DirectFisheye-GS: Enabling Native Fisheye Input in Gaussian Splatting with Cross-View Joint Optimization

2026-04-01

AI Summary

This work addresses the limitations of conventional 3D Gaussian splatting methods when applied to fisheye images, which typically require prior undistortion and often suffer from peripheral information loss and floating artifacts. The paper presents the first approach that directly integrates a fisheye camera model into the 3D Gaussian splatting framework, enabling native processing of fisheye imagery without preprocessing. It introduces a cross-view joint optimization strategy based on feature overlap to enhance multi-view geometric and photometric consistency. The proposed method achieves high-quality reconstruction on public benchmarks, matching or surpassing state-of-the-art performance while significantly improving detail fidelity and stability in fisheye scenes. Notably, the introduced optimization mechanism is also applicable to pinhole camera systems.

Abstract

3D Gaussian Splatting (3DGS) has enabled efficient 3D scene reconstruction from everyday images with real-time, high-fidelity rendering, greatly advancing VR/AR applications. Fisheye cameras, with their wider field of view (FOV), promise high-quality reconstructions from fewer inputs and have recently attracted much attention. However, since 3DGS relies on rasterization, most subsequent works involving fisheye camera inputs first undistort images before training, which introduces two problems: 1) Black borders at image edges cause information loss and negate the fisheye's large FOV advantage; 2) Undistortion's stretch-and-interpolate resampling spreads each pixel's value over a larger area, diluting detail density -- causes 3DGS overfitting these low-frequency zones, producing blur and floating artifacts. In this work, we integrate fisheye camera model into the original 3DGS framework, enabling native fisheye image input for training without preprocessing. Despite correct modeling, we observed that the reconstructed scenes still exhibit floaters at image edges: Distortion increases toward the periphery, and 3DGS's original per-iteration random-selecting-view optimization ignores the cross-view correlations of a Gaussian, leading to extreme shapes (e.g., oversized or elongated) that degrade reconstruction quality. To address this, we introduce a feature-overlap-driven cross-view joint optimization strategy that establishes consistent geometric and photometric constraints across views-a technique equally applicable to existing pinhole-camera-based pipelines. Our DirectFisheye-GS matches or surpasses state-of-the-art performance on public datasets.

Problem

• Fisheye Camera
• 3D Gaussian Splatting
• Image Distortion
• Cross-View Optimization
• Reconstruction Artifacts

Innovation

• Gaussian Splatting
• Fisheye Camera
• Cross-View Optimization
• Native Distortion Handling
• 3D Reconstruction