pith. sign in

arxiv: 2507.13486 · v2 · submitted 2025-07-17 · 💻 cs.CV

Uncertainty Quantification Framework for Aerial and UAV Photogrammetry through Error Propagation

Debao Huang , Rongjun Qin This is my paper

Pith reviewed 2026-05-19 03:49 UTC · model grok-4.3

classification 💻 cs.CV
keywords uncertainty quantificationphotogrammetrymulti-view stereostructure from motionerror propagationUAV imagerypoint cloud accuracydisparity uncertainty
0
0 comments X

The pith

A self-calibrating method regresses disparity uncertainty from MVS matching costs using reliable n-view points to produce per-point covariances for photogrammetric point clouds.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes a complete uncertainty quantification framework for aerial and UAV photogrammetry that accounts for error propagation through the two-step process of Structure-from-Motion with bundle adjustment followed by Multi-view Stereo. It closes the gap in the MVS stage by introducing a self-calibrating approach that selects reliable points visible in at least six views directly from the reconstruction and regresses their disparity uncertainty from matching-cost and related cues. A sympathetic reader would care because photogrammetric point cloud accuracy varies strongly with scene geometry and object complexity, unlike more uniform LiDAR returns, so per-point accuracy credentials improve usability in mapping and analysis tasks. The resulting covariances are shown to achieve high bounding rates on public datasets without systematic overestimation.

Core claim

The paper claims that reliable n-view points extracted directly from the MVS process can be used in a self-supervised regression to estimate disparity uncertainty from matching-cost cues, thereby yielding an error covariance matrix per point that respects the full error propagation path of SfM followed by MVS and supplies certifiable uncertainty across diverse scenes.

What carries the argument

Self-calibrating regression of disparity uncertainty from matching-cost cues on reliable n-view points (n>=6) per view extracted from the MVS process, which generates per-point covariances that close the MVS uncertainty gap.

If this is right

  • Per-point covariance matrices are produced that follow the natural error propagation path from SfM/BA through MVS.
  • The estimates achieve higher bounding rates than prior methods while avoiding overestimation on varied airborne and UAV imagery.
  • The approach remains self-supervised and requires no external calibration targets or labeled uncertainty data.
  • The framework supplies scene-dependent accuracy credentials that are directly usable for downstream tasks relying on photogrammetric point clouds.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Per-point uncertainties of this form could support weighted fusion of photogrammetric clouds with LiDAR or other sensors by providing explicit variance information for each point.
  • Downstream algorithms that filter or smooth point clouds could use the covariances to adaptively weight contributions according to local reliability.
  • Extending the regression cues to include additional MVS statistics such as consistency across more views might further tighten the uncertainty bounds on complex surfaces.

Load-bearing premise

The chosen reliable n-view points must be representative of the full point cloud and free of systematic bias so that the regression produces valid covariances everywhere.

What would settle it

On a held-out aerial or UAV dataset with independent ground-truth point positions, the predicted uncertainty ellipsoids would fail to contain the true errors at the reported high bounding rate.

read the original abstract

Uncertainty quantification of the photogrammetry process is essential for providing per-point accuracy credentials of the point clouds. Unlike airborne LiDAR, whose accuracy generally remains consistent with objects with varying geometric complexity, the accuracy of photogrammetric point clouds is rather object/scene-dependent, as it relies on algorithm-derived measurements. Generally, errors of the photogrammetric point clouds propagate through a two-step process: Structure-from-Motion (SfM) with Bundle adjustment (BA), followed by Multi-view Stereo (MVS). While uncertainty estimation in the SfM stage has been well studied using the first-order statistics of the reprojection error function, that in the MVS stage remains largely unsolved and non-standardized, primarily due to its non-differentiable and multi-modal nature (i.e., from pixel values to geometry). In this paper, we present an uncertainty quantification framework closing this gap by associating an error covariance matrix per point accounting for this two-step photogrammetry process. Specifically, to estimate the uncertainty in the MVS stage, we propose a novel, self-calibrating method by taking reliable n-view points (n>=6) per-view to regress the disparity uncertainty using highly relevant cues (such as matching cost values) from the MVS stage. Compared to existing approaches, our method uses self-contained, reliable 3D points extracted directly from the MVS process, with the benefit of being self-supervised and naturally adhering to error propagation path of the photogrammetry process, thereby providing a robust and certifiable uncertainty quantification across diverse scenes. We evaluate the framework using a variety of publicly available airborne and UAV imagery datasets. Results demonstrate that our method outperforms existing approaches by achieving high bounding rates without overestimating uncertainty.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript claims to introduce an uncertainty quantification framework for aerial and UAV photogrammetry that accounts for error propagation through the SfM/BA and MVS stages. For the MVS stage, it introduces a self-calibrating regression approach that selects reliable n-view points (n >= 6) from the MVS output to train a model predicting disparity uncertainty based on matching-cost cues and other relevant features, thereby generating per-point covariance matrices that are said to be self-supervised and consistent with the photogrammetry error propagation path. The framework is tested on various public airborne and UAV datasets, with results indicating superior performance over existing methods in terms of uncertainty bounding rates without overestimation.

Significance. If the proposed regression produces unbiased uncertainty estimates that generalize beyond the selected reliable points, this work would provide a valuable, practical tool for certifying the accuracy of photogrammetric point clouds in a manner that respects the underlying reconstruction pipeline. This is particularly relevant for applications in surveying and mapping where per-point uncertainty is crucial for downstream tasks, and the self-supervised nature avoids the need for external ground truth.

major comments (2)
  1. [Abstract] Abstract (self-calibrating method description): The central claim that regressing disparity uncertainty from matching-cost cues on reliable n-view points (n>=6) produces valid covariances for the full point cloud rests on the unverified assumption that these points are representative without systematic bias. Points selected for high view count may preferentially sample lower-error regimes, causing the fitted regressor to understate uncertainty on remaining points and breaking adherence to the two-step error-propagation path.
  2. [Evaluation] Evaluation (results description): The abstract states that the method achieves high bounding rates without overestimating uncertainty, but provides no quantitative validation details, error-bar analysis, cross-validation metrics, or ablation of the n-view selection rule. This absence makes it impossible to confirm that the learned mapping certifies uncertainties independently of the MVS output used for training.
minor comments (2)
  1. [Abstract] The abstract mentions 'highly relevant cues (such as matching cost values)' but does not specify the full set of input features or their preprocessing; a table listing the cues and their definitions would improve clarity.
  2. [Introduction] No diagram is referenced illustrating the two-step error propagation path from SfM/BA to MVS disparity uncertainty to final covariance; adding one would aid reader understanding of how the regression fits into the overall framework.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and detailed comments. We address each major comment below, indicating where we will revise the manuscript to incorporate the feedback.

read point-by-point responses

  1. Referee: [Abstract] Abstract (self-calibrating method description): The central claim that regressing disparity uncertainty from matching-cost cues on reliable n-view points (n>=6) produces valid covariances for the full point cloud rests on the unverified assumption that these points are representative without systematic bias. Points selected for high view count may preferentially sample lower-error regimes, causing the fitted regressor to understate uncertainty on remaining points and breaking adherence to the two-step error-propagation path.

    Authors: We thank the referee for raising this important point on potential selection bias. Our choice of reliable n-view points (n >= 6) is motivated by their multi-view consistency, which provides a self-supervised proxy for reliable disparity estimates directly from the MVS output. The regression leverages matching-cost cues and other features that are expected to reflect uncertainty sources independently of view count. While high-n points may on average exhibit lower errors, the framework is designed to propagate these estimates through the full photogrammetry pipeline. To address the concern, we will add to the revised manuscript a quantitative comparison of error statistics and uncertainty coverage between the selected n-view points and the remaining points, along with a discussion of how this supports generalization within the two-stage error-propagation model. revision: yes

  2. Referee: [Evaluation] Evaluation (results description): The abstract states that the method achieves high bounding rates without overestimating uncertainty, but provides no quantitative validation details, error-bar analysis, cross-validation metrics, or ablation of the n-view selection rule. This absence makes it impossible to confirm that the learned mapping certifies uncertainties independently of the MVS output used for training.

    Authors: We agree that the abstract is concise and that additional quantitative details would strengthen the evaluation. The full manuscript reports results across multiple public airborne and UAV datasets with superior bounding performance. In the revision we will expand the evaluation section to include cross-validation of the regression model (e.g., hold-out sets of n-view points), ablation studies on the n-view threshold, and any available statistical or error-bar analyses. These additions will provide clearer evidence that the learned mapping produces uncertainties that are robust and independent of the specific MVS training points. revision: yes

Circularity Check

0 steps flagged

No significant circularity; regression trained on high-redundancy subset generalizes via independent cues

full rationale

The derivation chain begins with SfM/BA uncertainty via reprojection statistics, then addresses MVS via a regressor trained exclusively on reliable n-view points (n>=6) selected by view count. These points supply training targets derived from multi-view geometric consistency, while inputs are matching-cost and related observables. The resulting mapping is applied to the remaining points. Because the selection criterion (minimum view count) is independent of the disparity uncertainty values and the regressor learns a cue-to-uncertainty function rather than re-expressing the same measurements, no step reduces by construction to its own inputs. The method is self-contained against external benchmarks such as held-out scenes or LiDAR comparison and invokes no load-bearing self-citations or imported uniqueness theorems.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The framework relies on standard first-order error propagation for the SfM/BA stage and on the assumption that matching-cost cues are informative for disparity uncertainty; no new physical constants or invented entities are introduced.

free parameters (1)
  • regression coefficients for disparity uncertainty
    The mapping from matching-cost cues to disparity uncertainty is learned by regression on selected n-view points and therefore constitutes fitted parameters.
axioms (2)
  • standard math First-order statistics of the reprojection error function suffice for SfM uncertainty
    Invoked when stating that SfM uncertainty estimation is already well studied.
  • domain assumption Reliable n-view points (n>=6) can be identified and used as ground truth for regressing disparity uncertainty
    Central modeling choice for the self-calibrating MVS step.

pith-pipeline@v0.9.0 · 5841 in / 1468 out tokens · 34877 ms · 2026-05-19T03:49:58.877917+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

82 extracted references · 82 canonical work pages

  1. [1]

    UA V photogrammetry for mapping and 3D modeling: Current status and future perspectives,

    F. Remondino, L. Barazzetti, F. Nex, M. Scaioni, and D. Sarazzi, "UA V photogrammetry for mapping and 3D modeling: Current status and future perspectives," in Proceedings of the International Conference on Unmanned Aerial Vehicle in Geomatics (UAV-g): 14-16 September 2011, Zurich, Switzerland, 2011: International Society for Photogrammetry and Remote Sens...

    work page 2011

  2. [2]

    UA V photogrammetry for topographic monitoring of coastal areas,

    J. A. Goncalves and R. Henriques, "UA V photogrammetry for topographic monitoring of coastal areas," ISPRS journal of Photogrammetry and Remote Sensing, vol. 104, pp. 101-111, 2015

    work page 2015

  3. [3]

    A volumetric change detection framework using UA V oblique photogrammetry–a case study of ultra -high-resolution monitoring of progressive building collapse,

    N. Xu et al. , "A volumetric change detection framework using UA V oblique photogrammetry–a case study of ultra -high-resolution monitoring of progressive building collapse," International Journal of Digital Earth, vol. 14, no. 11, pp. 1705- 1720, 2021

    work page 2021

  4. [4]

    Accurate mapping method for UA V photogrammetry without ground control points in the map projection frame,

    J. Liu, W. Xu, B. Guo, G. Zhou, and H. Zhu, "Accurate mapping method for UA V photogrammetry without ground control points in the map projection frame," IEEE Transactions on Geoscience and Remote Sensing, vol. 59, no. 11, pp. 9673-9681, 2021

    work page 2021

  5. [5]

    Accuracy of unmanned aerial vehicle (UA V) and SfM photogrammetry survey as a function of the number and location of ground control points used,

    E. Sanz-Ablanedo, J. H. Chandler, J. R. Rodríguez - Pérez, and C. Ordóñez, "Accuracy of unmanned aerial vehicle (UA V) and SfM photogrammetry survey as a function of the number and location of ground control points used," Remote Sensing, vol. 10, no. 10, p. 1606, 2018

    work page 2018

  6. [6]

    Accuracy assessment of 3D photogrammetric models from an unmanned aerial vehicle,

    S. Barba, M. Barbarella, A. Di Benedetto, M. Fiani, L. Gujski, and M. Limongiello, "Accuracy assessment of 3D photogrammetric models from an unmanned aerial vehicle," Drones, vol. 3, no. 4, p. 79, 2019

    work page 2019

  7. [7]

    Flightgoggles: Photorealistic sensor simulation for perception-driven robotics using photogrammetry and virtual reality,

    W. Guerra, E. Tal, V . Murali, G. Ryou, and S. Karaman, "Flightgoggles: Photorealistic sensor simulation for perception-driven robotics using photogrammetry and virtual reality," in 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2019: IEEE, pp. 6941- 6948

    work page 2019

  8. [8]

    Three -dimensional urban flood inundation simulation based on digital aerial photogrammetry,

    Y . Rong, T. Zhang, Y . Zheng, C. Hu, L. Peng, and P. Feng, "Three -dimensional urban flood inundation simulation based on digital aerial photogrammetry," Journal of Hydrology, vol. 584, p. 124308, 2020

    work page 2020

  9. [9]

    Lidar Point Cloud—USGS National Map 3DEP Downloadable Data Collection,

    U. G. Survey, "Lidar Point Cloud—USGS National Map 3DEP Downloadable Data Collection," ed: US Geological Survey, 2017

    work page 2017

  10. [10]

    Accuracy assessment of LiDAR‐derived digital elevation models,

    F. J. Aguilar and J. P . Mills, "Accuracy assessment of LiDAR‐derived digital elevation models," The Photogrammetric Record, vol. 23, no. 122, pp. 148- 169, 2008

    work page 2008

  11. [11]

    A quantitative assessment of LiDAR data accuracy,

    A. Elaksher, T. Ali, and A. Alharthy, "A quantitative assessment of LiDAR data accuracy," Remote Sensing, vol. 15, no. 2, p. 442, 2023

    work page 2023

  12. [12]

    Digital terrain models generated with low -cost UA V photogrammetry: Methodology and accuracy,

    S. I. Jiménez-Jiménez, W. Ojeda-Bustamante, M. d. J. Marcial -Pablo, and J. Enciso, "Digital terrain models generated with low -cost UA V photogrammetry: Methodology and accuracy," ISPRS International Journal of Geo-Information, vol. 10, no. 5, p. 285, 2021

    work page 2021

  13. [13]

    A comparative analysis of UA V photogrammetric software performance for forest 3D modeling: A case study using AgiSoft photoscan, PIX4DMapper, and DJI Terra,

    S. Jarahizadeh and B. Salehi, "A comparative analysis of UA V photogrammetric software performance for forest 3D modeling: A case study using AgiSoft photoscan, PIX4DMapper, and DJI Terra," Sensors, vol. 24, no. 1, p. 286, 2024

    work page 2024

  14. [14]

    Structure - from-motion revisited,

    J. L. Schonberger and J. -M. Frahm, "Structure - from-motion revisited," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 4104-4113

    work page 2016

  15. [15]

    Global structure -from-motion revisited,

    L. Pan, D. Baráth, M. Pollefeys, and J. L. Schönberger, "Global structure -from-motion revisited," in European Conference on Computer Vision, 2024: Springer, pp. 58-77

    work page 2024

  16. [16]

    Dust3r: Geometric 3d vision made easy,

    S. Wang, V . Leroy, Y . Cabon, B. Chidlovskii, and J. Revaud, "Dust3r: Geometric 3d vision made easy," in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 20697-20709

    work page 2024

  17. [17]

    Vggt: Visual geometry grounded transformer,

    J. Wang, M. Chen, N. Karaev, A. Vedaldi, C. Rupprecht, and D. Novotny, "Vggt: Visual geometry grounded transformer," in Proceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 5294-5306

    work page 2025

  18. [18]

    Lens distortion for close-range photogrammetry,

    J. G. Fryer and D. C. Brown, "Lens distortion for close-range photogrammetry," Photogrammetric engineering and remote sensing, vol. 52, pp. 51- 58, 1986

    work page 1986

  19. [19]

    Determining an initial image pair for fixing the scale of a 3d reconstruction from an image sequence,

    C. Beder and R. Steffen, "Determining an initial image pair for fixing the scale of a 3d reconstruction from an image sequence," in Joint Pattern Recognition Symposium , 2006: Springer, pp. 657- 666

    work page 2006

  20. [20]

    E. M. Mikhail, J. S. Bethel, and J. C. McGlone, Introduction to modern photogrammetry . John Wiley & Sons, 2001

    work page 2001

  21. [21]

    Hartley and A

    R. Hartley and A. Zisserman, Multiple view geometry in computer vision . Cambridge university press, 2003

    work page 2003

  22. [22]

    Bundle adjustment — a modern synthesis,

    B. Triggs, P . F. McLauchlan, R. I. Hartley, and A. W. Fitzgibbon, "Bundle adjustment — a modern synthesis," in Vision Algorithms: Theory and Practice: International Workshop on Vision Algorithms Corfu, Greece, September 21–22, 1999 Proceedings, 2000: Springer, pp. 298-372

    work page 1999

  23. [23]

    Accurate multiple view 3d reconstruction using patch-based stereo for large -scale scenes,

    S. Shen, "Accurate multiple view 3d reconstruction using patch-based stereo for large -scale scenes," IEEE transactions on image processing, vol. 22, no. 5, pp. 1901-1914, 2013

    work page 1901

  24. [24]

    Pixelwise view selection for unstructured multi -view stereo,

    J. L. Schönberger, E. Zheng, J. -M. Frahm, and M. Pollefeys, "Pixelwise view selection for unstructured multi -view stereo," in Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part III 14, 2016: Springer, pp. 501- 518

    work page 2016

  25. [25]

    OpenMVS: Multi -View Stereo Reconstruction Library

    D. Cernea. "OpenMVS: Multi -View Stereo Reconstruction Library." https://cdcseacave.github.io/openMVS (accessed

    work page

  26. [26]

    Bundle adjustment with motion constraints for uncalibrated multi-camera systems at the ground level,

    D. Huang, R. Qin, and M. Elhashash, "Bundle adjustment with motion constraints for uncalibrated multi-camera systems at the ground level," ISPRS Journal of Photogrammetry and Remote Sensing, vol. 211, pp. 452-464, 2024

    work page 2024

  27. [27]

    M. M. Thompson, R. C. Eller, W. A. Radlinski, and J. L. Speert, Manual of photogrammetry . American Society of Photogrammetry Falls Church, V A, 1966

    work page 1966

  28. [28]

    Best linear unbiased estimation and prediction under a selection model,

    C. R. Henderson, "Best linear unbiased estimation and prediction under a selection model," Biometrics, pp. 423-447, 1975

    work page 1975

  29. [29]

    Rigorous error modeling for sUAS acquired image-derived point clouds,

    C. A. Rodarmel, M. P . Lee, K. L. Brodie, N. J. Spore, and B. Bruder, "Rigorous error modeling for sUAS acquired image-derived point clouds," IEEE Transactions on Geoscience and Remote Sensing, vol. 57, no. 8, pp. 6240-6253, 2019

    work page 2019

  30. [30]

    Error Propagation in Satellite Multi-Image Geometry,

    J. L. Mundy and H. J. Theiss, "Error Propagation in Satellite Multi-Image Geometry," IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1-14, 2021

    work page 2021

  31. [31]

    Joint estimation of depth and its uncertainty from stereo images using bayesian deep learning,

    M. Mehltretter, "Joint estimation of depth and its uncertainty from stereo images using bayesian deep learning," ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. 2, pp. 69-78, 2022

    work page 2022

  32. [32]

    Learning the distribution of errors in stereo matching for joint disparity and uncertainty estimation,

    L. Chen, W. Wang, and P. Mordohai, "Learning the distribution of errors in stereo matching for joint disparity and uncertainty estimation," in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 17235-17244

    work page 2023

  33. [33]

    A tv prior for high-quality scalable multi- view stereo reconstruction,

    A. Kuhn, H. Hirschmüller, D. Scharstein, and H. Mayer, "A tv prior for high-quality scalable multi- view stereo reconstruction," International Journal of Computer Vision, vol. 124, pp. 2-17, 2017

    work page 2017

  34. [34]

    a Critical Analysis of Internal Reliability for Uncertainty Quantification of Dense Image Matching in Multi -View Stereo,

    D. Huang and R. Qin, "a Critical Analysis of Internal Reliability for Uncertainty Quantification of Dense Image Matching in Multi -View Stereo," ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. 1, pp. 517- 524, 2023

    work page 2023

  35. [35]

    UA V photogrammetry accuracy assessment for corridor mapping based on the number and distribution of ground control points,

    E. Ferrer -González, F. Agüera -Vega, F. Carvajal - Ramírez, and P . Martínez- Carricondo, "UA V photogrammetry accuracy assessment for corridor mapping based on the number and distribution of ground control points," Remote sensing, vol. 12, no. 15, p. 2447, 2020

    work page 2020

  36. [36]

    DEM generation with UA V Photogrammetry and accuracy analysis in Sahitler hill,

    M. Uysal, A. S. Toprak, and N. Polat, "DEM generation with UA V Photogrammetry and accuracy analysis in Sahitler hill," Measurement, vol. 73, pp. 539-543, 2015

    work page 2015

  37. [37]

    Assessment of photogrammetric mapping accuracy based on variation ground control points number using unmanned aerial vehicle,

    F. Agüera -Vega, F. Carvajal -Ramírez, and P . Martínez-Carricondo, "Assessment of photogrammetric mapping accuracy based on variation ground control points number using unmanned aerial vehicle," Measurement, vol. 98, pp. 221-227, 2017

    work page 2017

  38. [38]

    Computation of scalar accuracy metrics LE, CE, and SE as both predictive and sample -based statistics,

    J. Dolloff and J. Carr, "Computation of scalar accuracy metrics LE, CE, and SE as both predictive and sample -based statistics," in Proceedings of the ASPRS 2016 Annual Conference and Co-Located JACIE Workshop-Imaging Geospatial Technol. Forum Co-Located JACIE Work, Fort Worth, TX, USA, 2016, pp. 11-15

    work page 2016

  39. [39]

    C. R. Greenwalt and M. E. Schultz, Principles of error theory and cartographic applications (no. 96). Aeronautical Chart and Information Center, 1962

    work page 1962

  40. [40]

    A multi-view stereo benchmark with high-resolution images and multi -camera videos,

    T. Schops et al. , "A multi-view stereo benchmark with high-resolution images and multi -camera videos," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 3260-3269

    work page 2017

  41. [41]

    Are we ready for autonomous driving? the kitti vision benchmark suite,

    A. Geiger, P. Lenz, and R. Urtasun, "Are we ready for autonomous driving? the kitti vision benchmark suite," in 2012 IEEE conference on computer vision and pattern recognition, 2012: IEEE, pp. 3354-3361

    work page 2012

  42. [42]

    Discrete and distributed error assessment of UAS -SfM point clouds of roadways,

    Y . Liao and R. L. Wood, "Discrete and distributed error assessment of UAS -SfM point clouds of roadways," Infrastructures, vol. 5, no. 10, p. 87, 2020

    work page 2020

  43. [43]

    3‐D uncertainty‐based topographic change detection with structure‐from‐motion photogrammetry: precision maps for ground control and directly georeferenced surveys,

    M. R. James, S. Robson, and M. W. Smith, "3‐D uncertainty‐based topographic change detection with structure‐from‐motion photogrammetry: precision maps for ground control and directly georeferenced surveys," Earth Surface Processes and Landforms, vol. 42, no. 12, pp. 1769-1788, 2017

    work page 2017

  44. [44]

    3D reconstruction quality analysis and its acceleration on GPU clusters,

    L. Polok, V . Ila, and P. Smrz, "3D reconstruction quality analysis and its acceleration on GPU clusters," in 2016 24th European Signal Processing Conference (EUSIPCO), 2016: IEEE, pp. 1108- 1112

    work page 2016

  45. [45]

    Förstner and B

    W. Förstner and B. P. Wrobel, Photogrammetric computer vision. Springer, 2016

    work page 2016

  46. [46]

    Uncertainty ellipsoids calculations for complex 3D reconstructions,

    M. Lhuillier and M. Perriollat, "Uncertainty ellipsoids calculations for complex 3D reconstructions," in Proceedings 2006 IEEE International Conference on Robotics and Automation, 2006. ICRA 2006. , 2006: IEEE, pp. 3062-3069

    work page 2006

  47. [47]

    Gauges and gauge transformations for uncertainty description of geometric structure with indeterminacy,

    K.-i. Kanatani and D. D. Morris, "Gauges and gauge transformations for uncertainty description of geometric structure with indeterminacy," IEEE Transactions on Information Theory, vol. 47, no. 5, pp. 2017-2028, 2002

    work page 2017

  48. [48]

    Uncertainty computation in large 3d reconstruction,

    M. Polic and T. Pajdla, "Uncertainty computation in large 3d reconstruction," in Image Analysis: 20th Scandinavian Conference, SCIA 2017, Tromsø, Norway, June 12–14, 2017, Proceedings, Part I 20, 2017: Springer, pp. 110-121

    work page 2017

  49. [49]

    M. Z. Nashed, Generalized Inverses and Applications: Proceedings of an Advanced Seminar Sponsored by the Mathematics Research Center, the University of Wisconsin — Madison, October 8- 10, 1973 (no. 32). Elsevier, 2014

    work page 1973

  50. [50]

    Camera uncertainty computation in large 3d reconstruction,

    M. Polic and T. Pajdla, "Camera uncertainty computation in large 3d reconstruction," in 2017 International Conference on 3D Vision (3DV), 2017: IEEE, pp. 282-290

    work page 2017

  51. [51]

    Fast and accurate camera covariance computation for large 3d reconstruction,

    M. Polic, W. Forstner, and T. Pajdla, "Fast and accurate camera covariance computation for large 3d reconstruction," in Proceedings of the European Conference on Computer Vision (ECCV) , 2018, pp. 679-694

    work page 2018

  52. [52]

    A comparison and evaluation of multi - view stereo reconstruction algorithms,

    S. M. Seitz, B. Curless, J. Diebel, D. Scharstein, and R. Szeliski, "A comparison and evaluation of multi - view stereo reconstruction algorithms," in 2006 IEEE computer society conference on computer vision and pattern recognition (CVPR'06), 2006, vol. 1: IEEE, pp. 519-528

    work page 2006

  53. [53]

    Multiray photogrammetry and dense image matching,

    N. Haala, "Multiray photogrammetry and dense image matching," in Photogrammetric week , 2011, vol. 11: VDE Verlag Heidelberg, Germany, pp. 185- 195

    work page 2011

  54. [54]

    A quantitative evaluation of confidence measures for stereo vision,

    X. Hu and P. Mordohai, "A quantitative evaluation of confidence measures for stereo vision," IEEE transactions on pattern analysis and machine intelligence, vol. 34, no. 11, pp. 2121-2133, 2012

    work page 2012

  55. [55]

    A novel confidence measure for disparity maps by pixel -wise cost function analysis,

    R. O. H. Veld, T. Jaschke, M. Bätz, L. Palmieri, and J. Keinert, "A novel confidence measure for disparity maps by pixel -wise cost function analysis," in 2018 25th IEEE International Conference on Image Processing (ICIP), 2018: IEEE, pp. 644-648

    work page 2018

  56. [56]

    A stereo confidence metric using single view imagery with comparison to five alternative approaches,

    G. Egnal, M. Mintz, and R. P . Wildes, "A stereo confidence metric using single view imagery with comparison to five alternative approaches," Image and vision computing, vol. 22, no. 12, pp. 943-957, 2004

    work page 2004

  57. [57]

    The self-aware matching measure for stereo,

    P. Mordohai, "The self-aware matching measure for stereo," in 2009 IEEE 12th International Conference on Computer Vision, 2009: IEEE, pp. 1841-1848

    work page 2009

  58. [58]

    Distinctive similarity measure for stereo matching under point ambiguity,

    K.-J. Yoon and I. S. Kweon, "Distinctive similarity measure for stereo matching under point ambiguity," Computer Vision and Image Understanding, vol. 112, no. 2, pp. 173-183, 2008

    work page 2008

  59. [59]

    A progressive scheme for stereo matching,

    Z. Zhang and Y . Shan, "A progressive scheme for stereo matching," in 3D Structure from Images — SMILE 2000: Second European Workshop on 3D Structure from Multiple Images of Large -Scale Environments Dublin, Irleand, July 1–2, 2000 Revised Papers 2, 2001: Springer, pp. 68-85

    work page 2000

  60. [60]

    On the confidence of stereo matching in a deep -learning era: a quantitative evaluation,

    M. Poggi et al. , "On the confidence of stereo matching in a deep -learning era: a quantitative evaluation," IEEE transactions on pattern analysis and machine intelligence, vol. 44, no. 9, pp. 5293- 5313, 2021

    work page 2021

  61. [61]

    Learning to predict stereo reliability enforcing local consistency of confidence maps,

    M. Poggi and S. Mattoccia, "Learning to predict stereo reliability enforcing local consistency of confidence maps," in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017, pp. 2452-2461

    work page 2017

  62. [62]

    Stereo matching confidence learning based on multi -modal convolution neural networks,

    Z. Fu, M. Ardabilian, and G. Stern, "Stereo matching confidence learning based on multi -modal convolution neural networks," in Representations, Analysis and Recognition of Shape and Motion from Imaging Data: 7th International Workshop, RFMI 2017, Savoie, France, December 17–20, 2017, Revised Selected Papers 7 , 2019: Springer, pp. 69- 81

    work page 2017

  63. [63]

    What uncertainties do we need in bayesian deep learning for computer vision?,

    A. Kendall and Y . Gal, "What uncertainties do we need in bayesian deep learning for computer vision?," Advances in neural information processing systems, vol. 30, 2017

    work page 2017

  64. [64]

    Adversarial confidence estimation networks for robust stereo matching,

    S. Kim, D. Min, S. Kim, and K. Sohn, "Adversarial confidence estimation networks for robust stereo matching," IEEE Transactions on Intelligent Transportation Systems, vol. 22, no. 11, pp. 6875- 6889, 2020

    work page 2020

  65. [65]

    Laf -net: Locally adaptive fusion networks for stereo confidence estimation,

    S. Kim, S. Kim, D. Min, and K. Sohn, "Laf -net: Locally adaptive fusion networks for stereo confidence estimation," in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 205-214

    work page 2019

  66. [66]

    Beyond local reasoning for stereo confidence estimation with deep learning,

    F. Tosi, M. Poggi, A. Benincasa, and S. Mattoccia, "Beyond local reasoning for stereo confidence estimation with deep learning," in Proceedings of the European Conference on Computer Vision (ECCV), 2018, pp. 319-334

    work page 2018

  67. [67]

    Scalable 3D surface reconstruction by local stochastic fusion of disparity maps,

    A. Kuhn, "Scalable 3D surface reconstruction by local stochastic fusion of disparity maps," Universitätsbibliothek der Universität der Bundeswehr München, 2014

    work page 2014

  68. [68]

    Uncertainty estimation for dense stereo matching using Bayesian deep learning,

    M. Mehltretter, "Uncertainty estimation for dense stereo matching using Bayesian deep learning," Fachrichtung Geodäsie und Geoinformatik der Leibniz Unviersität Hannover, 2021

    work page 2021

  69. [69]

    Coral reef monitoring by scuba divers using underwater photogrammetry and geodetic surveying,

    E. Nocerino et al., "Coral reef monitoring by scuba divers using underwater photogrammetry and geodetic surveying," Remote Sensing, vol. 12, no. 18, p. 3036, 2020

    work page 2020

  70. [70]

    NGA Standardization Document

    (2015). NGA Standardization Document. The Generic Point-cloud Model (GPM): Implementation and Exploitation (Rev. 2023-11 -15) [Online] Available: https://nsgreg.nga.mil/doc/view?i=4197

    work page 2015

  71. [71]

    Select -and-Combine (SAC): A Novel Multi -Stereo Depth Fusion Algorithm for Point Cloud Generation via Efficient Local Markov Netlets,

    M. Elhashash and R. Qin, "Select -and-Combine (SAC): A Novel Multi -Stereo Depth Fusion Algorithm for Point Cloud Generation via Efficient Local Markov Netlets," IEEE Transactions on Geoscience and Remote Sensing, vol. 61, pp. 1-9, 2023

    work page 2023

  72. [72]

    Aerial triangulation with learning-based tie points,

    F. Remondino, L. Morelli, E. Stathopoulou, M. Elhashash, and R. Qin, "Aerial triangulation with learning-based tie points," The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. 43, pp. 77-84, 2022

    work page 2022

  73. [73]

    Multi -view stereo dense matching – A software package for dense point cloud and DSM generation from multi -view stereo images

    R. Qin. "Multi -view stereo dense matching – A software package for dense point cloud and DSM generation from multi -view stereo images." https://u.osu.edu/qin.324/msp/ (accessed 2025)

    work page 2025

  74. [74]

    Non-parametric local transforms for computing visual correspondence,

    R. Zabih and J. Woodfill, "Non-parametric local transforms for computing visual correspondence," in Computer Vision— ECCV'94: Third European Conference on Computer Vision Stockholm, Sweden, May 2–6 1994 Proceedings, Volume II 3, 1994: Springer, pp. 151-158

    work page 1994

  75. [75]

    Stereo processing by semiglobal matching and mutual information,

    H. Hirschmuller, "Stereo processing by semiglobal matching and mutual information," IEEE Transactions on pattern analysis and machine intelligence, vol. 30, no. 2, pp. 328-341, 2007

    work page 2007

  76. [76]

    A median- based depthmap fusion strategy for the generation of oriented points,

    M. Rothermel, N. Haala, and D. Fritsch, "A median- based depthmap fusion strategy for the generation of oriented points," ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. 3, pp. 115-122, 2016

    work page 2016

  77. [77]

    UseGeo -A U AV-based multi-sensor dataset for geospatial research,

    F. Nex et al. , "UseGeo -A U AV-based multi-sensor dataset for geospatial research," ISPRS Open Journal of Photogrammetry and Remote Sensing, p. 100070, 2024

    work page 2024

  78. [78]

    Benchmarking the extraction of 3D geometry from UA V images with deep learning methods,

    F. Nex, N. Zhang, F. Remondino, E. Farella, R. Qin, and C. Zhang, "Benchmarking the extraction of 3D geometry from UA V images with deep learning methods," International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. 48, pp. 123-130, 2023

    work page 2023

  79. [79]

    The ISPRS benchmark on urban object classification and 3D building reconstruction,

    F. Rottensteiner et al. , "The ISPRS benchmark on urban object classification and 3D building reconstruction," ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences; I-3, vol. 1, no. 1, pp. 293-298, 2012

    work page 2012

  80. [80]

    The DGPF test on digital aerial camera evaluation-Overview and test design. Photogrammetrie, Fernerkundung, Geoinformation, V ol. 2,

    M. Cramer, "The DGPF test on digital aerial camera evaluation-Overview and test design. Photogrammetrie, Fernerkundung, Geoinformation, V ol. 2," ed, 2010

    work page 2010

Showing first 80 references.