相机位姿求解的复用镜像约束法

doi:10.11947/j.AGCS.2024.20220362

测绘学报 ›› 2024, Vol. 53 ›› Issue (3): 548-557.doi: 10.11947/j.AGCS.2024.20220362

相机位姿求解的复用镜像约束法

阿晓荟¹, 李佳田¹, 刘佳音¹, 胡浩¹, 贺日兴^2,3, 陆美¹, 段烨¹

1. 昆明理工大学国土资源工程学院, 云南昆明 650093;
2. 首都师范大学资源环境与旅游学院, 北京 100048;
3. 首都师范大学三维数据获取与应用教育部重点实验室, 北京 100048

收稿日期:2022-05-30 修回日期:2023-05-16 发布日期:2024-04-08
通讯作者: 李佳田 E-mail:ljtwcx@163.com
作者简介:阿晓荟(1995—),女,博士生,研究方向为摄影测量与遥感。E-mail:1665369329@163.com
基金资助:
国家自然科学基金(41561082)

Solving camera pose with multiplexing mirror constraint

A Xiaohui¹, LI Jiatian¹, LIU Jiayin¹, HU Hao¹, HE Rixing^2,3, LU Mei¹, DUAN Ye¹

1. Faculty of Land and Resources Engineering, Kunming University of Science and Technology, Kunming 650093, China;
2. College of Resources Environment and Tourism, Capital Normal University, Beijing 100048, China;
3. Key Laboratory of 3D Information Acquisition and Application, MOE, Capital Normal University, Beijing 100048, China

Received:2022-05-30 Revised:2023-05-16 Published:2024-04-08
Supported by:
The National Natural Science Foundation of China (No. 41561082)

摘要/Abstract

摘要： 针对参照物不在视场范围内的问题,提出一种复用镜像约束求解相机位姿的方法。具体步骤为:①依据相机位姿求解方法得到镜像相机位姿;②根据平面镜反射对称性构建单张图像镜像相机与相机位姿间线性约束关系;③复用镜像相机姿态,建立多张图像之间约束关系,在获得3张图像情况下,可得18个线性约束条件,提高相机位姿求解精度。试验结果表明:旋转矩阵和平移向量均误差分别为0.009°和1.866mm,较对比方法,旋转矩阵均误差平均降低0.009°,平移向量均误差平均降低2.292mm,且具有较强适应性。在真实试验中,平均重投影误差为0.492像素,优于对比方法,能够准确求解参照物不在视场范围内的相机位姿。

关键词: 相机位姿, 平面镜反射对称性, 镜像相机, 最小二乘

Abstract: Aiming at the problem that the reference object is not in the field of view, a method with constraints of mirror camera to solve camera pose is proposed.The steps are:① Mirror camera pose is solved by using the method of camera pose estimation. ② The linear constraint relationship between the pose of camera and single image mirror camera is constructed with mirror reflection symmetry. ③ To establish constraints between multiple images, the mirror camera poses were multiplexed, 18 linear constraints can be obtained using 3 images. This makes the accuracy of camera pose solution improved. Experiments show that:the errors of rotation matrix and translation vector are 0.009° and 1.866mm respectively, and the corresponding errors are reduced by 0.009°and 2.292mm on average, and has a strong adaptability. In real experiments, the average reprojection error is 0.492 pixels. This suggests that the method in this paper is better than the comparison methods, and the pose of the camera where the reference object is not in the field of view can be accurately solved.

Key words: camera pose, mirror reflection symmetry, mirror camera, least squares

中图分类号:

P234

阿晓荟, 李佳田, 刘佳音, 胡浩, 贺日兴, 陆美, 段烨. 相机位姿求解的复用镜像约束法[J]. 测绘学报, 2024, 53(3): 548-557.

A Xiaohui, LI Jiatian, LIU Jiayin, HU Hao, HE Rixing, LU Mei, DUAN Ye. Solving camera pose with multiplexing mirror constraint[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(3): 548-557.

参考文献

[1] WANG Ping, CHOU Yongxin, AN Aimin, et al. Solving the PnL problem using the hidden variable method:an accurate and efficient solution[J]. The Visual Computer, 2022, 38(1):95-106.
[2] 李佳田, 贾成林, 牛一如, 等. 一种求解单像空间后方交会的监督学习方法[J]. 武汉大学学报(信息科学版), 2019, 44(8):1144-1152. LI Jiatian, JIA Chenglin, NIU Yiru, et al. A supervised learning method for solving space resection of single image[J]. Geomatics and Information Science of Wuhan University, 2019, 44(8):1144-1152.
[3] 尚洋, 孙晓亮, 张跃强, 等. 三维目标位姿跟踪与模型修正[J]. 测绘学报, 2018, 47(6):799-808.DOI:10.11947/j.AGCS.2018.20170626. SHANG Yang, SUN Xiaoliang, ZHANG Yueqiang, et al. Research on 3D target pose tracking and modeling[J]. Acta Geodaetica et Cartographica Sinica, 2018, 47(6):799-808.DOI:10.11947/j.AGCS.2018.20170626.
[4] 张帆, 黄印, 黄先锋, 等. 非直视区域的普通平面镜辅助地面三维激光扫描方法[J]. 测绘学报, 2017, 46(12):1950-1958.DOI:10.11947/j.AGCS.2016.20150411. ZHANG Fan, HUANG Yin, HUANG Xianfeng, et al. 3D laser scanning assisted by ordinary plane mirror for non-direct viewing area[J]. Acta Geodaetica et Cartographica Sinica, 2017, 46(12):1950-1958.DOI:10.11947/j.AGCS.2016.20150411.
[5] AGRAWAL A. Extrinsic camera calibration without a direct view using spherical mirror[C]//Proceedings of 2013 IEEE International Conference on Computer Vision. Sydney:IEEE, 2013:2368-2375.
[6] FRANCKEN Y, HERMANS C, BEKAERT P. Screen-camera calibration using a spherical mirror[C]//Proceedings of the 4th Canadian Conference on Computer and Robot Vision.Piscataway:ACM Press, 2007:11-20.
[7] FRANCKEN Y, HERMANS C, BEKAERT P. Screen-camera calibration using gray codes[C]//Proceedings of 2009 Canadian Confe-rence on Computer and Robot Vision. Kelowna:IEEE, 2009:155-161.
[8] 付生鹏, 赵吉宾, 夏仁波, 等. 基于环形镜面的相机外部参数自动标定方法[J]. 机器人, 2015, 37(3):277-285. FU Shengpeng, ZHAO Jibin, XIA Renbo, et al. Annular mirror based extrinsic camera automatic calibration[J]. Robot, 2015, 37(3):277-285.
[9] NITSCHKE C, NAKAZAWA A, TAKEMURA H. Display-camera calibration from eye reflections[C]//Proceedings of the 12th IEEE International Conference on Computer Vision. Kyoto:IEEE, 2009:1226-1233.
[10] FUNK N, YANG Y H. Using a raster display for photometric stereo[C]//Proceedings of the 4th Canadian Conference on Computer and Robot Vision. Montreal:IEEE, 2007:201-207.
[11] STURM P, BONFORT T. How to compute the pose of an object without a direct view?[C]//Proceedings of the 7th Asian Conference on Computer Vision-Volume Part II.Hyderabad:ACM Press, 2006:21-31.
[12] HESCH J A, MOURIKIS A I, ROUMELIOTIS S I. Mirror-based extrinsiccamera calibration[J]. Springer Tracts in Advanced Robotic,2009, 57:285-299.
[13] RODRIGUES R, BARRETO J P, NUNES U. Camera pose estimation using images of planar mirror reflections[C]//Proceedings of the 11th European conference on Computer vision:Part IV.Heraklion:ACM Press, 2010:382-395.
[14] TAKAHASHI K, NOBUHARA S, MATSUYAMA T. A new mirror-based extrinsic camera calibration using an orthogonality constraint[C]//Proceedings of 2012 IEEE Conference on Computer Vision and Pattern Recognition.Providence:IEEE, 2012:1051-1058.
[15] TAKAHASHI K, NOBUHARA S, MATSUYAMA T. Mirror-based camera pose estimation using an orthogonality constraint[J]. IPSJ Transactions on Computer Vision and Applications, 2016, 8:11-19.
[16] LONG Gucan, KNEIP L, LI Xin, et al. Simplified mirror-based camera pose computation via rotation averaging[C]//Proceedings of 2015 IEEE Conference on Computer Vision and Pattern Recognition.Boston:IEEE, 2015:1247-1255.
[17] KUMAR R K, ILIE A, FRAHM J M, et al. Simple calibration of non-overlapping cameras with a mirror[C]//Proceedings of 2008 IEEE Conference on Computer Vision and Pattern Recognition. Anchorage:IEEE, 2008:1-7.
[18] 付生鹏, 赵吉宾, 夏仁波, 等. 基于平面镜的相机-显示器快速标定方法[J]. 仪器仪表学报, 2013, 34(12):2770-2776. FU Shengpeng, ZHAO Jibin, XIA Renbo, et al. A mirror-based fast camera-screen calibration method[J]. Chinese Journal of Scientific Instrument, 2013, 34(12):2770-2776.
[19] ZHANG Zhengyou. A flexible new technique for camera calibration[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(11):1330-1334.
[20] 邓非, 吴幼丝, 胡玉雷, 等. 球形全景影像位姿估计的改进EPnP算法[J]. 测绘学报, 2016, 45(6):677-684.DOI:10.11947/j.AGCS.2016.20150411. DENG Fei, WU Yousi, HU Yulei,et al. Position and pose estimation of spherical panoramic image with improved EPnP algorithm[J]. Acta Geodaetica et Cartographica Sinica, 2016, 45(6):677-684.DOI:10.11947/j.AGCS.2016.20150411.
[21] LI Shiqi, XU Chi, XIE Ming. A robust O(n) solution to the perspective-n-point problem[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(7):1444-1450.
[22] GOLUB G H, VAN LOAN C F. Matrix computations[M]. 3rd ed. Baltimore:Johns Hopkins University Press, 1996.
[23] 李佳田, 王聪聪, 阿晓荟, 等. 多向前方交会与单隐层神经网络结合的近景前方交会法[J]. 测绘学报, 2020, 49(6):736-745.DOI:10.11947/j.AGCS.2020.20180600. LI Jiatian, WANG Congcong, A Xiaohui, et al. Method of close-range space intersection combining multi-image forward intersection with single hidden layer neural network[J]. Acta Geodaetica et Cartographica Sinica, 2020, 49(6):736-745.DOI:10.11947/j.AGCS.2020.20180600.
[24] LI Jiatian, WANG Congcong, JIA Cheng, et al. A hybrid conjugate gradient algorithm for solving relative orientation of big rotation angle stereo pair[J]. Journal of Geodesy and Geoinformation Science, 2020, 3(2):62-70.
[25] SATTLER T, MADDERN W, TOFT C, et al. Benchmarking 6DOF outdoor visual localization in changing conditions[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Salt Lake City:IEEE, 2018:8601-8610.

相机位姿求解的复用镜像约束法

Solving camera pose with multiplexing mirror constraint

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	杨玲, 朱金成, 孙楠, 喻杨康, 沈云中, 李博峰. GNSS高级接收机自主完好性监测随机模型精化[J]. 测绘学报, 2024, 53(2): 286-295.
[2]	刘昌振, 马威, 马红, 魏世轩. 建筑物轮廓方向计算和规则化的向量重组算法[J]. 测绘学报, 2023, 52(9): 1584-1594.
[3]	赵闯, 金涛勇, 秦朋波, 杨连俊. 顾及残差约束的多面函数法融合卫星测高和船载重力[J]. 测绘学报, 2023, 52(4): 605-613.
[4]	范雕, 李姗姗, 冯进凯, 黄炎, 范昊鹏, 张金辉, 李新星. 基于最小二乘配置的陌生海域海底地形反演方法[J]. 测绘学报, 2023, 52(12): 2039-2053.
[5]	王珂, 刘国林, 付政庆, 王路遥. 基于Moore-Penrose广义逆及立体矩阵的可分离非线性最小二乘解算方法[J]. 测绘学报, 2022, 51(3): 340-350.
[6]	鄢咏折, 彭松, 马友青, 张烁, 亓晨, 温博, 李昊, 贾阳, 刘少创. 行星探测器导航相机着陆后参数标定方法[J]. 测绘学报, 2022, 51(3): 437-445.
[7]	王乐洋, 陈涛. 加乘性混合误差模型精度评定的SUT法[J]. 测绘学报, 2022, 51(11): 2303-2316.
[8]	王路遥, 刘国林, 王凤云, 王珂, 韩宇. 基于矩阵分解的可分离非线性最小二乘问题求解方法及其应用[J]. 测绘学报, 2022, 51(11): 2317-2327.
[9]	闫凤池, 王振杰, 赵爽, 聂志喜, 孙振, 李伟嘉. 顾及双程声径的常梯度声线跟踪水下定位算法[J]. 测绘学报, 2022, 51(1): 31-40.
[10]	刘计洪, 胡俊, 李志伟, 朱建军. InSAR三维同震地表形变监测——窗口优化的SM-VCE算法[J]. 测绘学报, 2021, 50(9): 1222-1239.
[11]	范雕, 李姗姗, 欧阳永忠, 孟书宇, 陈成, 邢志斌, 张驰. 顾及海底地形非线性项的最小二乘配置反演方法[J]. 测绘学报, 2021, 50(7): 953-971.
[12]	郑凯, 李建胜, 王俊强, 欧阳文, 谷友艺, 张迅. DCLS-GAN:利用生成对抗网络的天绘一号卫星高原地区影像去云方法[J]. 测绘学报, 2021, 50(2): 248-259.
[13]	张兵兵, 牛继强, 王正涛, 徐丰, 田坤俊. Swarm系列卫星非差运动学厘米级精密定轨[J]. 测绘学报, 2021, 50(1): 27-36.
[14]	胡川, 方兴, 赵立都. 融合正交几何信息的非线性等式约束整体最小二乘平差及迭代算法[J]. 测绘学报, 2020, 49(7): 816-823.
[15]	解清华, 朱建军, 汪长城, 付海强, 张兵. 基于S-RVoG模型的PolInSAR森林高度非线性复数最小二乘反演算法[J]. 测绘学报, 2020, 49(10): 1303-1310.