市教育委员会科研计划一般项目(KM201710853005)