pith. machine review for the scientific record. sign in

arxiv: 2604.09199 · v1 · submitted 2026-04-10 · 💻 cs.CV

Recognition: unknown

Globally Optimal Pose from Orthographic Silhouettes

Agniva Sengupta , Dilara Ku\c{s} , Jianning Li , Stefan Zachow

Authors on Pith no claims yet

Pith reviewed 2026-05-10 16:54 UTC · model grok-4.3

classification 💻 cs.CV
keywords pose estimationsilhouetteorthographic projectionglobal optimizationresponse surfacecomputer vision3D pose
0
0 comments X

The pith

Precomputed silhouette-area response surfaces allow globally optimal pose recovery from orthographic silhouettes alone.

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

This paper demonstrates that the area of an object's silhouette changes continuously as it rotates, allowing the creation of a precomputed response surface that maps areas to rotations. By searching this surface at increasing resolutions and using fitted ellipse aspect ratios as additional signatures, the method efficiently identifies the best matching pose. It does so without relying on any point or feature correspondences between the 3D model and the 2D image. The approach applies to arbitrary shapes, including concave ones and those with holes. Validation on synthetic and real data shows better accuracy than existing silhouette-based methods.

Core claim

The paper claims that the continuity of silhouette area with rotation trajectories enables a branching search on a precomputed response surface, combined with auxiliary ellipse aspect ratio signatures, to recover the globally optimal pose from unoccluded orthographic silhouettes for any shape irrespective of convexity and genus.

What carries the argument

A pre-computed response surface of silhouette areas over rotation space, which supports resolution-guided candidate search by providing strong branching to narrow down possible poses.

If this is right

  • Accurate pose estimation becomes possible using only binary silhouette images for complex 3D models.
  • The method avoids failures common in correspondence-based approaches for non-convex or topologically complex shapes.
  • Global optimality is achieved through systematic search rather than local optimization or heuristic matching.
  • Computation remains practical because the area continuity allows coarse-to-fine refinement without exhaustive enumeration.

Where Pith is reading between the lines

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

  • The response surface idea might apply to other continuous projection properties, such as silhouette perimeter, for hybrid signatures.
  • In practice, this could support real-time applications in augmented reality if surfaces are precomputed for multiple objects.
  • The orthographic assumption suggests a potential extension by incorporating camera calibration to handle mild perspective effects.

Load-bearing premise

Silhouette areas vary continuously across rotation trajectories and inputs are perfect unoccluded orthographic projections.

What would settle it

Finding a shape and rotation where the measured silhouette area does not correspond to the global peak on the response surface, causing the guided search to converge to an incorrect pose.

Figures

Figures reproduced from arXiv: 2604.09199 by Agniva Sengupta, Dilara Ku\c{s}, Jianning Li, Stefan Zachow.

Figure 1
Figure 1. Figure 1: When a L-c trajectory from A to B on an unit sphere (a) is mapped to SOp3q and applied to two arbitrary shapes (b,c), the resulting evolution of orthographic AoS, as shown in (d,e) with its abscissa mapped as rA, Bs ÞÑ r0, 1s, is L-c 3.3.1. Area of silhouettes as shape signature We use the area of a region enclosed by the silhouette as a global geometric feature, to subdivide the search space. We introduce… view at source ↗
Figure 2
Figure 2. Figure 2: (a) Shows the Postel ball (Sπ) and disc (Dπ) inside a cube of length π with axes pXπ, Yπ, Zπq, (b) shows the 3D template of PD, (c) shows the mapping of Dπ to AoS for PD, and (d) shows the mapping of Dπ to elliptical aspect ratio for PD algorithm-1 of supplementary. An example of PARS for the triangulated 3D model of Phlegmatic Dragon (PD) [11] is shown in [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Example of pose estimation: (a) shows input silhouette [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Success rate of pose estimation across the seven best so [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: RMSE (top-row) and runtime (bottom-row) of GlOpti￾PoS under varying (a) ϵz, (b) ϵX, and (c) P. Blue curves denote interpolated mean-curves (please zoom-in). 7 [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: I) The six shapes A through F using spherical harmonics, and II) |C˜| plotted against shapes A through F, the red curve passes through mean values, the shaded area gives the range 5. Discussion Despite our method’s accuracy, strong occlusion or heavy noise can still induce failures. Such settings are fundamen￾tally unsolvable, no existing silhouette-only approach is re￾liable under such data corruption. Ou… view at source ↗
Figure 7
Figure 7. Figure 7: Qualitative results on the asymmetric shapes from [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
read the original abstract

We solve the problem of determining the pose of known shapes in $\mathbb{R}^3$ from their unoccluded silhouettes. The pose is determined up to global optimality using a simple yet under-explored property of the area-of-silhouette: its continuity w.r.t trajectories in the rotation space. The proposed method utilises pre-computed silhouette-signatures, modelled as a response surface of the area-of-silhouettes. Querying this silhouette-signature response surface for pose estimation leads to a strong branching of the rotation search space, making resolution-guided candidate search feasible. Additionally, we utilise the aspect ratio of 2D ellipses fitted to projected silhouettes as an auxiliary global shape signature to accelerate the pose search. This combined strategy forms the first method to efficiently estimate globally optimal pose from just the silhouettes, without being guided by correspondences, for any shape, irrespective of its convexity and genus. We validate our method on synthetic and real examples, demonstrating significantly improved accuracy against comparable approaches. Code, data, and supplementary in: https://agnivsen.github.io/pose-from-silhouette/

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 paper claims to solve for the globally optimal 3D pose of a known shape from a single unoccluded orthographic silhouette by exploiting the continuity of the projected area function over SO(3). It precomputes a silhouette-area response surface (termed a 'silhouette-signature'), performs resolution-guided branching search over this surface, and augments the search with the aspect ratio of an ellipse fitted to the silhouette. The method is asserted to be the first correspondence-free approach that works for arbitrary shapes irrespective of convexity or genus, with validation on synthetic and real data showing improved accuracy over prior methods.

Significance. If the global-optimality guarantee can be established, the work would be significant: it offers a practical, correspondence-free pipeline that leverages a simple geometric invariant (silhouette area continuity) together with precomputed signatures and auxiliary ellipse features. This could broaden applicability in vision tasks where feature matching is unreliable. The precomputation-plus-branching strategy is a concrete algorithmic contribution that merits attention if the discretization error is provably controlled.

major comments (2)
  1. [Abstract and §3 (method)] The central claim of global optimality (abstract and §3) rests on resolution-guided search over the precomputed area response surface. However, no Lipschitz constant, modulus of continuity, or branch-and-bound pruning argument is supplied to bound the discretization error; for non-convex or high-genus shapes the area map can possess multiple local maxima and steep gradients, so the discrete search may return a local rather than global solution. This is load-bearing for the 'globally optimal' assertion.
  2. [§4] §4 (validation) reports improved accuracy on synthetic and real examples, yet the manuscript provides neither quantitative error tables (e.g., mean angular error, success rate at 5°/10° thresholds) nor an ablation isolating the contribution of the area surface versus the ellipse auxiliary. Without these metrics it is impossible to verify whether the results actually support the global-optimality claim over local baselines.
minor comments (2)
  1. [§2–3] Notation for the rotation parameterization and the precise definition of the 'silhouette-signature' response surface should be introduced with an equation in §2 or §3 rather than left implicit.
  2. [§3] The supplementary material link is given, but the main text should state the exact resolution used for the precomputed surface and the branching criterion (e.g., area threshold or gradient magnitude) so that the method is reproducible from the paper alone.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We address each major comment below, clarifying the method's reliance on continuity and empirical validation while indicating planned revisions to strengthen the presentation.

read point-by-point responses

  1. Referee: [Abstract and §3 (method)] The central claim of global optimality (abstract and §3) rests on resolution-guided search over the precomputed area response surface. However, no Lipschitz constant, modulus of continuity, or branch-and-bound pruning argument is supplied to bound the discretization error; for non-convex or high-genus shapes the area map can possess multiple local maxima and steep gradients, so the discrete search may return a local rather than global solution. This is load-bearing for the 'globally optimal' assertion.

    Authors: We acknowledge that the manuscript does not supply a formal Lipschitz constant, modulus of continuity, or explicit branch-and-bound analysis to provably bound discretization error. The approach instead exploits the continuity of the projected-area function over SO(3) to enable a resolution-guided branching search over the precomputed silhouette-signature surface; the auxiliary ellipse aspect ratio is used to further prune candidates. While this strategy does not constitute a rigorous guarantee against local maxima for every non-convex or high-genus shape, the exhaustive nature of the precomputed surface at successively finer resolutions, combined with the auxiliary feature, is intended to locate the global solution in practice. We will revise §3 to expand the discussion of how continuity induces branching and to note the empirical character of the global-optimality claim for complex shapes. revision: partial

  2. Referee: [§4] §4 (validation) reports improved accuracy on synthetic and real examples, yet the manuscript provides neither quantitative error tables (e.g., mean angular error, success rate at 5°/10° thresholds) nor an ablation isolating the contribution of the area surface versus the ellipse auxiliary. Without these metrics it is impossible to verify whether the results actually support the global-optimality claim over local baselines.

    Authors: We agree that the validation section would benefit from more rigorous quantitative reporting. The current experiments demonstrate improved accuracy over prior methods on both synthetic and real data, but we will augment §4 with tables of mean angular error, success rates at 5° and 10° thresholds, and an ablation study that isolates the silhouette-signature response surface from the ellipse aspect-ratio auxiliary. These additions will make the empirical support for the method's performance and global-optimality behavior explicit and comparable to local baselines. revision: yes

Circularity Check

0 steps flagged

No circularity: algorithmic search on precomputed response surface is self-contained

full rationale

The paper presents a computational method that precomputes a response surface of silhouette areas over rotation space, then performs resolution-guided search exploiting the known continuity of projected area under orthographic projection. No equation or claim reduces the estimated pose to a fitted parameter defined from the target data, nor does any load-bearing step rely on a self-citation that itself assumes the result. The global-optimality assertion follows from the search procedure on the discrete surface rather than from a definitional identity or ansatz smuggled via prior work. Validation on synthetic and real examples supplies an external check independent of the derivation. This is a standard non-circular algorithmic contribution in computational geometry.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the continuity of silhouette area with rotation trajectories and the utility of precomputed response surfaces plus ellipse aspect ratios. No free parameters, additional axioms, or invented entities are explicitly quantified in the abstract.

axioms (1)
  • domain assumption The area of the silhouette is continuous with respect to trajectories in the rotation space.
    This property is invoked to justify the response-surface branching and resolution-guided search.

pith-pipeline@v0.9.0 · 5493 in / 1124 out tokens · 36099 ms · 2026-05-10T16:54:38.748742+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

42 extracted references · 2 canonical work pages · 1 internal anchor

  1. [1]

    Dronepose: photo- realistic uav-assistant dataset synthesis for 3d pose estima- tion via a smooth silhouette loss

    Georgios Albanis, Nikolaos Zioulis, Anastasios Dimou, Dimitrios Zarpalas, and Petros Daras. Dronepose: photo- realistic uav-assistant dataset synthesis for 3d pose estima- tion via a smooth silhouette loss. InComputer Vision–ECCV 2020 Workshops: Glasgow, UK, August 23–28, 2020, Pro- ceedings, Part II 16, pages 663–681. Springer, 2020. 1

    2020

  2. [2]

    Camera pose in sft and nrsfm under isometric and weaker deformation models

    Adrien Bartoli and Agniva Sengupta. Camera pose in sft and nrsfm under isometric and weaker deformation models. Computer Vision and Image Understanding, page 104488,

  3. [3]

    Least squares fitting of ellipsoid using orthogonal distances.Boletim de ciências geodésicas, 21(2): 329–339, 2015

    Sebahattin Bektas. Least squares fitting of ellipsoid using orthogonal distances.Boletim de ciências geodésicas, 21(2): 329–339, 2015. 5

    2015

  4. [4]

    Depth Pro: Sharp Monocular Metric Depth in Less Than a Second

    Aleksei Bochkovskii, Amaël Delaunoy, Hugo Germain, Marcel Santos, Yichao Zhou, Stephan R. Richter, and Vladlen Koltun. Depth pro: Sharp monocular metric depth in less than a second. InInternational Conference on Learn- ing Representations, 2025. URLhttps://arxiv.org/ abs/2410.02073. 5

    work page internal anchor Pith review arXiv 2025

  5. [5]

    Boumal, B

    N. Boumal, B. Mishra, P.-A. Absil, and R. Sepulchre. Manopt, a Matlab toolbox for optimization on manifolds. Journal of Machine Learning Research, 15(42):1455–1459,

  6. [6]

    URLhttps://www.manopt.org. 5

  7. [7]

    Robust 6dof pose tracking consid- ering contour and interior correspondence uncertainty for ar assembly guidance.arXiv preprint arXiv:2502.11971, 2025

    Jixiang Chen, Jing Chen, Kai Liu, Haochen Chang, Shan- feng Fu, and Jian Yang. Robust 6dof pose tracking consid- ering contour and interior correspondence uncertainty for ar assembly guidance.arXiv preprint arXiv:2502.11971, 2025. 1, 2

    work page arXiv 2025

  8. [8]

    Augmented reality guided laparoscopic surgery of the uterus.IEEE Transactions on Medical Imag- ing, 40(1):371–380, 2020

    Toby Collins, Daniel Pizarro, Simone Gasparini, Nicolas Bourdel, Pauline Chauvet, Michel Canis, Lilian Calvet, and Adrien Bartoli. Augmented reality guided laparoscopic surgery of the uterus.IEEE Transactions on Medical Imag- ing, 40(1):371–380, 2020. 1

    2020

  9. [9]

    Efficient and flexible sampling with blue noise properties of triangular meshes.IEEE transactions on visualization and computer graphics, 18(6):914–924, 2012

    Massimiliano Corsini, Paolo Cignoni, and Roberto Scopigno. Efficient and flexible sampling with blue noise properties of triangular meshes.IEEE transactions on visualization and computer graphics, 18(6):914–924, 2012. 2

    2012

  10. [10]

    Silhouette-based 6d object pose estimation

    Xiao Cui, Nan Li, Chi Zhang, Qian Zhang, Wei Feng, and Liang Wan. Silhouette-based 6d object pose estimation. In International Conference on Computational Visual Media, pages 157–179. Springer, 2024. 2, 5, 6

    2024

  11. [11]

    On the shape of a set of points in the plane.IEEE Trans- actions on information theory, 29(4):551–559, 1983

    Herbert Edelsbrunner, David Kirkpatrick, and Raimund Sei- del. On the shape of a set of points in the plane.IEEE Trans- actions on information theory, 29(4):551–559, 1983. 2, 3

    1983

  12. [12]

    3d model of a dragon released during eurographics 2007, 2007

    Ji ˇrí Filip, Radek Holub, Vlastimil Havran, Jaroslav Kˇrivánek, and Daniel Sýkora. 3d model of a dragon released during eurographics 2007, 2007. URLwww.dcgi.fel. cvut.cz/eg07/index.php?page=dragon. 4

    2007

  13. [13]

    Vincent Gaudillière, Gilles Simon, and Marie-Odile Berger. Perspective-1-ellipsoid: Formulation, analysis and solutions of the camera pose estimation problem from one ellipse- ellipsoid correspondence.International Journal of Computer Vision, 131(9):2446–2470, 2023. 2, 8

    2023

  14. [14]

    Courier Corporation, 2006

    Robert Gilmore.Lie groups, Lie algebras, and some of their applications. Courier Corporation, 2006. 6

    2006

  15. [15]

    Relative pose from cylinder silhouettes

    Anna Gummeson and Magnus Oskarsson. Relative pose from cylinder silhouettes. InProceedings of the Asian Con- ference on Computer Vision, pages 2545–2561, 2024. 2

    2024

  16. [16]

    Pose initialization of uncooperative spacecraft by template match- ing with sparse point cloud.Journal of Guidance, Control, and Dynamics, 44(9):1707–1720, 2021

    Wulong Guo, Weiduo Hu, Chang Liu, and Tingting Lu. Pose initialization of uncooperative spacecraft by template match- ing with sparse point cloud.Journal of Guidance, Control, and Dynamics, 44(9):1707–1720, 2021. 1

    2021

  17. [17]

    Global optimization through rotation space search.International Journal of Com- puter Vision, 82(1):64–79, 2009

    Richard I Hartley and Fredrik Kahl. Global optimization through rotation space search.International Journal of Com- puter Vision, 82(1):64–79, 2009. 1, 3

    2009

  18. [18]

    Com- bined shape, appearance and silhouette for simultaneous ma- nipulator and object tracking

    Paul Hebert, Nicolas Hudson, Jeremy Ma, Thomas Howard, Thomas Fuchs, Max Bajracharya, and Joel Burdick. Com- bined shape, appearance and silhouette for simultaneous ma- nipulator and object tracking. In2012 IEEE International Conference on Robotics and Automation, pages 2405–2412. IEEE, 2012. 1

    2012

  19. [19]

    Closed-form solution of absolute orientation us- ing orthonormal matrices.Journal of the Optical Society of America A, 5(7):1127–1135, 1988

    Berthold KP Horn, Hugh M Hilden, and Shahriar Negah- daripour. Closed-form solution of absolute orientation us- ing orthonormal matrices.Journal of the Optical Society of America A, 5(7):1127–1135, 1988. 6

    1988

  20. [20]

    Silhouette lookup for automatic pose tracking

    Nicholas R Howe. Silhouette lookup for automatic pose tracking. In2004 Conference on Computer Vision and Pat- tern Recognition Workshop, pages 15–22. IEEE, 2004. 2

    2004

  21. [21]

    Silhouette lookup for monocular 3d pose tracking.Image and Vision Computing, 25(3):331–341,

    Nicholas R Howe. Silhouette lookup for monocular 3d pose tracking.Image and Vision Computing, 25(3):331–341,

  22. [22]

    Using lo- cally corresponding cad models for dense 3d reconstructions from a single image

    Chen Kong, Chen-Hsuan Lin, and Simon Lucey. Using lo- cally corresponding cad models for dense 3d reconstructions from a single image. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 4857– 4865, 2017. 2

    2017

  23. [23]

    On pencils of tangent planes and the recognition of smooth 3d shapes from silhouettes

    Svetlana Lazebnik, Amit Sethi, Cordelia Schmid, David Kriegman, Jean Ponce, and Martial Hebert. On pencils of tangent planes and the recognition of smooth 3d shapes from silhouettes. InEuropean Conference on Computer Vision, pages 651–665. Springer, 2002. 2

    2002

  24. [24]

    Bcot: A markerless high-precision 3d object tracking benchmark

    Jiachen Li, Bin Wang, Shiqiang Zhu, Xin Cao, Fan Zhong, Wenxuan Chen, Te Li, Jason Gu, and Xueying Qin. Bcot: A markerless high-precision 3d object tracking benchmark. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 6697–6706, 2022. 6

    2022

  25. [25]

    Model-based shape from silhouette: A solution involving a small number of views

    JF Menudet, JM Becker, T Fournel, and C Mennessier. Model-based shape from silhouette: A solution involving a small number of views. InProceedings of the Second Inter- national Conference on Computer Vision Theory and Appli- cations, pages 379–386, 2007. 2 9

    2007

  26. [26]

    The levenberg-marquardt algorithm: imple- mentation and theory

    Jorge J Moré. The levenberg-marquardt algorithm: imple- mentation and theory. InNumerical analysis: proceedings of the biennial Conference held at Dundee, June 28–July 1, 1977, pages 105–116. Springer, 2006. 6

    1977

  27. [27]

    Optimal coherent point selection for 3d quality inspection from silhouette-based reconstructions.Mathematics, 11(21):4419, 2023

    Javier Pérez Soler, Jose-Luis Guardiola, Alberto Perez Jimenez, Pau Garrigues Carbó, Nicolás García Sastre, and Juan-Carlos Perez-Cortes. Optimal coherent point selection for 3d quality inspection from silhouette-based reconstructions.Mathematics, 11(21):4419, 2023. 1

    2023

  28. [28]

    Fast and controllable 3d modelling from silhouettes

    Mukta Prasad, Andrew W Fitzgibbon, and Andrew Zisser- man. Fast and controllable 3d modelling from silhouettes. In Eurographics (Short Presentations), pages 9–12, 2005. 2

    2005

  29. [29]

    Convex solutions to sft and nrsfm under algebraic deformation models.IEEE Transactions on Pattern Analysis and Machine Intelligence,

    Agniva Sengupta and Adrien Bartoli. Convex solutions to sft and nrsfm under algebraic deformation models.IEEE Transactions on Pattern Analysis and Machine Intelligence,

  30. [30]

    Shape-from-template with generalised camera.Image and Vision Computing, page 105579, 2025

    Agniva Sengupta and Stefan Zachow. Shape-from-template with generalised camera.Image and Vision Computing, page 105579, 2025. 1, 2

    2025

  31. [31]

    The Mathworks.MATLAB Function Reference

    Inc. The Mathworks.MATLAB Function Reference. The Mathworks, Inc., 2024. 4

    2024

  32. [32]

    Silhouette-based variational methods for single view reconstruction

    Eno Töppe, Martin R Oswald, Daniel Cremers, and Carsten Rother. Silhouette-based variational methods for single view reconstruction. InVideo Processing and Computa- tional Video: International Seminar, Dagstuhl Castle, Ger- many, October 10-15, 2010. Revised Papers, pages 104–123. Springer, 2011. 2

    2010

  33. [33]

    Scatter search and local nlp solvers: A multistart framework for global optimization.IN- FORMS Journal on computing, 19(3):328–340, 2007

    Zsolt Ugray, Leon Lasdon, John Plummer, Fred Glover, James Kelly, and Rafael Martí. Scatter search and local nlp solvers: A multistart framework for global optimization.IN- FORMS Journal on computing, 19(3):328–340, 2007. 6

    2007

  34. [34]

    Invariant- based recognition of complex curved 3d objects from image contours.Computer Vision and Image Understanding, 72(3): 287–303, 1998

    B Vijayakumar, David Kriegman, and Jean Ponce. Invariant- based recognition of complex curved 3d objects from image contours.Computer Vision and Image Understanding, 72(3): 287–303, 1998. 2

    1998

  35. [35]

    Deep active contours for real-time 6-dof object tracking

    Long Wang, Shen Yan, Jianan Zhen, Yu Liu, Maojun Zhang, Guofeng Zhang, and Xiaowei Zhou. Deep active contours for real-time 6-dof object tracking. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 14034–14044, 2023. 2, 8

    2023

  36. [36]

    Directshape: Direct photometric alignment of shape pri- ors for visual vehicle pose and shape estimation

    Rui Wang, Nan Yang, Joerg Stueckler, and Daniel Cremers. Directshape: Direct photometric alignment of shape pri- ors for visual vehicle pose and shape estimation. In2020 IEEE International Conference on Robotics and Automation (ICRA), pages 11067–11073. IEEE, 2020. 1

    2020

  37. [37]

    Azimuthal equidistant projection, Dec

    Wikipedia. Azimuthal equidistant projection, Dec

  38. [38]

    [Online; accessed 10-April-2026]

    URLhttps://en.wikipedia.org/wiki/ Azimuthal_equidistant_projection. [Online; accessed 10-April-2026]. 3

    2026

  39. [39]

    Spherical harmonics.http : / / en

    Wikipedia. Spherical harmonics.http : / / en . wikipedia . org / w / index . php ? title = Spherical % 20harmonics & oldid = 1321561306,

  40. [40]

    [Online; accessed 13-November-2025]. 7

    2025

  41. [41]

    Using silhou- ette for pose estimation of object with surface of revolution

    Ming Zhang, Yinqiang Zheng, and Yuncai Liu. Using silhou- ette for pose estimation of object with surface of revolution. In2009 16th IEEE International Conference on Image Pro- cessing (ICIP), pages 333–336. IEEE, 2009. 2

    2009

  42. [42]

    Mc-lrf based pose measurement system for shipborne aircraft automatic landing.Chinese Journal of Aeronautics, 36(8):298–312, 2023

    Zhuo Zhang, W ANG Qiufu, BI Daoming, SUN Xiaoliang, and YU Qifeng. Mc-lrf based pose measurement system for shipborne aircraft automatic landing.Chinese Journal of Aeronautics, 36(8):298–312, 2023. 1 10

    2023