发表机构
Colorado School of Mines（科罗拉多矿业大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对无GNSS城市环境中的无人机定位，提出基于建筑足迹几何匹配的视觉定位系统，利用几何驱动描述符实现大范围鲁棒定位，在113-452平方公里搜索区域内Recall@1达100%至71.4%。
AI 中文摘要
在无GNSS城市环境中运行的无人机（UAV）需要替代方法进行位置估计。现有的基于卫星图像检索或学习描述符的方法对外观变化敏感，并且随着搜索区域的扩大而迅速退化。我们提出了一种基于视觉的定位系统，该系统将向下看的无人机摄像头观察到的建筑模式与参考建筑足迹数据库进行匹配。我们的方法在航空图像中检测建筑物，将跨帧的观测累积到统一地图中，并使用一种新颖的几何驱动描述符将局部建筑布局与参考足迹进行匹配，该描述符通过每栋建筑的形状特征增强局部三角形结构。通过编码附近建筑之间的空间关系而非视觉外观，该系统对外观变化具有鲁棒性，并在大搜索区域内保持区分性。在大型都市区四个市镇的七次飞行评估中，在约113平方公里和254平方公里的搜索区域内实现了100%的Recall@1，当扩展到约452平方公里（涵盖多达277,000栋建筑）时，Recall@1为71.4%。相比之下，基线方法迅速退化，在254平方公里和452平方公里时达到0%的Recall@1。
英文摘要
Unmanned aerial vehicles (UAVs) operating in GNSS-denied urban environments require alternative methods for position estimation. Existing approaches based on satellite image retrieval or learned descriptors are sensitive to appearance variation and degrade rapidly as the search area grows. We present a vision-based localization system that matches building patterns observed from a downward-facing UAV camera against a reference building footprint database. Our approach detects buildings in aerial imagery, accumulates observations across frames into a unified map, and matches local building arrangements against reference footprints using a novel geometry-driven descriptor that augments local triangle structure with per-building shape features. By encoding spatial relationships between nearby buildings rather than visual appearance, the system is robust to appearance variations and remains discriminative over large search areas. Evaluations on seven flights across four municipalities in a large metropolitan area demonstrate 100% Recall@1 at search areas of approximately 113 km$^2$ and 254 km$^2$, and 71.4% Recall@1 when expanded to approximately 452 km$^2$, encompassing up to 277,000 buildings. In contrast, baseline methods degrade rapidly and achieve 0% Recall@1 at 254 km$^2$ and 452 km$^2$.
CommentsAccepted IROS 2026
