ULSR-GS: Ultra Large-scale Surface Reconstruction Gaussian Splatting with Multi-View Geometric Consistency

TL;DR

ULSR-GS introduces a novel framework for high-fidelity surface reconstruction in ultra-large-scale scenes, leveraging multi-view geometric consistency and optimal view matching to improve accuracy over existing Gaussian Splatting methods.

Contribution

The paper proposes a new large-scale surface reconstruction method combining point-to-photo partitioning, multi-view optimal view matching, and geometric densification, addressing limitations of prior Gaussian Splatting approaches.

Findings

Outperforms state-of-the-art GS-based methods on large-scale aerial datasets.

Significantly improves surface extraction accuracy in complex urban environments.

Effective in handling ultra-large-scale scene reconstruction tasks.

Abstract

While Gaussian Splatting (GS) demonstrates efficient and high-quality scene rendering and small area surface extraction ability, it falls short in handling large-scale aerial image surface extraction tasks. To overcome this, we present ULSR-GS, a framework dedicated to high-fidelity surface extraction in ultra-large-scale scenes, addressing the limitations of existing GS-based mesh extraction methods. Specifically, we propose a point-to-photo partitioning approach combined with a multi-view optimal view matching principle to select the best training images for each sub-region. Additionally, during training, ULSR-GS employs a densification strategy based on multi-view geometric consistency to enhance surface extraction details. Experimental results demonstrate that ULSR-GS outperforms other state-of-the-art GS-based works on large-scale aerial photogrammetry benchmark datasets, significantly improving surface extraction accuracy in complex urban environments. Project page: https://ulsrgs.github.io.