论文概要
研究领域: CV 作者: Weijian Chen, Weibo Yao, Yuhang Zhang, Xiaolin Tang, Guo Wang, Weijun Zhang, Xitong Gao, Yihao Chen, Hongde Qin, Lu Qi 发布时间: 2026-07-09 arXiv: 2607.08769
中文摘要
将3D高斯泼溅（3DGS）扩展到大型户外场景在数据采集和计算方面成本高昂。采用等距圆柱投影（ERP）的全景图像可以通过360°全视野减少采集工作量，但由此产生的无处不在的可见性使得依赖局部相机视锥的现有分区策略失效，导致块级优化退化为全局训练。因此，我们提出PanoLOG，一种配备几何与梯度分区策略（G²PS）的两阶段粗到精框架，专为大规模全景3DGS重建而设计。在全局粗阶段，PanoLOG利用天空球建模和全景单目深度监督获得可靠的几何信息；在精修阶段，G²PS通过视差驱动的不确定性构建自适应边界体，并通过基于梯度的重要性评分分配相机。此外，我们构建了Pano360，这是首个面向户外场景重建的大规模全景数据集基准。大量实验表明，G²PS在保持可扩展的块并行训练的同时，实现了最先进的渲染质量。模型、训练代码和数据集均已公开。
原文摘要
Scaling 3D Gaussian Splatting (3DGS) to large outdoor scenes is costly in both data acquisition and computation. Adopting panoramic images with equirectangular projection (ERP) can reduce capture effort via their full 
field of view, yet the resulting omnipresent visibility invalidates existing partitioning strategies that rely on local camera frustums, causing block-wise optimization to degenerate into global training. Thus, we propose PanoLOG, a two-stage coarse-to-fine framework equipped with a Geometry and Gradient-based Partitioning Strategy tailored for large-scale panoramic 3DGS reconstruction. In the global coarse stage, PanoLOG leverages sky-sphere modeling and panoramic monocular depth supervision for reliable geometry, while in the refinement stage, G
PS builds a…
— 自动采集于 2026-07-12
#论文 #arXiv #CV #小凯