Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

3D Gaussian Splatting with Fisheye Images: Field of View Analysis and Depth-Based Initialization

Ulas Gunes, Matias Turkulainen, Mikhail Silaev, Juho Kannala, Esa Rahtu

TL;DR

Results demonstrate the feasibility of fisheye-based 3D Gaussian Splatting and provides a benchmark for future research on wide-angle reconstruction from sparse and distorted inputs.

Abstract

We present the first evaluation of 3D Gaussian Splatting methods on real fisheye imagery with fields of view above 180\textdegree{}. Our study evaluates Fisheye-GS \cite{liao2024fisheyegslightweightextensiblegaussian} and 3DGUT \cite{wu20253dgut} on indoor and outdoor scenes captured with 200\textdegree{} fisheye cameras, with the aim of assessing the practicality of wide-angle reconstruction under severe distortion. By comparing reconstructions at 200\textdegree{}, 160\textdegree{}, and 120\textdegree{} field-of-view, we show that both methods achieve their best results at 160\textdegree{}, which balances scene coverage with image quality, while distortion at 200\textdegree{} degrades performance. To address the common failure of Structure-from-Motion (SfM) initialization at such wide angles, we introduce a depth-based alternative using UniK3D (Universal Camera Monocular 3D Estimation) \cite{piccinelli2025unik3d}. This represents the first application of UniK3D to fisheye imagery beyond 200\textdegree{}, despite the model not being trained on such data. With the number of predicted points controlled to match SfM for fairness, UniK3D produces geometrically accurate reconstructions that rival or surpass SfM, even in challenging scenes with fog, glare, or open sky. These results demonstrate the feasibility of fisheye-based 3D Gaussian Splatting and provides a benchmark for future research on wide-angle reconstruction from sparse and distorted inputs.

3D Gaussian Splatting with Fisheye Images: Field of View Analysis and Depth-Based Initialization

TL;DR

Results demonstrate the feasibility of fisheye-based 3D Gaussian Splatting and provides a benchmark for future research on wide-angle reconstruction from sparse and distorted inputs.

Abstract

We present the first evaluation of 3D Gaussian Splatting methods on real fisheye imagery with fields of view above 180\textdegree{}. Our study evaluates Fisheye-GS \cite{liao2024fisheyegslightweightextensiblegaussian} and 3DGUT \cite{wu20253dgut} on indoor and outdoor scenes captured with 200\textdegree{} fisheye cameras, with the aim of assessing the practicality of wide-angle reconstruction under severe distortion. By comparing reconstructions at 200\textdegree{}, 160\textdegree{}, and 120\textdegree{} field-of-view, we show that both methods achieve their best results at 160\textdegree{}, which balances scene coverage with image quality, while distortion at 200\textdegree{} degrades performance. To address the common failure of Structure-from-Motion (SfM) initialization at such wide angles, we introduce a depth-based alternative using UniK3D (Universal Camera Monocular 3D Estimation) \cite{piccinelli2025unik3d}. This represents the first application of UniK3D to fisheye imagery beyond 200\textdegree{}, despite the model not being trained on such data. With the number of predicted points controlled to match SfM for fairness, UniK3D produces geometrically accurate reconstructions that rival or surpass SfM, even in challenging scenes with fog, glare, or open sky. These results demonstrate the feasibility of fisheye-based 3D Gaussian Splatting and provides a benchmark for future research on wide-angle reconstruction from sparse and distorted inputs.

Paper Structure

This paper contains 21 sections, 3 equations, 1 figure, 2 tables.

Table of Contents

  1. INTRODUCTION
  2. Contributions
  3. RELATED WORK
  4. Novel View Synthesis and 3D Gaussian Splatting
  5. Fisheye Extensions of 3D Gaussian Splatting
  6. Learning-Based Methods for Monocular Fisheye Depth Estimation
  7. METHOD
  8. Camera Calibration and Data
  9. Evaluated Gaussian Splatting Methods for Fisheye Cameras
  10. Structure-from-Motion (SfM)
  11. Fisheye-GS Implementation
  12. 3DGUT Implementation
  13. Field of View Adjustment
  14. Gaussian Initialization: SfM vs UniK3D
  15. RESULTS
  16. ...and 6 more sections

Figures (1)

  • Figure 1: Example renderings from Hall and Building Out using SfM- and UniK3D-based initialization (FGS: Fisheye-GS). Towards the periphery, depth-based results degrade for both methods due to projection ambiguity, while in central regions they often recover sharper details.