Perspective transformation method for position-attitude calculation of crowdsourcing aerial panoramic imaging

doi:10.13474/j.cnki.11-2246.2023.0100

Abstract

Abstract: For the lack of position-attitude information, crowdsourcing aerial panoramic imaging is not able to precisely locate and orient in the geographic information space.A perspective transformation-based method of position-attitude calculation is proposed. First of all, the principle of proposed method is illustrated and the calculation scheme is designed. Secondly, spherical panorama model is constructed and “flat-sphere-flat” model transformation is implemented by perspective transformation Meanwhile, iterative reweighted least squares(IRLS) is designed for position and attitude calculation based on the deformation characteristics of model transformation. Finally, feasibility and reliability of the method are experimentally verified by using aerial panoramic images of Zhifang reservoir as a case.

Key words: crowdsourcing aerial panoramic imaging, perspective transformation, IRLS, position-attitude calculation, overseas battlefield environment reconnaissance

CLC Number:

P237

PU Hongyi, LI Feng, XU Zheng. Perspective transformation method for position-attitude calculation of crowdsourcing aerial panoramic imaging[J]. Bulletin of Surveying and Mapping, 2023, 0(4): 29-34,40.

References

[1] TONG G,PANG X,YE N,et al. A precise spherical camera model based on multi-camera system[J]. Journal of Computational Information Systems,2013,9(3):897-805.
[2] 季顺平,史云. 多镜头组合型全景相机两种成像模型的定位精度比较[J]. 测绘学报,2014,43(12):1252-1258.
[3] 季顺平,史云. 车载全景相机的影像匹配和光束法平差[J]. 测绘学报,2013,42(1):94-100.
[4] 石丽梅,赵红蕊,李明海,等. 车载移动测图系统外方位元素标定方法[J]. 测绘学报,2015,44(1):52-58.
[5] ZHANG X,ZHAO P,HU Q,et al. A UAV-based panoramic oblique photogrammetry (POP) approach using spherical projection[J]. ISPRS Journal of Photogrammetry and Remote Sensing,2020,159:198-219.
[6] FANGI G. Towards an easier orientation for spherical photogrammetry[J]. International Archives of the Photogrammetry,Remote Sensing and Spatial Information Sciences,2015,XL-5/W4(5):279-283.
[7] FANGI G. Multiscale multiresolution spherical photog-rammetry with long focal lenses for architectural surveys[J]. International Archives of Photogrammetry,Remote Sensing and Spatial Information Sciences,2010,38:225-233.
[8] 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.
[9] 朱宁宁,杨必胜,陈驰,等. 点-线特征联合的全景图像位姿解算方法[J/OL]. 测绘学报.[2022-03-16]. http://kns.cnki.net/kcms/detail/11.2089.P.20220124.2004.002.html.
[10] 刘帅,陈军,孙敏,等. 一种球形立体全景的三维量测算法与实验[J]. 地球信息科学学报,2014,16(1):15-22.
[11] 刘帅,陈军,孙敏,等. 双球面投影几何可量测全景模型的构建[J]. 计算机辅助设计与图形学学报,2015,27(4):657-665.
[12] 谢东海,钟若飞,吴俣,等. 球面全景影像相对定向与精度验证[J]. 测绘学报,2017,46(11):1822-1829.
[13] 国雷. 基于球面全景图像的机器人定位[D]. 重庆:西南大学,2021.
[14] 李加元,胡庆武,艾明耀. 以重心坐标为基准的空间后方交会非迭代法[J]. 测绘学报,2015,44(9):988-994.
[15] 张保明,龚志辉,郭海涛. 摄影测量学[M]. 北京:测绘出版社,2008.
[16] 周润,张征宇,黄叙辉. 相机位姿估计的加权正交迭代算法[J]. 光学学报,2018,38(5):193-199.
[17] DU Z Y,ROBLES-KELLY A,LU F F. Robust surface reconstruction from gradient field using the L1 norm[C]//Proceedings of the 9th Biennial Conference of the Australian Pattern Recognition Society on Digital Image Computing Techniques and Applications. New York:ACM,2007:203-209.
[18] 吴幼丝,邓非,万方. 基于本质矩阵鲁棒估计的球形全景匹配[J]. 测绘地理信息,2018,43(3):119-123.
[19] 吴幼丝. 球形全景影像位姿估计[D]. 武汉:武汉大学,2017.