3D Gaussian Splatting with Self-Constrained Priors for High Fidelity Surface Reconstruction

2026-03-20

AI Summary

Existing 3D Gaussian splatting methods still face challenges in achieving sufficient geometric accuracy for high-fidelity surface reconstruction. This work proposes a self-constrained prior mechanism that iteratively fuses depth maps rendered from the current Gaussian representation to generate a truncated signed distance function (TSDF) mesh, which dynamically constrains the positions and opacities of the Gaussians. This process establishes a progressively tightening band-like geometric prior, enabling geometry-aware Gaussian optimization. The proposed approach significantly outperforms current state-of-the-art methods across multiple standard benchmarks, effectively enhancing both the accuracy and completeness of reconstructed surface details.

Technology Category

• Computer Vision: 3D Computer Vision
• Search and Optimization: Sampling/Simulation-based Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Rendering 3D surfaces has been revolutionized within the modeling of radiance fields through either 3DGS or NeRF. Although 3DGS has shown advantages over NeRF in terms of rendering quality or speed, there is still room for improvement in recovering high fidelity surfaces through 3DGS. To resolve this issue, we propose a self-constrained prior to constrain the learning of 3D Gaussians, aiming for more accurate depth rendering. Our self-constrained prior is derived from a TSDF grid that is obtained by fusing the depth maps rendered with current 3D Gaussians. The prior measures a distance field around the estimated surface, offering a band centered at the surface for imposing more specific constraints on 3D Gaussians, such as removing Gaussians outside the band, moving Gaussians closer to the surface, and encouraging larger or smaller opacity in a geometry-aware manner. More importantly, our prior can be regularly updated by the most recent depth images which are usually more accurate and complete. In addition, the prior can also progressively narrow the band to tighten the imposed constraints. We justify our idea and report our superiority over the state-of-the-art methods in evaluations on widely used benchmarks.

Problem

• 3D Gaussian Splatting
• surface reconstruction
• high fidelity
• depth rendering
• radiance fields

Innovation

• 3D Gaussian Splatting
• self-constrained prior
• TSDF
• surface reconstruction
• depth rendering