Batched 3D Reconstruction of UAV Images Fused Low Precision Position and Orientation Parameters

Abstract

Abstract: The images sequence captured by UAV has the high resolution and high overlapping characteristics, so it is used in a wide range of scene reconstruction. However, the result of these state-of-art multi-view 3D reconstruction algorithms is unstable when dealing with large-scale UAV thousands of images. In this regard, the UAV low accuracy GPS / IMU parameters sequence was fused to the large-scale UAV image reconstruction process and the UAV sequence batched image reconstruction method was proposed. The image matching time and false matching number was reduced by using low-precision GPS / IMU priori information. Through the establishment of a polar graph and drawing the midpoint of the track multi-view, combined with low-accuracy GPS / IMU information, the global coordinate system rotation matrix of all the images was disposably solved, executing only when a bundle adjustment function, reducing the reconstruction optimization of time complexity. Experimental verification of this method ensures accuracy while improving efficiency.

Key words: image sequence, 3D reconstruction, GPS/IMU parameters, PMVS

CLC Number:

TP346.2

Cao Xuefeng, Wang Qinghe, Xie Lixiang. Batched 3D Reconstruction of UAV Images Fused Low Precision Position and Orientation Parameters[J]. Journal of System Simulation, 2015, 27(10): 2409-2415.

References

[1] Kersten T P, Lindstaedt M.Potential of Automatic 3D Object Reconstruction from Multiple Images for Applications in Architecture, Cultural Heritage and Archaeology[J]. International Journal of Heritage in the Digital Era (S2047-4970), 2012, 1(3): 399-420.
[2] Sawyer S M, Ni K, Bliss N T.Cluster-based 3D reconstruction of aerial video[C]// High Performance Extreme Computing, 2012 IEEE Conference on IEEE. USA: IEEE, 2012: 1-6.
[3] 李立春, 张恒, 李由, 等. 一种基于无人机序列成像的地形地貌重建方法[J]. 遥感技术与应用, 2008, 23(5): 505-510.
[4] Shum H, Ke Q, Zhang Z.Efficient Bundle Adjustment with Virtual Key Frames: A Hierarchical Approach to Multi-Frame Structure from Motion[C]// IEEE Conference on Computer Vision and Pattern Recognition. USA: IEEE, 1999: 538-543.
[5] C Strecha, T Pylvanainen, P Fua.Dynamic and scalable large scale image reconstruction[C]// IEEE Conference on Computer Vision and Pattern Recognition. USA: IEEE, 2010: 406-413.
[6] Carceroni R, Kumar A, Daniilidis K.Structure from Motion with Known Camera Positions[C]// IEEE Conference on Computer Vision and Pattern Recognition. USA: IEEE, 2006: 477-484.
[7] Yasutaka Furukawa, Jean Ponce.Accurate, dense, and robust multiview stereopsis[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence (S0162-8828), 2010, 32(8): 1362-1376.
[8] Lowe D G.Distinctive Image Features from Scale-Invariant Keypoints[J]. International Journal of Computer Vision (S0920-5691), 2004, 60(2): 91-110.
[9] Mikolajczyk K, Tuytelaars T, Schmid C, et al. Gool. A comparison of affine region detectors[J]. International Journal of Computer Vision (S0920-5691), 2005, 65(1/2): 43-72.
[10] D Nister.An efficient solution to the five-point relative pose problem[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence (S0162-8828), 2004, 26(6): 756-770.
[11] M A Fischler, R C Bolles.Random sample consensus: a paradigm for model fitting with application to image analysis and automated cartography[J]. Communication Association and Computing Machine (S0001-0782), 1981, 24(6): 381-395.
[12] Martinec D, Pajdla T. Robust Rotation, Translation Estimation in Multiview Reconstruction[C]// IEEE Conference on Computer Vision and Pattern Recognition. USA: IEEE, 2007: 1-8.
[13] Peter N Belhumeur.A bayesian approach to binocular stereopsis[J]. International Journal of Computer Vision (S0920-5691), 1996, 19(3): 237-260.