WHU-Stereo: A Challenging Benchmark for Stereo Matching of High-Resolution Satellite Images

Li, Shenhong; He, Sheng; Jiang, San; Jiang, Wanshou; Zhang, Lin

doi:10.1109/tgrs.2023.3245205

Cited by 8 publications

(4 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Additionally, we used two versions of ImMPI [20] during comparison. One version, denoted as 'ImMPI * ', was first pre-trained on WHU [36] and then trained with 11 views. The other version, denoted as 'ImMPI † ', was not pre-trained, but was trained using all 21 images.…”

Section: Baselinementioning

confidence: 99%

UniRender: Reconstructing 3D Surfaces from Aerial Images with a Unified Rendering Scheme

Yan,

Zhou,

et al. 2023

Remote Sensing

View full text Add to dashboard Cite

While recent advances in the field of neural rendering have shown impressive 3D reconstruction performance, it is still a challenge to accurately capture the appearance and geometry of a scene by using neural rendering, especially for remote sensing scenes. This is because both rendering methods, i.e., surface rendering and volume rendering, have their own limitations. Furthermore, when neural rendering is applied to remote sensing scenes, the view sparsity and content complexity that characterize these scenes will severely hinder its performance. In this work, we aim to address these challenges and to make neural rendering techniques available for 3D reconstruction in remote sensing environments. To achieve this, we propose a novel 3D surface reconstruction method called UniRender. UniRender offers three improvements in locating an accurate 3D surface by using neural rendering: (1) unifying surface and volume rendering by employing their strengths while discarding their weaknesses, which enables accurate 3D surface position localization in a coarse-to-fine manner; (2) incorporating photometric consistency constraints during rendering, and utilizing the points reconstructed by structure from motion (SFM) or multi-view stereo (MVS), to constrain reconstructed surfaces, which significantly improves the accuracy of 3D reconstruction; (3) improving the sampling strategy by locating sampling points in the foreground regions where the surface needs to be reconstructed, thus obtaining better detail in the reconstruction results. Extensive experiments demonstrate that UniRender can reconstruct high-quality 3D surfaces in various remote sensing scenes.

show abstract

Section: Baselinementioning

confidence: 99%

UniRender: Reconstructing 3D Surfaces from Aerial Images with a Unified Rendering Scheme

Yan,

Zhou,

et al. 2023

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Typically, creating high-quality LoD2 models involves a manual and very expensive process, while recent research efforts aim to automate this process. Out of many sources, veryhigh-resolution (VHR) satellite stereo imagery (with ground sampling distance (GSD) < 1 m) is beneficial due to its global coverage and low cost per unit area (Facciolo et al, 2017;Li et al, 2023b). Previous works have shown that it is possible to reconstruct LoD2 (Gui and Qin, 2021;Gui et al, 2022;Partovi et al, 2019) from such data, which typically follow a standard process takes pre-processed digital surface model (DSM) and orthophotos from stereo satellite imagery as input data: first, perform building detection to obtain building masks; second, vectorize individual building masks with topologically consistent line primitives, third, determine the types of roofs and then join individual small buildings into more complex building models.…”

Section: Introduction 11 Backgroundmentioning

confidence: 99%

Unit-level LoD2 Building Reconstruction from Satellite-derived Digital Surface Model and Orthophoto

Gui,

Schuegraf,

Bittner

et al. 2024

ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

Abstract. Recent advancements in deep learning have enabled the possibility to identify unit-level building sections from very high resolution satellite images. By learning from the examples, deep models can capture patterns from the low-resolution roof textures to separate building units from duplex buildings. This paper demonstrates that such unit-level segmentation can further advance level of details (LoD)2 modeling. We extend a building boundary regularization method by adapting noisy unit-level segmentation results. Specifically, we propose a novel polygon composition approach to ensure the individually segmented units within a duplex building or dense adjacent buildings are consistent in their shared boundaries. Results of the experiments show that, our unit-level LoD2 modeling has favorably outperformed the state-of-the-art LoD2 modeling results from satellite images.

show abstract

“…To evaluate the performance in the high-resolution satellite datasets, we cannot just rely on the results in close-range datasets (Li et al, 2023a). Therefore, this paper studies several classic deep learning models and evaluates the effectiveness of deep learning algorithms on several datasets, including SceneFlow (Mayer et al, 2016), KITTI 2015 (Menze and Geiger, 2015), US3D (Bosch et al, 2019, Le Saux et al, 2019, and WHU-Stereo (Li et al, 2023c).…”

Section: Introductionmentioning

confidence: 99%

Deep Learning-Based Stereo Matching for High-Resolution Satellite Images: A Comparative Evaluation

He,

Jiang,

et al. 2023

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

Abstract. Dense matching plays an important role in 3D modeling from satellite images. Its purpose is to establish pixel-by-pixel correspondences between two stereo images. The most well-known algorithm is the semi-global matching (SGM), which can generate high-quality 3D models with high computational efficiency. Due to the complex coverage and imaging condition, SGM cannot cope with these situation well. In recent years, deep learning-based stereo matching has attracted wide attention and shown overwhelming benefits over traditional algorithms in terms of precision and completeness. However, existing models are usually evaluated by using close-ranging datasets. Thus, this study investigates the recent deep learning models and evaluate their performance on both close-ranging and satellite image datasets. The results demonstrate that deep learning network can better adapt to the satellite dataset than the typical SGM. Meanwhile, the generalization ability of deep learning-based models is still low for the real application at recent time.

show abstract

WHU-Stereo: A Challenging Benchmark for Stereo Matching of High-Resolution Satellite Images

Cited by 8 publications

References 37 publications

UniRender: Reconstructing 3D Surfaces from Aerial Images with a Unified Rendering Scheme

UniRender: Reconstructing 3D Surfaces from Aerial Images with a Unified Rendering Scheme

Unit-level LoD2 Building Reconstruction from Satellite-derived Digital Surface Model and Orthophoto

Deep Learning-Based Stereo Matching for High-Resolution Satellite Images: A Comparative Evaluation

Contact Info

Product

Resources

About