GTLR-GS: Geometry-Texture Aware LiDAR-Regularized 3D Gaussian Splatting for Realistic Scene Reconstruction

2026-03-24

AI Summary

This work proposes a LiDAR-centric 3D Gaussian splatting framework that overcomes key limitations of conventional approaches, which rely on sparse Structure-from-Motion (SfM) point clouds and suffer from scale ambiguity, geometric inconsistency, and view dependency. By explicitly incorporating metric geometric priors during optimization, the reconstruction problem is reformulated as a geometry-conditioned allocation and refinement task under a fixed representation budget. The core contributions include a geometry- and texture-aware Gaussian initialization strategy, a curvature-adaptive Gaussian splitting mechanism, and a confidence-aware LiDAR-driven depth regularization. Experiments demonstrate that the proposed method achieves high-fidelity, metric-accurate reconstructions on both the ScanNet++ benchmark and a newly collected real-world dataset, outperforming existing state-of-the-art techniques.

Technology Category

• Computer Vision: 3D Computer Vision
• Search and Optimization: Learning to Search
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work

Abstract

Recent advances in 3D Gaussian Splatting (3DGS) have enabled real-time, photorealistic scene reconstruction. However, conventional 3DGS frameworks typically rely on sparse point clouds derived from Structure-from-Motion (SfM), which inherently suffer from scale ambiguity, limited geometric consistency, and strong view dependency due to the lack of geometric priors. In this work, a LiDAR-centric 3D Gaussian Splatting framework is proposed that explicitly incorporates metric geometric priors into the entire Gaussian optimization process. Instead of treating LiDAR data as a passive initialization source, 3DGS optimization is reformulated as a geometry-conditioned allocation and refinement problem under a fixed representational budget. Specifically, this work introduces (i) a geometry-texture-aware allocation strategy that selectively assigns Gaussian primitives to regions with high structural or appearance complexity, (ii) a curvature-adaptive refinement mechanism that dynamically guides Gaussian splitting toward geometrically complex areas during training, and (iii) a confidence-aware metric depth regularization that anchors the reconstructed geometry to absolute scale using LiDAR measurements while maintaining optimization stability. Extensive experiments on the ScanNet++ dataset and a custom real-world dataset validate the proposed approach. The results demonstrate state-of-the-art performance in metric-scale reconstruction with high geometric fidelity.

Problem

• 3D Gaussian Splatting
• LiDAR
• metric reconstruction
• geometric consistency
• scale ambiguity

Innovation

• LiDAR-regularized
• 3D Gaussian Splatting
• geometry-texture aware allocation
• curvature-adaptive refinement
• metric depth regularization