Jun 2026· Photogrammetric Engineering & Remote Sensing· Vol abs/2606.28581· 1 citation· 35 references
Computer Science
Abstract
High-resolution satellite imagery demands three-dimensional (3D) reconstruction methods that deliver both speed and geometric accuracy. Recent adaptations of 3D Gaussian splatting (3DGS) to satellite imagery demonstrate strong efficiency, but reconstruction quality often degrades under diverse illumination across multi-date, high-altitude acquisitions (with small intersection angles), limiting applicability to remote sensing and vision tasks. We present SatSplat, the first framework to adapt 2D Gaussian splatting (2DGS) to satellite photogrammetry, with online camera adjustment. We approximated satellite cameras with an affine model and learned a minimal delta parameterization for in-splat camera refinement from dense observations. The formulation was implemented with a 2DGS scene representation. To handle time-varying shadows and illumination changes, we integrated geometric shadow mapping and per-camera color correction during training. Across the evaluated DFC2019 and IARPA2016 benchmark sites, SatSplat achieved strong geometric accuracy while significantly outperforming prior 3DGS-based baselines. On our processed DFC2019 benchmark, SatSplat reduced mean absolute error by 11.93% and peak video memory by 31% relative to the previous state of the art. Our approach enabled large-scale digital surface modeling with practical computational efficiency. The project page is available at https://gdaosu.github.io/satsplat.
Abstract. Satellite imagery offers a distinct advantage in Earth observation by providing expansive coverage and enabling the monitoring of inaccessible regions without physical on-site intervention, serving as a significantly more cost-effective and scalable alternative to traditional aerial or ground-based surveys. The task of 3D reconstruction from multi-view satellite images has therefore been a pivotal point of research at the intersection of photogrammetry and remote sensing. Recently, novel-view synthesis techniques such as Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS) have accelerated the accuracy and speed of topographic modeling. Among these, Earth Observation Gaussian Splatting (EOGS) has emerged as a state-of-the-art approach by adapting 3DGS to handle the unique geometric and radiometric characteristics of satellite data, including Rational Polynomial Coefficients (RPCs) and varying solar conditions. However, the standard EOGS pipeline relies on stochastic initialization, where Gaussians are distributed uniformly within a volumetric bounding box, leading to high computational overhead and dependency on aggressive pruning that can inadvertently remove critical geometric features, particularly in areas with complex urban structures. To address these limitations, we propose Bundle-Adjusted Initialization for Earth Observation Gaussian Splatting, which leverages sparse point clouds from bundle adjustment as geometric priors for Gaussian initialization. Combined with an adaptive densification strategy, our method achieves faster convergence and improved DSM accuracy on the DFC2019 dataset compared to the EOGS baseline.
Jiyong Kim, Shuang Song, Rongjun Qin· The International Archives o...· 0 citations
Abstract. Satellite imagery acquired over large areas from multiple viewpoints introduces subtle geometric misalignments that degrade the quality of derived products such as Digital Surface Models (DSMs). This paper presents a global block adjustment workflow designed to correct these errors across overlapping stereo acquisitions from the “Constellation Optique 3D” (CO3D) constellation, which captures Earth's surface at 50 cm resolution. The proposed pipeline operates in three stages: individual acquisition refinement using Space Reference Points (SRPs) as Ground Control Points; tie point extraction between overlapping scenes through two-pass image correlation; and a weighted global spatio-triangulation simultaneously optimizing attitude biases, attitude drifts, and per-satellite magnification parameters. Applied to a large stereo acquisition dataset over the Aorounga crater, Chad, the method demonstrates strong geometric performance. The results highlight that careful parameterization — combining observation weighting, n-tuple point filtering, and per-satellite sensor refinement — is key to producing accurate, geometrically consistent large-scalemosaics from bi-satellite stereo imagery. This paper does not include the in-orbit performances due to confidentiality agreement.
Michaël Erblang, Emelyne Saulnier, Guillaume Laurent et al.· The International Archives o...· 2 citations
True digital orthophoto maps (TDOMs) accurately represent the true spatial positions and visual appearances of ground objects at the urban scale, and constitute a fundamental remote sensing product for large-scale urban digitization and fine-grained geographic modeling. Conventional TDOM generation methods that rely on digital elevation models (DEMs) or digital surface models (DSMs) are highly susceptible to occlusions and elevation errors in complex urban environments, often resulting in geometric distortions and visual artifacts. Neural radiance field (NeRF)-based approaches can mitigate these issues, but their high computational cost limits practical deployment on large-scale satellite imagery. To address these challenges, we present Tortho–SatGS, a pure vision-based framework that, to the best of our knowledge, is among the first to systematically integrate 3-D Gaussian Splatting (3DGS) into satellite true orthophoto generation. Specifically, we design a 3DGS-based geometric modeling pipeline tailored to satellite imaging geometry and introduce an orthographic rasterization-based rendering scheme to effectively resolve building side facades and edge curvature artifacts, enabling DEM/DSM-free TDOM generation with improved geometric consistency. In addition, a VGG-based perceptual loss is incorporated to complement pixel-level supervision, improving radiometric consistency and fine-grained texture fidelity, particularly in shadowed and low-texture regions. Experimental results on two real-world satellite datasets demonstrate that Tortho–SatGS consistently outperforms conventional methods in terms of geometric accuracy, texture quality, and radiometric consistency. Compared with NeRF-based approaches, our method achieves approximately $27\times $ faster training speed and $15\times $ faster rendering speed, effectively balancing reconstruction accuracy and computational efficiency. These results validate the effectiveness and practical advantages of 3DGS-based approaches for the generation of true orthophotos of satellites.
Junxing Yang, Wenya Bian, Xingcheng Liu et al.· IEEE Transactions on Geoscie...· 0 citations
Accurate 3D reconstruction from satellite imagery typically relies on near-simultaneous stereo pairs, limiting its applicability to diachronic settings where multi-date images exhibit varying seasonal and illumination conditions. Training dense stereo matching models robust to appearance changes is a long-standing challenge, as aligned multi-date imagery and ground-truth geometry are costly to obtain at scale. We propose SeasonStereo, a scalable framework that addresses disparity estimation from diachronic satellite images by training on synthetic image pairs with controlled seasonal appearance variation, while leveraging zero-shot geometric priors from foundation models. SeasonStereo matches the accuracy of state-of-the-art LiDAR-supervised models, while producing sharper geometric details without requiring aligned real multi-date training products or LiDAR-derived labels. As a result, SeasonStereo offers a practical path toward large-scale 3D reconstruction from heterogeneous satellite images with reduced supervision cost.
Álvaro Díaz-Laureano, Roger Mar'i, El'ias Masquil et al.· 0 citations
Novel view synthesis, which is essential for three-dimensional (3D) reconstruction, traditionally requires dense image sets and pre-calibrated camera parameters. However, in industrial digital twin applications spatial constraints often limit data acquisition to sparse views that fail to provide the feature correspondences necessary for accurate modeling. To overcome this bottleneck, a semantic-guided 3D Gaussian splatting (3DGS) framework tailored to sparse-view industrial reconstruction was introduced. Rather than relying on precise initialization and dense inputs, the proposed method couples explicit 3D Gaussian representations with the dense and unconstrained stereo 3D reconstruction (DUSt3R) end-to-end pose estimation model. The severe occlusions, typical of factory environments, are handled by incorporating Segment Anything Model 2 (SAM2) to hierarchically decompose the scene, yielding a structured representation that improves optimization stability. A probability density field-driven algorithm is subsequently applied to extract lightweight meshes directly from the optimized Gaussian point clouds. Evaluations on the MipNeRF360 benchmark and a custom industrial dataset demonstrated substantial improvements in the structural similarity index measure. By enabling robust reconstruction from limited viewpoints, this pipeline offers a practical geometric foundation for automated inspection and remote equipment monitoring.
Boyang Li, Tianhan Gao, Zuan Gu et al.· Visual Computing for Industr...· 0 citations
SLAM methods based on 3D Gaussian Splatting (3DGS) have demonstrated impressive tracking and mapping performance, but typically require additional geometric information from external depth sensors. Meanwhile, recent SLAM systems that leverage geometric priors from pre-trained feed-forward models enable real-time dense reconstruction, yet often discard original RGB information during optimization, thus degrading overall reconstruction quality. We present GeoGS-SLAM, an online monocular dense reconstruction system that combines the 3DGS-based map representation with learned geometric priors. Given uncalibrated RGB input, we first employ a feed-forward visual geometry model to predict camera and scene priors. The Gaussian scene map is then expanded by directly sampling Gaussian primitives from both RGB input and geometric priors. Camera poses and the scene map are jointly optimized through a coarse-to-fine strategy that minimizes both photometric and geometric losses. To ensure global consistency, we further incorporate online loop closure detection and pose graph optimization. Extensive experiments across indoor and outdoor benchmarks demonstrate that GeoGS-SLAM achieves superior rendering quality and tracking accuracy compared to state-of-the-art methods while maintaining online real-time performance. Project page: https://rlgao.github.io/geogs_slam.