EOGS++: Earth Observation Gaussian Splatting with Internal Camera Refinement and Direct Panchromatic Rendering

2025-11-20

AI Summary

Existing 3D reconstruction methods from high-resolution satellite panchromatic imagery suffer from insufficient geometric accuracy and rely heavily on external preprocessing for camera pose estimation. Method: We propose EOGS++, the first end-to-end 3D Gaussian splatting framework that directly takes raw panchromatic images as input. It innovatively embeds bundle adjustment (BA) into Gaussian point cloud training to jointly optimize camera poses and Gaussian parameters; integrates optical-flow-guided inlier constraints to enhance BA robustness; and combines early stopping with TSDF fusion to improve geometric consistency. Results: On the IARPA 2016 and DFC2019 benchmarks, EOGS++ reduces the mean absolute error (MAE) over building regions from 1.33 m to 1.19 m—outperforming both the prior EOGS and NeRF-based approaches. It achieves a unified trade-off between high-fidelity novel-view synthesis and metrically accurate 3D reconstruction.

Technology Category

• Computer Vision: 3D Computer Vision
• Search and Optimization: Sampling/Simulation-based Search
• Intelligent Robots: State Estimation

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Recently, 3D Gaussian Splatting has been introduced as a compelling alternative to NeRF for Earth observation, offering com- petitive reconstruction quality with significantly reduced training times. In this work, we extend the Earth Observation Gaussian Splatting (EOGS) framework to propose EOGS++, a novel method tailored for satellite imagery that directly operates on raw high-resolution panchromatic data without requiring external preprocessing. Furthermore, leveraging optical flow techniques we embed bundle adjustment directly within the training process, avoiding reliance on external optimization tools while improving camera pose estimation. We also introduce several improvements to the original implementation, including early stopping and TSDF post-processing, all contributing to sharper reconstructions and better geometric accuracy. Experiments on the IARPA 2016 and DFC2019 datasets demonstrate that EOGS++ achieves state-of-the-art performance in terms of reconstruction quality and effi- ciency, outperforming the original EOGS method and other NeRF-based methods while maintaining the computational advantages of Gaussian Splatting. Our model demonstrates an improvement from 1.33 to 1.19 mean MAE errors on buildings compared to the original EOGS models

Problem

• Improving 3D reconstruction from satellite imagery using Gaussian Splatting
• Refining camera poses internally without external optimization tools
• Processing raw panchromatic data directly without preprocessing steps

Innovation

• Direct panchromatic rendering without preprocessing
• Internal camera refinement using optical flow
• Early stopping and TSDF post-processing enhancements