一种基于机器视觉的无人机倾斜摄影三维重建方法及系统

By employing UAV oblique photogrammetry 3D reconstruction, combined with oblique camera and LiDAR data, and using a hierarchical window attention backbone network and cross-modal cross-attention mechanism, the problem of reconstructing weak texture regions and highly reflective surfaces was solved, and high-precision 3D reconstruction model generation was achieved.

CN122244333BActive Publication Date: 2026-07-17HUNAN CHANGSHUN ENG CONSTRUCT JIANLI CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HUNAN CHANGSHUN ENG CONSTRUCT JIANLI CO LTD
Filing Date
2026-05-18
Publication Date
2026-07-17

Smart Images

  • Figure CN122244333B_ABST
    Figure CN122244333B_ABST
Patent Text Reader

Abstract

本发明公开了一种基于机器视觉的无人机倾斜摄影三维重建方法及系统，涉及摄影三维重建技术领域。包括：生成同步影像数据集与同步点云数据集；生成增强影像数据集；采用包含SwinTransformer‑V2与DCNv4可变形卷积耦合结构的分层式窗口注意力骨干网络，构建高分辨率纹理特征、中分辨率结构特征与低分辨率语义特征组成的多尺度特征金字塔；形成几何锚点引导标记；在改进DUSt3R三维重建网络中，输出融合特征集；生成初始视觉稠密点云；生成融合增强稠密点云；输出满足工程测绘精度要求的三维重建成果。本发明在复杂工地场景中可显著降低点云空洞率并提升连续表面覆盖能力。
Need to check novelty before this filing date? Find Prior Art