GeoGS-SLAM: Online Monocular Reconstruction Using Gaussian Splatting with Geometric Priors
Ruilan Gao, Letian Jin, Yu Zhang
AI summary
Problem
Existing monocular 3DGS-based SLAM degrades without depth sensors, while geometric prior-based SLAM often discards RGB data during optimization, compromising reconstruction fidelity and visual consistency.
Approach
The system predicts camera and scene priors from uncalibrated RGB using a pre-trained model, directly samples Gaussian primitives from both images and priors, and jointly optimizes poses and the map by minimizing photometric and geometric losses with online loop closure.
Key results
- State-of-the-art rendering quality with PSNR gains up to +4.94
- Tracking error reduced by up to 64.6% versus prior-based SLAM
- Real-time online performance across indoor and outdoor benchmarks
- Robust global consistency via integrated loop closure and pose graph optimization
Why it matters
Provides a robust, sensor-light solution for high-fidelity 3D mapping and localization, advancing robotics, autonomous driving, and AR systems that rely solely on monocular cameras.
Abstract
SLAM methods based on 3D Gaussian Splatting (3DGS) have demonstrated impressive tracking and mapping performance, but typically require additional geometric infor- mation from external depth sensors. Meanwhile, recent SLAM systems that leverage geometric priors from pre-trained feed- forward models enable real-time dense reconstruction, yet often discard original RGB information during optimization, thus degrading overall reconstruction quality. We present GeoGS- SLAM, an online monocular dense reconstruction system that combines the 3DGS-based map representation with learned ge- ometric priors. Given uncalibrated RGB input, we first employ a feed-forward visual geometry model to predict camera and scene priors. The Gaussian scene map is then expanded by di- rectly sampling Gaussian primitives from both RGB input and geometric priors. Camera poses and the scene map are jointly optimized through a coarse-to-fine strategy that minimizes both photometric and geometric losses. To ensure global consistency, we further incorporate online loop closure detection and pose graph optimization. Extensive experiments across indoor and outdoor benchmarks demonstrate that GeoGS-SLAM achieves superior rendering quality and tracking accuracy compared to state-of-the-art methods while maintaining online real-time performance. Project page: https://rlgao.github.io/ geogs_slam. This work was supported by the National Natural Science Foundation of China (Grant No. 62576311), in part by NSFC 62088101 Autonomous Intelligent Unmanned Systems, and in part by Zhejiang Provincial Natural Science Foundation of China under Grant No. LD24F030001. 1State Key Laboratory of Industrial Control Technology, College of Control Science and Engineering, Zhejiang University, Hangzhou, China, 310027. 2Key Laboratory of Collaborative Sensing and Autonomous Unmanned Systems of Zhejiang Province, Hangzhou, China, 310027. ̊Corresponding author: Yu Zhang (Email: zhangyu80@zju.edu.cn).