Raindrop GS: A Benchmark for 3D Gaussian Splatting under Raindrop Conditions

2025-10-20

AI Summary

Under realistic raindrop interference, 3D Gaussian Splatting (3DGS) reconstruction suffers severe degradation—including lens occlusion, optical distortion, inaccurate camera pose estimation, and unreliable point cloud initialization. Method: This paper introduces the first comprehensive evaluation benchmark for 3DGS under unconstrained real-rain conditions. It provides a multi-focus-aligned image dataset (raindrop-focused, background-focused, and rain-free ground truth) and establishes an end-to-end evaluation pipeline covering image deraining, pose estimation, point cloud initialization, and 3DGS reconstruction. Contribution/Results: Experiments quantitatively reveal the critical impact of lens focus position and pose initialization on reconstruction fidelity, exposing significant performance bottlenecks of existing methods under real rain. The benchmark fills a critical gap in robustness evaluation of 3DGS under realistic precipitation, offering a reproducible, quantitative foundation for algorithmic advancement.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Calibration & Uncertainty Quantification
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Security and Privacy: Large-scale security measurements
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

3D Gaussian Splatting (3DGS) under raindrop conditions suffers from severe occlusions and optical distortions caused by raindrop contamination on the camera lens, substantially degrading reconstruction quality. Existing benchmarks typically evaluate 3DGS using synthetic raindrop images with known camera poses (constrained images), assuming ideal conditions. However, in real-world scenarios, raindrops often interfere with accurate camera pose estimation and point cloud initialization. Moreover, a significant domain gap between synthetic and real raindrops further impairs generalization. To tackle these issues, we introduce RaindropGS, a comprehensive benchmark designed to evaluate the full 3DGS pipeline-from unconstrained, raindrop-corrupted images to clear 3DGS reconstructions. Specifically, the whole benchmark pipeline consists of three parts: data preparation, data processing, and raindrop-aware 3DGS evaluation, including types of raindrop interference, camera pose estimation and point cloud initialization, single image rain removal comparison, and 3D Gaussian training comparison. First, we collect a real-world raindrop reconstruction dataset, in which each scene contains three aligned image sets: raindrop-focused, background-focused, and rain-free ground truth, enabling a comprehensive evaluation of reconstruction quality under different focus conditions. Through comprehensive experiments and analyses, we reveal critical insights into the performance limitations of existing 3DGS methods on unconstrained raindrop images and the varying impact of different pipeline components: the impact of camera focus position on 3DGS reconstruction performance, and the interference caused by inaccurate pose and point cloud initialization on reconstruction. These insights establish clear directions for developing more robust 3DGS methods under raindrop conditions.

Problem

• Evaluating 3D Gaussian Splatting performance under raindrop occlusion and distortion
• Addressing domain gap between synthetic and real raindrop image conditions
• Analyzing impact of inaccurate camera pose estimation on 3D reconstruction

Innovation

• Introduces real-world raindrop reconstruction benchmark dataset
• Evaluates full 3DGS pipeline from corrupted images to reconstruction
• Analyzes impact of camera focus and pose estimation issues