基于视觉测量的机翼三维形变检测技术研究

机翼的形变检测对于飞行器安全监测和空气动力特性验证有着十分重要的价值，而视觉测量具有高精度、非接触、大范围、实时检测等优点，可以直接对被摄物的三维形变进行反演，近年来得到了广泛应用和发展。为了突破视觉测量在机翼形变检测应用中的关键技术，本文开展了对该技术的广泛调研，从三维形变测量系统设计、摄像机标定、图像相关匹配以及形变实验验证四个方面展开研究。
  对于三维形变测量系统设计，本文采用双目立体视觉+数字图像相关的软件方案，采用两台分辨率为1200万像素的工业相机搭配12mm焦距镜头实现图像采集。按照由机体观测机翼的方向进行系统布局，可以对600mm*230mm的机翼区域进行测量。通过实验验证了所搭建系统景深和视野范围的可行性。
  对于摄像机标定关键技术，本文以针孔摄像模型和畸变模型为基础，通过坐标系变换推导了标定具体方法，利用C++和OpenCV设计了系统的相机标定软件，实现了被摄物到图像投影关系的求解。本文提出了一种带方位点的圆点阵列特征提取算法，通过滤除受透视畸变影响较大的圆点特征，减小偏心误差的影响。通过与传统标定图案的实验对比，证明了其在大景深场景下的鲁棒性。
  对于图像相关匹配技术，本文对以下方面进行深入研究：（1）3D-DIC的数学模型；（2）相关标准选取；（3）非线性迭代优化算法，包括FA-GN和IC-GN两种方法及其停机准则；（4）几种初值估计方法以及对其的改进；（5）提出一种带松弛因子的IC-GN算法，利用因子改变更新步长提升迭代速率，通过标准图像证明了该算法结果收敛且在该数据集中速率提升82.2%。在此基础上，本文在MATLAB上开发了3D-DIC形变测量软件，在iDICs的标准图像集验证下，软件的精确帧更新均方根测量误差为0.0518mm，模糊帧更新均方根误差0.0934mm。
  最后，本文对所设计系统的三维重建和形变测量效果进行了验证。通过对半径48.72mm的圆柱体进行测量，圆拟合所得点云的均方根半径误差为0.129mm；通过对翼展1300mm的无人机机翼附加载荷时的离面位移进行测量，600mm测量范围下的均方根误差为0.11mm，在相机景深中段误差仅0.10mm，证明了该系统较高的形变测量精度以及在机翼形变检测应用中的可行性。

 The deformation measurement of wing is very important for aircraft safety inspection and aerodynamic characteristics verification. Vision technology has the advantages of high precision, non-contact, large-scale, real-time detection, etc. It can directly measure the three-dimensional deformation of surface so that has been widely used and developed in recent years. In order to breakthrough the key part of vision technology in the application of wing measurement, the research was carried out from four aspects: design of 3D deformation measurement system, camera calibration, image matching and experiment verification.
   
   The design of 3D deformation measurement system adopts the software solution related to binocular stereo vision + digital image correlation. The system realizes image acquisition through two industrial cameras with a resolution of 12 million and a 12mm focal length lens. The system is arranged according to the direction of observing wing from airframe. The feasibility of basic function of the system was verified through prototype production and experiments.
   
   The camera calibration technology deduces its algorithm principle from the perspective of coordinate system transformation based on the pinhole camera model and the distortion model. In this paper, the camera calibration software of the deformation measurement system was designed with C++ and OpenCV, which link the physical location of surface point to the pixel position on the image plane. In this paper, a feature extraction algorithm of dot array with azimuth points is proposed, which can reduce the influence of eccentricity error by filtering out the dot features which are greatly affected by perspective distortion. By comparing with the traditional calibration pattern, the reliability of the software and the robustness of this algorithm in the scene with large depth of field were proved.
   
   The image matching is discussed from several aspects, including: (1) the mathematical model of three-Dimensional digital image correlation(3D-DIC); (2) correlation criterion selection; (3) non-linear optimization algorithm; (4) initial guess method; (5) the derivation of newly proposed inverse compositional Gauss-Newton algorithm with the relaxation factor which can increase the speed by changing the increment. After experimental validation on the standard image set, this algorithm have 82.2% increase in speed on this set. Based on this algorithm, this paper developed 3D-DIC deformation measurement software on MATLAB. Verified by iDICs' standard image set, RMS measurement error of the software is 0.0518mm under the precise frame update strategy and 0.0934mm under the fuzzy frame update strategy.
   Two real experiments were carried out to verify the accuracy of 3D reconstruction and deformation measurement of this system. Through the 3D reconstruction of the cylinder, RMS error of the cylinder radius obtained by circle fitting to 3D pointcloud is 0.129mm, which is close to the true value. By measuring off-surface displacement of wing with additional load, the RMS deformation measurement error of the system is 0.11mm, and the RMS error at the camera focus position is only 0.10mm, which proves the high accuracy of the system and its feasibility in the application of wing deformation measurement.

[1] 韩涛. 大型结构件弹性形变组网摄影测量[J]. 科学技术与工程, 2016, 16(32): 101-106.
[2] 何超. 基于光纤光栅的飞行器结构健康监测技术研究[D]. 南京: 南京航空航天大学, 2016.
[3] 于洋, 黄俊斌, 顾宏灿, 等. 基于光纤光栅传感器的飞行器结构健康监测[J]. 信息通信,2013(2): 27-29.
[4] 魏传达. 基于应变信息的飞机机翼变形测量及形变重构理论研究[D]. 西安电子科技大学,2015.
[5] 张新华. 基于光纤光栅的结构形状传感技术研究[D]. 南京航空航天大学, 2018.
[6] BODEN F, WOLF T, LANARI C, et al. Ipct ground vibration measurements on a small aircraft[M]//Advanced In-Flight Measurement Techniques. Springer, 2013: 33-55.
[7] MA Z, CHEN X. Fiber bragg gratings sensors for aircraft wing shape measurement: Recentapplications and technical analysis[J]. Sensors, 2019, 19(1): 55.
[8] KOMPENHANS J, SCHRÖDER A, ENGLER R, et al. Development and application of imagebased measurement techniques for aerodynamic investigations in wind tunnels[J]. High-SpeedFlow Fundamental Problems, 2004.
[9] MICHAELIS D, FRAHNERT H, STASICKI B. Accuracy of combined 3d surface deformationmesurement and 3d position tracking in a wind tunnel[Z]. 2004.
[10] 李巧真, 李刚, 韩钦泽. 电阻应变片的实验与应用[J]. 实验室研究与探索, 2011, 30(4):134-137.
[11] 阎楚良, 张书明. 飞机机翼结构载荷测量试验力学模型与数据处理[J]. 航空学报, 2000, 21(1): 56-59.
[12] 韩小进, 郭志强, 温晓月, 等. 大展弦比飞机机翼弯曲变形分析和测试方法[J]. 科学技术与工程, 2017, 17(30): 343-347.
[13] WISSMAN J, PEREZ-ROSADO A, EDGERTON A, et al. New compliant strain gauges forself-sensing dynamic deformation of flapping wings on miniature air vehicles[J]. Smart materials and structures, 2013, 22(8): 085031.
[14] JONES R, BERKOFF T A, BELLEMORE D G, et al. Cantilever plate deformation monitoringusing wavelength division multiplexed fiber bragg grating sensors[C]//Smart Structures andMaterials 1996: Smart Sensing, Processing, and Instrumentation: volume 2718. InternationalSociety for Optics and Photonics, 1996: 258-268.
[15] JONES R, BELLEMORE D, BERKOFF T, et al. Determination of cantilever plate shapes usingwavelength division multiplexed fiber bragg grating sensors and a least-squares strain-fittingalgorithm[J]. Smart materials and structures, 1998, 7(2): 178.
[16] DAVIS M, KERSEY A, SIRKIS J, et al. Shape and vibration mode sensing using a fiber opticbragg grating array[J]. Smart Materials and Structures, 1996, 5(6): 759.
[17] BANG H J, SHIN H K, JU Y C. Structural health monitoring of a composite wind turbineblade using fiber bragg grating sensors[C]//Sensors and Smart Structures Technologies for Civil,Mechanical, and Aerospace Systems 2010: volume 7647. SPIE, 2010: 1331-1342.
[18] CUI P, ZHANG W, WANG Z, et al. Measurement of aircraft wing deformation using fiber bragg gratings[C]//2017 16th International Conference on Optical Communications and Networks (ICOCN). IEEE, 2017: 1-3.
[19] MURAYAMA H, TACHIBANA K, HIRANO Y, et al. Distributed strain and load monitoring of 6 m composite wing structure by fbg arrays and long-length fbgs[C]//OFS2012 22nd International Conference on Optical Fiber Sensors: volume 8421. International Society for Optics and Photonics, 2012: 84212D.
[20] MILLER E J, PENA F, JORDAN A, et al. Evaluation of wing load calibration and sensing methods using conventional strain gages and a fiber optic sensing system installed on a straight tapered wing[C]//AIAA Scitech 2019 Forum. 2019: 0227.
[21] 德安, 万, 技术测量方法. 激光基准高精度测量技术[M]. 国防工业出版社, 1999.
[22] LI X, XU L, TAN C, et al. Real-time measurement of aerodynamic deformation of wing by laserrangefinder[C]//2010 IEEE Instrumentation & Measurement Technology Conference Proceedings. IEEE, 2010: 1581-1584.
[23] 季红侠. 飞机大部件对接中的自动测量技术研究与系统开发 [D][D]. 南京: 南京航空航天大学, 2012.
[24] KURITA M, KOIKE S, NAKAKITA K, et al. In-flight wing deformation measurement[C]//51st AIAA Aerospace Sciences Meeting including the New Horizons Forum and AerospaceExposition. 2013: 967.
[25] 谈高超. 基于机器视觉的批量零件自动检测系统的研制与工程应用[D]. 上海: 东华大学,2014.
[26] CHANG M, HU C P, LAM P, et al. High precision deformation measurement by digital phaseshifting holographic interferometry[J]. Applied optics, 1985, 24(22): 3780-3783.
[27] KHAN A S, WANG X. Strain measurements and stress analysis[M]. 2001.
[28] FEIN H. Holographic interferometry: nondestructive tool[J]. The Industrial Physicist, 1997, 3(3): 37-39.
[29] CHU T, RANSON W, SUTTON M A. Applications of digital-image-correlation techniques to experimental mechanics[J]. Experimental mechanics, 1985, 25(3): 232-244.
[30] KHOO S W, KARUPPANAN S, TAN C S. A review of surface deformation and strain measurement using two-dimensional digital image correlation[J]. Metrology and Measurement Systems,2016, 23(3).
[31] PETERS W, RANSON W. Digital imaging techniques in experimental stress analysis[J]. Optical engineering, 1982, 21(3): 427-431.
[32] 周轶昊. 基于双目视觉的物体表面三维复杂运动重建及其应用[D]. 复旦大学, 2012.
[33] FAYYAD T M, LEES J M. Application of digital image correlation to reinforced concrete fracture[J]. Procedia Materials Science, 2014, 3: 1585-1590.
[34] WANG Y, CUITIÑO A M. Full-field measurements of heterogeneous deformation patterns onpolymeric foams using digital image correlation[J]. International Journal of Solids and Structures, 2002, 39(13-14): 3777-3796.
[35] XU X, WANG K, GU G. An improved method for shape measurement using two-dimensionaldigital image correlation[J]. Optik, 2013, 124(20): 4097-4099.
[36] KAHN-JETTER Z, CHU T. Three-dimensional displacement measurements using digital imagecorrelation and photogrammic analysis[J]. Experimental Mechanics, 1990, 30(1): 10-16.
[37] LUO P, CHAO Y, SUTTON M, et al. Accurate measurement of three-dimensional deformationsin deformable and rigid bodies using computer vision[J]. Experimental mechanics, 1993, 33(2): 123-132.
[38] HELM J D, MCNEILL S R, SUTTON M A. Improved three-dimensional image correlation forsurface displacement measurement[J]. Optical Engineering, 1996, 35(7): 1911-1920.
[39] PAZMINO J, CARVELLI V, LOMOV S V, et al. 3d digital image correlation measurementsduring shaping of a non-crimp 3d orthogonal woven e-glass reinforcement[J]. Internationaljournal of material forming, 2014, 7(4): 439-446.
[40] 李婧宇, 朱飞鹏, 雷冬. 三维 DIC 在铸铁拉伸试验中的应变测量精度研究[J]. 力学季刊,2015, 36(3): 434-441.
[41] 高越. 三维数字图像相关法的关键技术及应用研究[J]. 博士论文]. 合肥: 中国科学技术大学, 2014.
[42] LEVENBERG K. A method for the solution of certain non-linear problems in least squares[J].Quarterly of applied mathematics, 1944, 2(2): 164-168.
[43] MARQUARDT D W. An algorithm for least-squares estimation of nonlinear parameters[J].Journal of the society for Industrial and Applied Mathematics, 1963, 11(2): 431-441.
[44] ZHANG Z. A flexible new technique for camera calibration[J]. IEEE Transactions on patternanalysis and machine intelligence, 2000, 22(11): 1330-1334.
[45] RAVN O, ANDERSEN N A, SØRENSEN A T. Auto-calibration in automation systems usingvision[M]//Experimental robotics III. Springer, 1994: 206-218.
[46] 起峰, 测量学, 宏伟, 等. 基于图像的精密测量与运动测量[M]. 科学出版社, 2002.
[47] MALLON J, WHELAN P F. Which pattern? biasing aspects of planar calibration patterns anddetection methods[J]. Pattern recognition letters, 2007, 28(8): 921-930.
[48] 张鑫, 习俊通. 双目立体测量系统标定及精度评价[J]. 计算机应用与软件, 2014, 31(7):160-163.
[49] DATTA A, KIM J S, KANADE T. Accurate camera calibration using iterative refinement ofcontrol points[C]//2009 IEEE 12th International Conference on Computer Vision Workshops,ICCV Workshops. IEEE, 2009: 1201-1208.
[50] ZHANG G, WEI Z. A position-distortion model of ellipse centre for perspective projection[J].Measurement Science and Technology, 2003, 14(8): 1420.
[51] SHEN Y, ZHANG X, CHENG W, et al. Quasi-eccentricity error modeling and compensation in vision metrology[J]. Measurement Science and Technology, 2018, 29(4): 045006.
[52] LOWE D G. Distinctive image features from scale-invariant keypoints[J]. International journalof computer vision, 2004, 60(2): 91-110.
[53] LUCAS B D, KANADE T, et al. An iterative image registration technique with an applicationto stereo vision[C]//Vancouver, 1981.
[54] 张华俊. 基于图像特征匹配技术的数字图像相关法研究[D]. 硕士论文], 安徽大学, 2014.
[55] SUTTON M A, ORTEU J J, SCHREIER H. Image correlation for shape, motion and deformation measurements: basic concepts, theory and applications[M]. Springer Science & BusinessMedia, 2009.
[56] PAN B, QIAN K, XIE H, et al. Two-dimensional digital image correlation for in-plane displacement and strain measurement: a review[J]. Measurement science and technology, 2009,20(6): 062001.
[57] GIACHETTI A. Matching techniques to compute image motion[J]. Image and Vision Computing, 2000, 18(3): 247-260.
[58] TONG W. An evaluation of digital image correlation criteria for strain mapping applications[J]. Strain, 2005, 41(4): 167-175.
[59] 陈亚军, 孙胜洁, 季春明. 三维数字图像相关技术 (3D DIC) 在材料形变研究中的应用进展[J]. 航空材料学报, 2017, 37(4): 90-100.
[60] PAN B, YU L, ZHANG Q. Review of single-camera stereo-digital image correlation techniquesfor full-field 3d shape and deformation measurement[J]. Science China Technological Sciences,2018, 61(1): 2-20.
[61] TANG Z, LIANG J, XIAO Z, et al. Large deformation measurement scheme for 3d digitalimage correlation method[J]. Optics and lasers in engineering, 2012, 50(2): 122-130.
[62] HU Z, XIE H, LU J, et al. Study of the performance of different subpixel image correlationmethods in 3d digital image correlation[J]. Applied optics, 2010, 49(21): 4044-4051.
[63] 柴兴华. 结合数字散斑干涉和数字图像相关的三维形变测量研究[D]. 北京: 北京信息科技大学, 2010.
[64] POOZESH P, SARRAFI A, NIEZRECKI C, et al. Extracting high frequency operating shapesfrom 3d dic measurements and phased-based motion magnified images[M]//International Digital Imaging Correlation Society. Springer, 2017: 81-83.
[65] BEBERNISS T J, EHRHARDT D A. High-speed 3d digital image correlation vibration measurement: Recent advancements and noted limitations[J]. Mechanical Systems and Signal Processing, 2017, 86: 35-48.
[66] BODEN F, KIRMSE T, STASICKI B, et al. Advanced optical in-flight measurements on deformation of wings and propeller blades[C]//SFTE (EC) Symposium Manching. 2008: 22-24.
[67] ALBERTANI R, STANFORD B, HUBNER J, et al. Aerodynamic coefficients and deformation measurements on flexible micro air vehicle wings[J]. Experimental Mechanics, 2007, 47(5):625-635.
[68] BODEN F, LANARI C, TORRES A, et al. In-flight ipct wing deformation measurements on a small aircraft[M]//Advanced In-Flight Measurement Techniques. Springer, 2013: 57-72.
[69] BODEN F, JENTINK H, PETIT C. Ipct wing deformation measurements on a large transportaircraft[M]//Advanced In-Flight Measurement Techniques. Springer, 2013: 93-115.
[70] KIRMSE T. Recalibration of a stereoscopic camera system for in-flight wing deformation measurements[J]. Measurement Science and Technology, 2016, 27(5): 054001.
[71] BAKUNOWICZ J, MEYER R. In-flight wing deformation measurements on a glider[J]. TheAeronautical Journal, 2016, 120(1234): 1917-1931.
[72] LI L G, LIANG J, GUO X, et al. Full-field wing deformation measurement scheme for inflight cantilever monoplane based on 3d digital image correlation[J]. Measurement Scienceand Technology, 2014, 25(6): 065202.
[73] 李湘萍, 刘海淙. 基于数字图像相关的机翼有限元模型试验验证[J]. 科学技术与工程,2017, 17(14): 97-104.
[74] 李湘萍, 刘海淙. 某无人机机翼压载变形有限元分析与光学测试[J]. 现代制造工程, 2018(9): 88-93.
[75] 赵凯华, 钟锡华. 光学[M]. 北京大学出版社, 1984.
[76] KAEHLER A, BRADSKI G. Learning opencv 3: computer vision in c++ with the opencvlibrary[M]. ” O’Reilly Media, Inc.”, 2016.
[77] BROWN D C. Decentering distortion of lenses[J]. Photogrammetric Engineering and RemoteSensing, 1966.
[78] HARTLEY R, ZISSERMAN A. Multiple view geometry in computer vision[M]. Cambridgeuniversity press, 2003.
[79] DUANE C B. Close-range camera calibration[J]. Photogramm. Eng, 1971, 37(8): 855-866.
[80] HARTLEY R I. Theory and practice of projective rectification[J]. International Journal ofComputer Vision, 1999, 35(2): 115-127.
[81] BOUGUET J Y, PERONA P. Camera calibration from points and lines in dual-space geometry[C]//Proc. 5th European Conf. on Computer Vision. Citeseer, 1998: 2-6.
[82] TURNER D, CROZIER P, REU P. Digital image correlation engine[R]. Sandia National Laboratories (SNL), Albuquerque, NM, and Livermore, CA …, 2015.
[83] ATKINSON D, BECKER T. A 117 line 2d digital image correlation code written in matlab[J].Remote Sensing, 2020, 12(18): 2906.
[84] PAN B, XIE H, WANG Z. Equivalence of digital image correlation criteria for pattern matching[J]. Applied optics, 2010, 49(28): 5501-5509.
[85] HOU H, ANDREWS H. Cubic splines for image interpolation and digital filtering[J]. IEEETransactions on acoustics, speech, and signal processing, 1978, 26(6): 508-517.
[86] BLABER J, ADAIR B, ANTONIOU A. Ncorr: open-source 2d digital image correlation matlab software[J]. Experimental Mechanics, 2015, 55(6): 1105-1122.
[87] GAO Y, CHENG T, SU Y, et al. High-efficiency and high-accuracy digital image correlationfor three-dimensional measurement[J]. Optics and Lasers in Engineering, 2015, 65: 73-80.
[88] LU H, CARY P. Deformation measurements by digital image correlation: implementation of asecond-order displacement gradient[J]. Experimental mechanics, 2000, 40(4): 393-400.
[89] ATKINSON D, BECKER T H. Stereo digital image correlation in matlab[J]. Applied Sciences,2021, 11(11): 4904.
[90] YANG J, BHATTACHARYA K. Augmented lagrangian digital image correlation[J]. Experimental Mechanics, 2019, 59(2): 187-205.
[91] HE B, YUAN X. A class of admm-based algorithms for three-block separable convex programming[J]. Computational Optimization and Applications, 2018, 70(3): 791-826.
[92] HARTLEY R I, STURM P. Triangulation[J]. Computer vision and image understanding, 1997,68(2): 146-157.
[93] TAUBIN G. Estimation of planar curves, surfaces, and nonplanar space curves defined by implicit equations with applications to edge and range image segmentation[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 1991, 13(11): 1115-1138.