Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
169 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Handbook on Leveraging Lines for Two-View Relative Pose Estimation (2309.16040v1)

Published 27 Sep 2023 in cs.CV

Abstract: We propose an approach for estimating the relative pose between calibrated image pairs by jointly exploiting points, lines, and their coincidences in a hybrid manner. We investigate all possible configurations where these data modalities can be used together and review the minimal solvers available in the literature. Our hybrid framework combines the advantages of all configurations, enabling robust and accurate estimation in challenging environments. In addition, we design a method for jointly estimating multiple vanishing point correspondences in two images, and a bundle adjustment that considers all relevant data modalities. Experiments on various indoor and outdoor datasets show that our approach outperforms point-based methods, improving AUC@10$\circ$ by 1-7 points while running at comparable speeds. The source code of the solvers and hybrid framework will be made public.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (87)
  1. L2d2: Learnable line detector and descriptor. In 3DV, 2021.
  2. Ceres solver. http://ceres-solver.org.
  3. Building rome in a day. Commun. ACM, 54(10), 2011.
  4. Efficient recovery of essential matrix from two affine correspondences. IEEE TIP, 27(11), 2018.
  5. Relative pose from sift features. ECCV, 2022.
  6. Graph-cut ransac. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 6733–6741, 2018.
  7. Progressive-X: Efficient, anytime, multi-model fitting algorithm. In ICCV, 2019.
  8. Magsac++, a fast, reliable and accurate robust estimator. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2020, Seattle, WA, USA, June 13-19, 2020, 2020.
  9. Efficient initial pose-graph generation for global sfm. In CVPR, 2021.
  10. Structure-from-motion using lines: Representation, triangulation, and bundle adjustment. CVIU, 100(3), 2005.
  11. Line multiview varieties, 2022.
  12. Hybrid camera pose estimation. In CVPR, 2018.
  13. Point-based multi-view stereo network. In ICCV, 2019.
  14. Locally optimized ransac. In Joint Pattern Recognition Symposium, pages 236–243, 2003.
  15. Scannet: Richly-annotated 3d reconstructions of indoor scenes. In CVPR, 2017.
  16. Toward geometric deep SLAM. CoRR, 2017.
  17. SuperPoint: Self-supervised interest point detection and description. In CVPR, 2018.
  18. Flownet: Learning optical flow with convolutional networks. In ICCV, 2015.
  19. PLMP - point-line minimal problems in complete multi-view visibility. In ICCV, 2019.
  20. Pl11{}_{\mbox{1}}start_FLOATSUBSCRIPT 1 end_FLOATSUBSCRIPTp - point-line minimal problems under partial visibility in three views. In ECCV, 2020.
  21. Line-based relative pose estimation. In The 24th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2011, Colorado Springs, CO, USA, 20-25 June 2011, pages 3049–3056. IEEE Computer Society, 2011.
  22. LSD-SLAM: large-scale direct monocular SLAM. In ECCV, 2014.
  23. TRPLP - trifocal relative pose from lines at points. In CVPR, 2020.
  24. Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography. In Readings in Computer Vision, 1987.
  25. A minimal case solution to the calibrated relative pose problem for the case of two known orientation angles. In ECCV, 2010.
  26. Multi-view stereo: A tutorial. FTCGV, 9(1-2), 2015.
  27. Towards internet-scale multi-view stereo. In CVPR, 2010.
  28. Two efficient solutions for visual odometry using directional correspondence. IEEE Transactions on Pattern Analysis & Machine Intelligence, 34(04), 2012.
  29. Pose refinement with joint optimization of visual points and lines. arXiv preprint arXiv:2110.03940, 2021.
  30. Are we ready for autonomous driving? the kitti vision benchmark suite. In Conference on Computer Vision and Pattern Recognition (CVPR), 2012.
  31. Vision meets robotics: The kitti dataset. International Journal of Robotics Research (IJRR), 2013.
  32. Privacy preserving structure-from-motion. In ECCV, 2020.
  33. Privacy preserving localization and mapping from uncalibrated cameras. In CVPR, 2021.
  34. Pl-slam: A stereo slam system through the combination of points and line segments. IEEE Transactions on Robotics, 35(3), 2019.
  35. From lines to homographies between uncalibrated images. In IX Symposium on Pattern Recognition and Image Analysis, VO4, pages 233–240, 2001.
  36. Multiple View Geometry in Computer Vision. Cambridge University Press, 2004.
  37. Reconstructing the world* in six days. In CVPR, 2015.
  38. Berthold K. P. Horn and Brian G. Schunck. Determining optical flow. Artif. Intell., 17(1-3), 1981.
  39. Learning to solve hard minimal problems. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, LA, USA, June 18-24, 2022, pages 5522–5532. IEEE, 2022.
  40. VSAC: efficient and accurate estimator for H and F. In ICCV, 2021.
  41. Image Matching across Wide Baselines: From Paper to Practice. IJCV, 2020.
  42. A new solution to the relative orientation problem using only 3 points and the vertical direction. J. Math. Imaging Vis., 39(3), 2011.
  43. Learning a multi-view stereo machine. In NeurIPS, 2017.
  44. Polynomial eigenvalue solutions to the 5-pt and 6-pt relative pose problems. In BMVC, 2008.
  45. Kenneth Levenberg. A method for the solution of certain non-linear problems in least squares. Quarterly of Applied Mathematics, 2(2), 1944.
  46. GAPS: generator for automatic polynomial solvers. CoRR, abs/2004.11765, 2020.
  47. Five-point motion estimation made easy. In ICPR, 2006.
  48. Avoiding degeneracy for monocular visual slam with point and line features. In ICRA, 2021.
  49. David G. Lowe. Distinctive image features from scale-invariant keypoints. IJCV, 60(2), 2004.
  50. Large-scale, real-time visual-inertial localization revisited. IJRR, 39(9), 2020.
  51. Scale & affine invariant interest point detectors. IJCV, 60(1), 2004.
  52. Mods: Fast and robust method for two-view matching. CVIU, 2015.
  53. Repeatability is Not Enough: Learning Affine Regions via Discriminability. In ECCV, 2018.
  54. ORB-SLAM: A versatile and accurate monocular SLAM system. IEEE Trans. Robotics, 31(5), 2015.
  55. David Nistér. An efficient solution to the five-point relative pose problem. In CVPR, 2003.
  56. Visual odometry. In CVPR, 2004.
  57. Visual odometry for ground vehicle applications. J. Field Robotics, 23(1), 2006.
  58. Meshloc: Mesh-based visual localization. In ECCV, 2022.
  59. SOLD2: Self-supervised occlusion-aware line description and detection. In CVPR, 2021.
  60. Deeplsd: Line segment detection and refinement with deep image gradients. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17327–17336, 2023a.
  61. Gluestick: Robust image matching by sticking points and lines together. Internation Conference on Computer Vision, 2023b.
  62. Pl-slam: Real-time monocular visual slam with points and lines. In ICRA, 2017.
  63. USAC: A universal framework for random sample consensus. IEEE TPAMI, 35(8), 2013.
  64. R2D2: reliable and repeatable detector and descriptor. In NeurIPS, 2019.
  65. Robust and accurate line- and/or point-based pose estimation without manhattan assumptions. In ECCV 2016, 2016.
  66. SuperGlue: Learning feature matching with graph neural networks. In CVPR, 2020.
  67. LaMAR: Benchmarking Localization and Mapping for Augmented Reality. In ECCV, 2022.
  68. Improving image-based localization by active correspondence search. In ECCV, 2012.
  69. Benchmarking 6dof outdoor visual localization in changing conditions. In CVPR, 2018.
  70. Homography based visual odometry with known vertical direction and weak manhattan world assumption, 2012.
  71. Structure-from-motion revisited. In CVPR, 2016.
  72. A multi-view stereo benchmark with high-resolution images and multi-camera videos. In CVPR, 2017.
  73. Scene coordinate regression forests for camera relocalization in rgb-d images. In CVPR, 2013.
  74. Structure plp-slam: Efficient sparse mapping and localization using point, line and plane for monocular, rgb-d and stereo cameras. arXiv, 2022.
  75. Photo tourism: Exploring photo collections in 3d. In ACM SIGGRAPH Conference, 2006.
  76. Modeling the world from internet photo collections. IJCV, 80(2), 2008.
  77. Recent developments on direct relative orientation. ISPRS Journal of Photogrammetry and Remote Sensing, 60(4):284–294, 2006.
  78. Loftr: Detector-free local feature matching with transformers. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8922–8931, 2021.
  79. Solving for relative pose with a partially known rotation is a quadratic eigenvalue problem. In 2nd International Conference on 3D Vision, 3DV 2014, Tokyo, Japan, December 8-11, 2014, Volume 1, pages 483–490. IEEE Computer Society, 2014.
  80. Jean-Philippe Tardif. Non-iterative approach for fast and accurate vanishing point detection. In ICCV, 2009.
  81. DISK: learning local features with policy gradient. In NeurIPS, 2020.
  82. Real-time monocular visual slam by combining points and lines. In ICME, 2019.
  83. Line as a Visual Sentence: Context-aware line descriptor for visual localization. IEEE Robotics and Automation Letters, 2021.
  84. ASIFT: an algorithm for fully affine invariant comparison. Image Process. Line, 1, 2011.
  85. An efficient and robust line segment matching approach based on lbd descriptor and pairwise geometric consistency. Journal of Visual Communication and Image Representation, 24, 2013.
  86. Very large-scale global sfm by distributed motion averaging. In CVPR, 2018.
  87. Robust visual slam with point and line features. In IROS, 2017.
Citations (2)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now