Robust and High-Fidelity 3D Gaussian Splatting: Fusing Pose Priors and Geometry Constraints for Texture-Deficient Outdoor Scenes

TL;DR

This paper tackles the challenges of robust pose estimation and geometry-consistent 3D Gaussian Splatting (3DGS) in large outdoor scenes with weak or repetitive textures. It couples LiDAR-Inertial Odometry (LIO) pose priors with COLMAP triangulation to achieve metric-scale, robust camera poses, and introduces normal vector constraints along with an effective rank regularization to guide Gaussian primitives toward coherent geometry. The proposed map optimization combines photometric loss with normal supervision from Omnidata and shape regularization, yielding improved rendering fidelity and geometric consistency, especially in texture-deficient regions. Experiments on public and self-collected datasets show faster, more robust pose optimization and superior geometric results compared to state-of-the-art baselines, highlighting the practical impact for real-world digital asset creation in challenging outdoor environments.

Abstract

3D Gaussian Splatting (3DGS) has emerged as a key rendering pipeline for digital asset creation due to its balance between efficiency and visual quality. To address the issues of unstable pose estimation and scene representation distortion caused by geometric texture inconsistency in large outdoor scenes with weak or repetitive textures, we approach the problem from two aspects: pose estimation and scene representation. For pose estimation, we leverage LiDAR-IMU Odometry to provide prior poses for cameras in large-scale environments. These prior pose constraints are incorporated into COLMAP's triangulation process, with pose optimization performed via bundle adjustment. Ensuring consistency between pixel data association and prior poses helps maintain both robustness and accuracy. For scene representation, we introduce normal vector constraints and effective rank regularization to enforce consistency in the direction and shape of Gaussian primitives. These constraints are jointly optimized with the existing photometric loss to enhance the map quality. We evaluate our approach using both public and self-collected datasets. In terms of pose optimization, our method requires only one-third of the time while maintaining accuracy and robustness across both datasets. In terms of scene representation, the results show that our method significantly outperforms conventional 3DGS pipelines. Notably, on self-collected datasets characterized by weak or repetitive textures, our approach demonstrates enhanced visualization capabilities and achieves superior overall performance. Codes and data will be publicly available at https://github.com/justinyeah/normal_shape.git.

Figures (3)

• Figure 1: Overview of our system.We utilize LiDAR-Inertial Odometry for initial metric-aware camera poses. Initial and relative constraints refine these poses. By integrating normal and shape constraints, the geometric texture consistency of the scene representation is significantly improved.
• Figure 2: Illustration of the angle between the Gaussian normal and the camera optical axis.
• Figure 3: Comparison of Rendering Results.