测绘地理信息

随着大数据、数字孪生、物联网等新兴技术的发展，校园场景模拟技术的迭代演进呈现“虚实结合、智能化”的趋势。本文采用倾斜摄影测量数据与即时定位与地图构建（simultaneous localization and mapping,SLAM）激光点云融合建模，基于数字孪生技术搭建三维可视化平台来实现校园发展的数字化推演。为满足建模的精细化要求，通过QEMICP(quadric error metrics-iterative closest point)算法实现点云数据轻量化与分步精确配准加强点云精度，结合局部曲率分析与自适应密度阈值的去冗余算法实现点云二次轻量化，完成两种点云的融合。将融合后的模型结合UE5(unreal engine 5)引擎，利用其跨平台兼容协同和多源数据集成渲染能力，构建一个校园数字孪生管理平台。实验结果表明，本文点云数据处理流程与现有方法相比提升模型精度的同时消除多源点云冗余，加快模型在UE5中的运行效率，为校园数字孪生平台提供关键技术支撑。

With the development of emerging technologies such as big data, digital twins, and the Internet of Things, the iterative evolution of campus scene simulation technology presents a trend of "virtual real integration and intelligence".This article employs a modeling approach that integrates oblique photogrammetry data with SLAM laser point clouds, and uses digital twin technology to build a 3D visualization platform for simulating campus development. To meet the requirements of fine-grained modeling, the QEM-ICP algorithm is used to achieve lightweight point cloud data and stepby-step precise registration to enhance point cloud accuracy.The local curvature analysis and adaptive density threshold denoising algorithm are combined to achieve secondary lightweight point cloud, completing the fusion of the two types of point clouds. By combining the fused model with the UE5 engine and utilizing its cross platform compatibility, collaboration, and multi-source data integration rendering capabilities, a campus digital twin management platform is constructed.The experimental results show that compared with existing methods, the point cloud data processing flow in this paper improves model accuracy while eliminating multi-source point cloud redundancy, thereby accelerating the running efficiency of the model in UE5, and providing key technical support for the campus digital twin platform.