一种基于单目SLAM的无人机序列图像拼接方法

系统仿真学报 ›› 2016, Vol. 28 ›› Issue (10): 2349-2355.

一种基于单目SLAM的无人机序列图像拼接方法

马跃龙¹, 赵勇², 曹雪峰¹

1.信息工程大学,河南郑州 450002;
2.西北工业大学,陕西西安 710072

收稿日期:2016-05-23 修回日期:2016-07-16 出版日期:2016-10-08 发布日期:2020-08-13
作者简介:马跃龙(1984-),男,内蒙古开鲁,博士生,研究方向为计算机视觉;赵勇(1984-),男,湖南湘潭,硕士,研究方向为计算机视觉。
基金资助:
国家自然科学基金(41401465,41371384)

Mono-SLAM Based Method for Stitching Images of Sequence From UAV

Ma Yuelong¹, Zhao Yong², Cao Xuefeng¹

1. Information Engineering University, Zhengzhou 450002, China;
2. Northwestern Polytechnical University, Xi’an 710072, China

Received:2016-05-23 Revised:2016-07-16 Online:2016-10-08 Published:2020-08-13

摘要/Abstract

摘要： 由无人机序列图像生成正射影像图是获取大范围正射影像图的一种有效方法。基于SfM生成正射影像图的方法计算量大,耗时长,所生成的正射影像图可视化效果较差。设计了基于单目SLAM的无人机序列图像增量式拼接算法,实现了大范围场景正射影像图的实时生成。通过单目SLAM系统实时估计相机位置与姿态,获取无人机飞行轨迹,同时生成3D点云地图。通过SLAM环路检测与基于位姿图的局部优化,有效降低了估计误差累积的影响,在保持估计精度的同时实现了优化操作的最小化。利用所提出的自适应权重的多频带算法,实现正射影像图的实时拼接融合与显示。通过实验验证了本方法的实时性与有效性。

关键词: 序列图像, 图像拼接, 正射影像图, 单目SLAM

Abstract: It is an effective method to generate orthoimage from image sequence obtained from UAV. Most of the SfM-based methods for generating orthoimage are expensive computation and need hours or more to process image sequence, but the orthoimages generated finally are not satisfactory most of the time. An algorithm to stitch image sequence obtained from UAV incrementally was designed based on monocular SLAM generating orthoimage for large-scale instantaneously. Through the flow of SLAM, the system could estimate the position and attitude of the camera and the trajectory of UAV, meanwhile, generate a map of 3D point cloud. With the detection of loop-closure within SLAM and local optimization based on pose-graph, the system could minimize the impact of error accumulation and the computation for optimization, meanwhile, retain the accuracy. With the proposed adaptive weighted multiband algorithm, the system could fuse and visualize the orthoimage instantaneously. The experiment results demonstrate the efficient and effective of the proposed algorithm.

Key words: image sequence, image stitching, orthoimage, mono slam

中图分类号:

TP346.2

马跃龙, 赵勇, 曹雪峰. 一种基于单目SLAM的无人机序列图像拼接方法[J]. 系统仿真学报, 2016, 28(10): 2349-2355.

Ma Yuelong, Zhao Yong, Cao Xuefeng. Mono-SLAM Based Method for Stitching Images of Sequence From UAV[J]. Journal of System Simulation, 2016, 28(10): 2349-2355.

参考文献 30

[1]	万刚, 余旭初, 布树辉, 等. 无人机测绘技术及应用 [M]. 北京: 测绘出版社, 2014.
[2]	Kneip L, Furgale P.OpenGV: A unified and generalized approach to real-time calibrated geometric vision[C]//Robotics and Automation (ICRA), 2014 IEEE International Conference on. USA: IEEE, 2014: 1-8.
[3]	Verhoeven G.Taking computer vision aloft-archaeological three‐dimensional reconstructions from aerial photographs with photoscan[J]. Archaeological Prospection (S1099-0763), 2011, 18(1): 67-73.
[4]	Triggs B, McLauchlan P F, Hartley R I, et al. Bundle adjustment—a modern synthesis[M]// Vision algorithms: theory and practice. Germany: Springer Berlin Heidelberg, 1999: 298-372.
[5]	Bailey T, Durrant-Whyte H.Simultaneous localization and mapping (SLAM): Part I[J]. IEEE Robot. Autom. Mag.(S1070-9932), 2006, 13(2): 99-110.
[6]	Bailey T, Durrant-Whyte H.Simultaneous localization and mapping (SLAM): Part II[J]. IEEE Robot. Autom. Mag., 2006, 13(3): 108-117.
[7]	Smith R, M. S, Cheeseman P. Estimating uncertain spatial realtionships in robotics[J]. Autonomous Robot Vehicles, 1990: 167-193.
[8]	Castellanos J A, Montiel J M M, Neira J, et al. The SPmap: A probabilistic framework for simultaneous localization and map building[J]. IEEE Transactions on Robotics and Automation, 1999, 15(5): 948-952.
[9]	Montemerlo M, Thrun S, Koller D, et al.FastSLAM: A factored solution to Simultaneous localisation and mapping [C]// Proc. of the National Conference on Artificial Intelligence, 2002. California: The AAAI Press, 2002: 593-598.
[10]	Hahnel D, Burgard W, Fox D, et al.An efficient algorithm for generating maps of large-scale cyclic environments from new laser range measurements[C]// Proc. of the IEEE/RSJ Int.Conf. on Intelligent Robots and Systems, 2003. USA: IEEE, 2003.
[11]	Thrun S, Liu Y, Koller D.Simultaneous localisation and mapping with sparse extended information filters[J]. Journal of Robotics Research (S0278-3649), 2004, 23(7): 693-716.
[12]	Eustice R, Singh H, Leonard J J.Exactly sparse delayed-state filters[C]// Proc. of the IEEE Int. Conf. on Robotics & Automation, 2005. USA: IEEE, 2005.
[13]	Lu F, Milios E.Globally consistent range scan alignment for environment mapping[J]. Autonomous Robots (S0929-5593), 1997, 4(4): 333-349.
[14]	Kümmerle R, Grisetti G, Strasdat H, et al.G2o: A general framework for graph optimization[C]// Proceedings - IEEE International Conference on Robotics and Automation. USA: IEEE, 2011: 3607-3613.
[15]	Andrew J D, Ian D M D Reid, et al. MonoSLAM: Real-Time Single Camera SLAM[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence (S0162-8828), 2007, 29(6): 1052-1067.
[16]	Klein G, Murray D.Parallel tracking and mapping for small AR workspaces[C]// Mixed and Augmented Reality, 2007. ISMAR 2007, 6th IEEE and ACM International Symposium on. USA: IEEE, 2007: 225-234.
[17]	Ra´ul Mur-Artal J M M Montiel, Tard´ J D. ORB-SLAM: a Versatile and Accurate Monocular[J]. IEEE Transactions on Robotics (S1552-3098), 2015, 31(5): 1147-1163.
[18]	Gálvez-López D, Tardós J D.Bags of binary words for fast place recognition in image sequences[J]. IEEE Transactions on Robotics (S1552-3098), 2012, 28(5): 1188-1197.
[19]	Richard A Newcombe, Andrew J Davison.Live dense reconstruction using a single moving camera[J]. Computer Vision and Pattern Recognition (CVPR), 2010 IEEE Conference on (S1063-6919), 2010, 1498-1505.
[20]	Engel J, Schöps T, Cremers D.LSD-SLAM: Large-scale direct monocular SLAM[M]// Computer Vision-ECCV 2014. Geermany: Springer, 2014: 834-849.
[21]	Lovegrove S, Davison A J.Real-time spherical mosaicing using whole image alignment[M]// Computer Vision-ECCV 2010. Germany: Springer Berlin Heidelberg, 2010: 73-86.
[22]	Maki T, Kondo H, Ura T, et al.Photo mosaicing of tagiri shallow vent area by the auv" tri-dog 1" using a slam based navigation scheme[C]// OCEANS 2006. USA: IEEE, 2006: 1-6.
[23]	Brown M, Lowe D G.Automatic panoramic image stitching using invariant features[J]. International Journal of Computer Vision (S0920-5691), 2007, 74(1): 59-73.
[24]	Zaragoza J, Chin T-J, Tran Q-H, et al.As-projective-as-possible image stitching with moving DLT[J]. Pattern Analysis and Machine Intelligence, IEEE Transactions on, IEEE (S0162-8828), 2014, 36(7): 1285-1298.
[25]	Kwatra V, Schödl A, Essa I, et al.Graphcut textures: image and video synthesis using graph cuts[J]. ACM Transactions on Graphics (ToG)(S0730-0301), 2003, 22(3): 277-286.
[26]	Zhao Y, Xu D.Fast Image Blending Using Seeded Region Growing[M]// Advances in Image and Graphics Technologies. Germany: Springer, 2015: 408-415.
[27]	Burt P J, Adelson E H.A multiresolution spline with application to image mosaics[J]. ACM Transactions on Graphics (TOG)(S0730-0301), 1983, 2(4): 217-236.
[28]	P Perez M Gangnet, Blake A. Poisson image editing[C]// Geoscience and Remote Sensing Symposium, 2008. IGARSS 2008. USA: IEEE International, 2003.
[29]	Rublee E, Rabaud V, Konolige K, et al.ORB: an efficient alternative to SIFT or SURF[C]// Computer Vision (ICCV), 2011 IEEE International Conference on. USA: IEEE, 2011: 2564-2571.
[30]	Strasdat H, Davison A J, Montiel J M M, et al. Double window optimisation for constant time visual SLAM[C]// Computer Vision (ICCV), 2011 IEEE International Conference on. USA: IEEE, 2011: 2352-2359.