含有平面结构场景的捆绑调整

谢远帆; 吴毅红; 范力欣

doi:10.3724/SP.J.1004.2014.01601

含有平面结构场景的捆绑调整

doi: 10.3724/SP.J.1004.2014.01601

1.
中国科学院自动化研究所模式识别国家重点实验室北京 100190, 中国;
2.
诺基亚研究中心坦佩雷 33720, 芬兰

基金项目:

国家重点基础研究发展计划（973计划）（2012CB316302），国家自然科学基金（61070107）资助

详细信息

作者简介:
吴毅红中国科学院自动化研究所研究员. 2001 年获中国科学院系统所理学博士学位. 主要研究方向为摄像机标定，摄像机定位，三维重建.E-mail：yhwu@nlpr.ia.ac.cn

通讯作者:
谢远帆中国科学院自动化研究所博士研究生. 2007 年获中南大学信息科学与工程学院自动化专业学士学位. 主要研究方向为基于视觉的同步定位与地图创建.E-mail：yfxie@nlpr.ia.ac.cn

计量
- 文章访问数: 2521
- HTML全文浏览量: 73
- PDF下载量: 1513
- 被引次数: 0
出版历程
- 收稿日期: 2012-12-13
- 修回日期: 2013-08-13
- 刊出日期: 2014-08-20

Bundle Adjustment for Scenes Containing Planes

1.
National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China;
2.
Nokia Research Center, Tampere 33720, Finland

Funds:

Supported by National Basic Research Program of China (973 Program) (2012CB316302), and National Natural Science Foundation of China (61070107)

摘要

摘要: 捆绑调整是计算机视觉中三维结构恢复过程的重要步骤. 捆绑调整通常将空间中点与点坐标的调整视为相互独立的过程，但是在包含有自然物和人工物的场景中，由于存在多余的自由度，这种调整方法会导致优化结果偏离真值. 提出了一种带有共面约束和平面夹角约束的捆绑调整，用于优化带有平面的场景. 借助新的参数化方法，共面约束和夹角约束可以方便地进行表示，并且带有这两类约束的捆绑调整求解过程，仍然是一个无约束的非线性最小二乘问题. 实验结果表明，这种带有先验信息的捆绑调整提供了对结构的更准确估计. 由于先验信息的加入，增强型法方程的维度变高，借助了稀疏的求解技术和预条件子方法，大大降低了求解时间. 最后，为了在实际应用中能够自动寻找出夹角约束，提出了一种基于最大完全图的贪心方法，该方法尽可能多地保留了夹角约束.
- 捆绑调整 /
- 三维重建 /
- L-M（Levenberg-Marquardt）方法 /
- 稀疏方程求解
Abstract: Bundle adjustment has been considered as one of the most important components in computer vision systems where three-dimensional structures are needed. A general bundle adjustment can optimize the coordinates of space points independently. But for a scene composed of both natural and structured objects, this often leads to an over parametrization the result and which deviates from truth. In this paper, a bundle adjustment with planar constraints and angle constraints is proposed for recovering the structures of environments with planes. By the aid of a new parametrization, the optimization remains an unconstrained non-linear least squares problem even if these two kinds of constraints are added. Experiments show that this new bundle adjustment method with prior knowledge provides an accurate estimation of the structures. Since prior information is added, the dimensionality of the augmented normal equation increases. A sparse solver is used after preconditioning in order to alleviate this problem. Moreover, a graph based angle constraint inference is devised for automatically finding constraints in a greedy manner once all planes are identified. This greedy method can preserve as many angle constraints as possible.
- Bundle adjustment /
- 3D reconstruction /
- Levenberg-Marquardt (L-M) method /
- sparse solver

HTML全文

参考文献(18)

[1]	Lhuillier M. Fusion of GPS and structure-from-motion using constrained bundle adjustments. In: Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Colorado Springs, USA: IEEE, 2011. 3025-3032
[2]	[2] Wong K H, Chang M M Y. 3D model reconstruction by constrained bundle adjustment. In: Proceedings of the 17th International Conference on Pattern Recognition. Cambridge, UK: IEEE, 2004, 3: 902-905
[3]	[3] Di K, Xu F, Li R. Constrained bundle adjustment of panoramic stereo images for Mars landing site mapping. In: Proceedings of the 4th International Symposium on Mobile Mapping Technology. Kunming, China: MMT, 2004. 29-31
[4]	[4] Brlin N, Grussenmeyer P, Eriksson J, Lindstrom P. Pros and cons of constrained and unconstrained formulation of the bundle adjustment problem. In: International Archives of ISPRS, 2004, XXXV(B3): 589-594
[5]	[5] Zhou Z H, Jin H L, Ma Y. Robust plane-based structure from motion. In: Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Rhode Island, USA: IEEE, 2012. 1482-1489
[6]	[6] Shan Y, Liu Z C, Zhang Z Y. Model-based bundle adjustment with application to face modeling. In: Proceedings of the 8th IEEE International Conference on Computer Vision. Vancouver, Canada: IEEE, 2001, 2: 644-651
[7]	[7] Fua P. Regularized bundle-adjustment to model heads from image sequences without calibration data. International Journal of Computer Vision, 2000, 38(2): 153-171
[8]	[8] Szeliski R, Torr P H S. Geometrically constrained structure from motion: points on planes. Lecture Notes in Computer Science, 1998, 1506: 171-186
[9]	[9] Bartoli A, Sturm P. Constrained structure and motion from multiple uncalibrated views of a piecewise planar scene. International Journal of Computer Vision, 2003, 52(1): 45-64
[10]	Gerke M. Using horizontal and vertical building structure to constrain indirect sensor orientation. ISPRS Journal of Photogrammetry and Remote Sensing, 2011, 66(3): 307-316
[11]	McGlone J C. Bundle adjustment with geometric constraints for hypothesis evaluation. sl ISPRS Journal of Photogrammetry and Remote Sensing, 1996. B3-III529-534
[12]	Hartley R I, Zisserman A. Multiple View Geometry in Computer Vision. Cambridge: Cambridge University Press, 2004
[13]	Triggs B, McLauchlan P F, Hartley R I, Fitzgibbon A W. Bundle adjustment-a modern synthesis. In: Proceedings of the International Workshop on Vision Algorithms: Theory and Practice. London, UK: IEEE, 2000. 298-372
[14]	Wedderburn J H M. Lectures on Matrices. Providence: American Mathematical Society, 1934
[15]	Timothy A D. Direct Methods for Sparse Linear Systems. Philadelphia: Society for Industrial and Applied Mathematics, 2006
[16]	Rotkin V, Toledo S. The design and implementation of a new out-of-core sparse cholesky factorization method. ACM Transactions on Mathematical Software, 2004, 30(1): 19-46
[17]	Toldo R, Fusiello A. Robust multiple structures estimation with J-linkage. In: Proceedings of the 10th European Conference on Computer Vision: Part I. Marseille. France: Springer-Verlag, 2008. 537-547
[18]	Torr P H S, Zisserman A. MLESAC: a new robust estimator with application to estimating image geometry. Computer Vision and Image Understanding, 2000, 78(1): 138-156

施引文献

资源附件(0)

访问统计

计量

文章访问数: 2521
HTML全文浏览量: 73
PDF下载量: 1513
被引次数: 0

姓名
邮箱
手机号码
标题
留言内容
验证码

留言板

含有平面结构场景的捆绑调整

doi: 10.3724/SP.J.1004.2014.01601

作者简介:
吴毅红中国科学院自动化研究所研究员. 2001 年获中国科学院系统所理学博士学位. 主要研究方向为摄像机标定，摄像机定位，三维重建.E-mail：yhwu@nlpr.ia.ac.cn

通讯作者:
谢远帆中国科学院自动化研究所博士研究生. 2007 年获中南大学信息科学与工程学院自动化专业学士学位. 主要研究方向为基于视觉的同步定位与地图创建.E-mail：yfxie@nlpr.ia.ac.cn

计量

Bundle Adjustment for Scenes Containing Planes

计量

目录

留言板

含有平面结构场景的捆绑调整

doi: 10.3724/SP.J.1004.2014.01601

作者简介: 吴毅红 中国科学院自动化研究所研究员. 2001 年获中国科学院系统所理学博士学位. 主要研究方向为摄像机标定，摄像机定位，三维重建.E-mail：yhwu@nlpr.ia.ac.cn

通讯作者: 谢远帆 中国科学院自动化研究所博士研究生. 2007 年获中南大学信息科学与工程学院自动化专业学士学位. 主要研究方向为基于视觉的同步定位与地图创建.E-mail：yfxie@nlpr.ia.ac.cn

计量

出版历程

Bundle Adjustment for Scenes Containing Planes

计量

出版历程

目录

作者简介:
吴毅红中国科学院自动化研究所研究员. 2001 年获中国科学院系统所理学博士学位. 主要研究方向为摄像机标定，摄像机定位，三维重建.E-mail：yhwu@nlpr.ia.ac.cn

通讯作者:
谢远帆中国科学院自动化研究所博士研究生. 2007 年获中南大学信息科学与工程学院自动化专业学士学位. 主要研究方向为基于视觉的同步定位与地图创建.E-mail：yfxie@nlpr.ia.ac.cn