摘要
无人飞行器(Unmanned Aerial Vehicle, UAV)作为智能移动机器人之一因在航空摄影、灾情调查、紧急救援、国土资源勘察等方面的出色表现,以及军事上的特殊作用,倍受航空大国和军事强国的青睐。在UAV的各项关键技术中,视觉导航技术能重建UAV飞行环境,用于引导UAV在复杂环境下安全自主飞行,是近年来研究的热点。
本文利用视觉传感器,瞄准UAV视觉导航技术发展方向,以双目立体视觉和基于单目视觉的运动恢复结构(SFM)为基础,侧重研究具有高精度的单目合成视觉基本原理,开发鲁棒的合成视觉算法,构建完整的软件处理系统。
基于运动视差分解思想,提出了一种基于SURF特征的单目合成视觉三维重建算法。首先,通过基于SURF特征的运动补偿消除关于参考平面的视差部分；然后,创建拉普拉斯图像金字塔降低光照变化影响,利用残留视差算法在金字塔各层级上估计结构参数并逐级求精；最后,在给定摄像机参数和位置信息情况下完成三维重建。通过对模拟场景重建结果的定量分析和对室外光照变化场景重建结果的定性分析,证明该算法比双目立体视觉和SFM具有更高的精度,且具有较好的鲁棒性。
Unmanned Aerial Vehicle (UAV) is one of automatic intelligence robots, due to the outstanding performance on aerial photography, disaster investigation, emergency and natural resource exploration, as well as the special role of the military, it attracts more attentions in military and air power. As one of the UAV key technologies, visual navigation was studied a lot in recent years, which could rebuild UAV flight environment for secure and autonomous flight in complex environments.
Using vision sensor and aiming the development direction of UAV visual navigation, while stereo vision and structure from motion (SFM) are investigated, we focus on the research with the basic principle of high precision monocular synthetic vision, exploiting the robust algorithm and establishing the complete software system.
On base of motion parallax decomposition, a synthetic vision three-dimensional (3D) reconstruction algorithm based on SURF (Speed up Robust Features) is proposed. First, remove the part of planar plane parallax by SURF characterization motion compensation. Then, the image Laplace pyramid is created to reduce the lighting variation, and the structure parameters are estimated on each pyramid level with residual parallax algorithm and refinement. At last, the 3D reconstruction is completed with the parameters given camera matrix and location information. With the quantitative analysis of the simulation scene and the qualitative analysis of the outdoor scene, the approach with higher precision and better robustness are obtained in comparison with stereo vision and SFM.
引文
[1]Geyer C., Templeton T. et al. The recursive multi-frame planar parallax algorithm. 2006 IEEE Proceedings of the Third International Symposium on 3D Data Processing, Visualization and Transmission,2006, pp.17-24.
[2]Stefan H., Gaurav S.S.. A comparison of two camera configurations for optic-flow based navigation of a UAV through urban canyons. IEEE International Conference on IRS,2004, pp.2673-2680.
[3]Stefan H., Gaurav S.S., Prter C.. Combined optic-flow and stereo-based navigation of urban canyons for a UAV. IEEE ICRA,2005, pp.1-7.
[4]Hanna E., Straznicky P., Goubran R.. Obstacle detection for low flying unmanned aerial vehicles using stereoscopic imaging.2008 IMTC 2008 IEEE Instrumentation and Measurement Technology Conference Proceedings,2008,3, pp.113-118.
[5]Hrabar S..3D path planning and stereo-based obstacle avoidance for rotorcraft UAVs.2008 IROS 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems,2008,9, pp.807-814.
[6]Byrne Jeffrey, Cosgrove Martin, Mehra Raman. Real time stereo based obstacle detection for UAV threat avoidance. http://www.jeffreybyrne.com/docs/byrne-icra06.pdf.
[7]Call B., Beard R., Taylor C.. Obstacle avoidance for unmanned air vehicles using image feature tracking. AIAA Conference on Guidance, Navigation and Control, Keystone, CO, USA, Aug,2006.
[8]Shakernia O., Vidal R., Sharp C.S., et al. Multiple view motion estimation and control for landing an unmanned aerial vehicle. Proceedings of 2002 IEEE International Conference on Robotics and Automation, Washington DC,2002, pp.2793-2798.
[9]Watanabe Y., Calise A.J., Johnson E.N.. Vision-based obstacle avoidance for UAVs. AIAA Guidance, Navigation and Control Conference and Exhibit, South Carolina,2007,8.
[10]Kendoul F., Fantoni I., Dherbomez G.. Three nested Kalman filters-based algorithm for real-time estimation of optical flow, UAV motion and obstacles detection.2007 IEEE International Conference on Robotics and Automation, 2007,pp.4746-4751.
[11]Irani M., Anandan P., Cohen M.. Direct recovery of planar-parallax from multiple frames. IEEE Transactions on Pattern Analysis and Machine Intelligence,2002, 24(11), pp.1528-1534.
[12]Templeton T., Shim D. H., Geyer C. et al. Autonomous vision-based landing and terrain mapping using an MPC-controlled unmanned rotorcraft.2007 IEEE International Conference on Robotics and Automation,2007, pp.1349-1356.
[13]Xiang Pan, De-qiang Ma, Li-ling Jin et al. Vision-based approach angle and height estimation for UAV landing.2008 Congress on Image and Signal Processing, Sanya, Hainan, China,2008.
[14]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.遥感技术与应用,2008,23(05),pp.505-510.
[15]陈磊,陈宗基.基于稀疏光流的无人机自主导航方案.航空学报,2008,29,增刊,pp.s128-s134.
[16]周宇,黄显林,介鸣等.Bayesian分类方法在微小型飞行器视觉导航中的应用.2007,28,增刊,pp.s136-s140.
[17]徐贵力,倪立学,程月华.基于合作目标和视觉的无人飞行器全天候自动着陆导引关键技术.航空学报,2008,29(02),pp.437-442.
[18]吴益明,卢京潮,连玮,等.基于双谱相位算法的无人机图像定位研究.飞行器测控学报,2006,25(03),pp.20-24.
[19]张光富.基于单目合成长基线的3D重建技术研究.浙江大学硕士论文,2008.
[20]Kumar R., Anandan P., Hanna K. Direct recovery of shape from multiple views:a parallax based approach. Proceedings of the 12th IARR International Conference on Computer Vision and Image Processing,1994(1), pp.685-688.
[21]Sawhney H.S.3D geometry from planar parallax.1994 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Seattle, WA,1994, pp.9292-934.
[22]Lucas B., Kanade T.. An iterative image registration technique with an application to stereo vision. Proceedings of 7th International Joint Conference on Artificial Intelligence,1981.
[23]Koga T., Linuma K., Hirano A. et al. Motion compensated inter-frame coding for video. Proceedings of NTC 81, New Orleans, LA,1981.
[24]Reoxiang Li, Bing Zeng, Liou M.L. A new three-step search algorithm for block motion estimation.1994 IEEE Transactions on Circuits and Systems for Video Technology,1994,4(04), pp.438-442.
[25]Lai-Man Po, Wing-Chung Ma. A novel four-step search algorithm for fast block motion estimation.1996 IEEE Transactions on Circuits and Systems for Video Technology,1996,6(3), pp.313-317.
[26]Bergen J. R., Anandan P., Hanna K.J. et al. Hierarchical model-based motion estimation. Proceedings of the Second European Conference on Computer Vision, 1992,pp.237-252.
[27]Hill L., Vlachos T.. On the estimation of global motion using phase correlation for broadcast applications.1999 Seventh International Conference on Image Processing and Its Applications, Manchester,1999,2, pp.721-725.
[28]Szeliski R. Image alignment and stitching:A tutorial. Foundations and Trends in Computer Graphics and Vision,2006,2(1), pp.1-104.
[29]Gordon I., Lowe D. G. Scene modelling, recognition and tracking with invariant image features. Proceedings of the 3rd IEEE/ACM International Symposium on Mixed and Augmented Reality (ISMAR),2004, pp.110-119.
[30]杜歆.用于导航的立体视觉系统.浙江大学博士论文,2003.
[31]张光富.基于合成视觉的3D重建技术研究.浙江大学硕士论文,2008.
[32]邬领东.ALV立体视觉导航三维地形重建算法研究.浙江大学硕士论文,2006.
[33]Zheng Y.Z.. Flexible camera calibration by viewing a plane from unknown orientations. The Proceedings of the 7th IEEE International Conference on Computer Vision,1999, 1,pp.666-673.
[34]Kanade T., Okutomi M.. A stereo matching algorithm with an adaptive window: theory andexperiment. Pattern Analysis and Machine Intelligence, IEEE Transactions,1994(16), pp.920-932.
[35]Anandan. P. A computational framework and an algorithm for the measurement of visual motion. International Journal of Computer Vision.1989, pp.283-310.
[36]Faugeras O., Vieville T., Theron E. et al. Real-time correlation-based stereo: algorithm, implementations and applications. Institut national de recherche en informatiue et en automatique,1993(4):35-48.
[37]Wu Q. X., S. J McNeill, Pairman D. Fast algorithms for correlation-relaxation technique to determine cloud motion fields[J].Digital Image Computing: Techniques and Applications,1995(47):330-350.
[38]Hirschmuller H.. Improvements in real-time correlation-based stereo vision. Stereo and Multi-Baseline Vision, Proceedings. IEEE Workshop on,2001,12, pp.141-148.
[39]徐光佑.计算机视觉.清华大学出版社,1999.
[40]周围.基于单目视觉的运动估计与结构重建.浙江大学硕士论文,2007.
[41]Shirai Y.. Three-dimensional computer vision. Springer-Verlag New York, Inc., 1987.
[42]Matthies L., SHAFER S.. Error modeling in stereo navigation. IEEE Journal of Robotics and Automation,1987,3(3), pp.239-248.
[43]马颂德,张正友.计算机视觉-计算理论与算法基础.科学出版社,1998.
[44]Irani M., Anandan P.. A unified approach to moving object detection in 2 d and 3d scenes. IEEE Transactions on Pattern Analysis and Machine Intelligence,1998, 20(6), pp.577-589.
[45]Irani M., Anandan P., Weinshall D.. From reference frames to reference planes: Multi-view parallax geometry and applications. Lecture Notes in Computer Science,1998,1407, pp.829-845.
[46]Irani M., Anandan P.. Parallax geometry of pairs of points for 3D scene analysis. Proceedings of the 4th European Conference on Computer Vision.1996,1064, pp.17-30.
[47]贾云得.机器视觉.科学出版社,2000.
[48]Horn B.K.P, Schunck B.G.. Determing optical flow. Artificial Intelligence, 1981,17(18), pp.185-203.
[49]章毓晋.图像工程(上册)图像处理第2版.清华大学出版社,2006.
[50]张锐娟,张建奇,杨翠等.基于SURF的图像配准方法研究.红外与激光工程,2009,38(1).
[51]Bay H., Tuvtelaars T. et al. SURF:speed up robust features. Proceedings of the European Conference on Computer Vision,2006.
[52]Ma Yi, Soatto Stefano, Kosecka Jana. An Invitation to 3-D Vision:From Images to Geometric Models.2003.
[53]Horn B.K.P., Schunck B.G. Determining optical flow. Artificial Intelligence, 1981,17(18), pp.185-203.
[54]孙鑫.VC++深入详解.电子工业出版社,2008.
[55]Bradski G., Kaebler A.. Learning OpenCV:computer vision with OpenCV Library.2008.
[56]于仕琪.OpenCV中文网站.http://www.opencv.org.cn.