发表机构
Wuhan University; Hangzhou Dianzi University（武汉大学; 杭州电子科技大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
CoMVS-GS结合多视图立体与3D高斯溅射，通过新的初始化方式和互监督机制优化表面重建，在室外场景提升几何精度与网格紧凑性，同时保持高渲染质量。
AI 中文摘要
3D高斯溅射（3D Gaussian Splatting）可实现高效的新视图合成，但在观测不足和遮挡区域，准确的网格重建仍存在困难，此处高斯基元可能生长为不稳定或几何不一致的结构。我们提出CoMVS-GS，这是一种结合多视图立体（Multi-View Stereo）与高斯溅射的通用表面重建框架。CoMVS-GS从具有预展平尺度和法向对齐方向的密集多视图立体点初始化高斯基元，提供比稀疏运动恢复结构（structure-from-motion）初始化更强的几何先验，减少早期优化中的歧义。它还引入PatchMatch-3DGS互监督机制：高斯渲染的深度和法向初始化PatchMatch优化，优化后的PatchMatch深度监督高斯优化，以改善弱约束几何。对于表面提取，CoMVS-GS用Delaunay图割网格划分流程替代截断符号距离体素融合，降低对体素分辨率的敏感性，同时保留与可见性一致的表面证据。在DTU、GauU-Scene V2和MatrixCity数据集上的实验表明，CoMVS-GS在物体级重建上仍具竞争力，在保持高渲染质量的同时，提升了室外场景的几何精度和网格紧凑性。
英文摘要
3D Gaussian Splatting enables efficient novel view synthesis, but accurate mesh reconstruction remains difficult in weakly observed and occluded regions, where Gaussian primitives may grow into unstable or geometrically inconsistent structures. We propose CoMVS-GS, a general surface reconstruction framework that combines Multi-View Stereo with Gaussian splatting. CoMVS-GS initializes Gaussian primitives from dense multi-view stereo points with pre-flattened scales and normal-aligned orientations, providing stronger geometric priors than sparse structure-from-motion initialization and reducing ambiguity during early optimization. It further introduces PatchMatch-3DGS Mutual Supervision, where Gaussian-rendered depths and normals initialize PatchMatch refinement, and refined PatchMatch depths supervise Gaussian optimization to improve weakly constrained geometry. For surface extraction, CoMVS-GS replaces truncated signed distance field voxel fusion with a Delaunay graph-cut meshing pipeline, reducing sensitivity to voxel resolution while preserving visibility-consistent surface evidence. Experiments on DTU, GauU-Scene V2, and MatrixCity show that CoMVS-GS remains competitive on object-level reconstruction and improves geometric accuracy and mesh compactness in outdoor scenes while maintaining high rendering quality.
