Robust LiDAR-Camera Calibration with 2D Gaussian Splatting

Shuyi Zhou, Shuxiang Xie, Ryoichi Ishikawa, Takeshi Oishi

The University of Tokyo

Accepted to IEEE Robotics and Automation Letters (Ra-L), 2025

Arxiv Code Data IEEE Xplore Appendix

Introduction

We present a robust method for LiDAR-camera calibration leveraging 2D Gaussian Splatting.The proposed method begins by reconstructing colorless 2DGS using LiDAR point clouds. Subsequently, we update the colors of the Gaussian splats by minimizing the photometric loss. The extrinsic parameters are optimized during this process. Additionally, we address the limitations of the photometric loss by incorporating the reprojection and triangulation losses, thereby enhancing the calibration robustness and accuracy.

Teaser Image

Training Process of 2DGS Geometry

Visualization of 2DGS geometry training process.

Calibration Process

Simultaneous LiDAR-camera cailbration and coloring for 2DGS.

BibTeX

@article{zhou2025robust,
  title={Robust LiDAR-Camera Calibration With 2D Gaussian Splatting},
  author={Zhou, Shuyi and Xie, Shuxiang and Ishikawa, Ryoichi and Oishi, Takeshi},
  journal={IEEE Robotics and Automation Letters},
  year={2025},
  publisher={IEEE}
}