飞机零部件数字化检测系统中的关键技术研究
|
摘要
在飞机制造中,复杂轮廓表面的加工与测量技术水平已成为一个国家航空科技水平的标志之一。虽然包含自由曲面的飞机零部件与模具的数控加工技术取得了显著进展,但其形状检测和质量保证一直是一个自动化程度很低的过程。本文对飞机零部件的数字化检测技术进行了讨论,提出了在CMM精确控制下的线结构光与计算机视觉相集成检测方案,并对其中的关键技术并进行了深入研究。
主要研究内容包括:
1.根据飞机零部件的结构和检测要求,构建了多传感器检测规划模型,分配CMM接触式测量、线结构光视觉测量系统和双目视觉测量系统的检测任务,获得完备的零件外形数据。建立复杂形状质量评价方法与指标,将所获测量数据与零件数字化模型进行比较,来完成飞机典型部件与零件的质量检测。
2.实现了CCD摄像机、图像采集卡、线结构光激光传感器以及供电电源等设备与计算机系统的连接,编制了获取图像数据的接口程序,实现了摄像机的拍摄参数的配置,图像拍摄模式、帧速率以及图像文件的保存等功能。
3.研究了图像处理的算法,包括图像数据的随机噪声的剔除,激光光条图像的分割以及光条中心点的提取。在随机噪声的剔除算法中,使用一种改进的基于5×5滤波窗口的中值滤波技术,大大简化了后续的图像处理过程,并降低了条纹中心提取的误判率,从而提高了测量数据的可靠性。
4.在建立的基于小孔成像的摄像机成像数学模型中,考虑了径向畸变误差对测量数据的影响,在此基础上实现了一种基于两步法的摄像机标定方法。该方法只需对标定模板进行一次拍摄,即可获得摄像机内外参数。通过投影面上三个不共线点的坐标来确定结构光平面方程,结合前面获得的标定结果,即可完成线结构激光传感器的标定。
5.使用VC+6.0开发了实验系统。通过获取的计算机图像信息得到了测量数据的三维信息,以此验证系统的测量精度。
In aircraft manufacturing, techniques of processing and detecting complex surfaces have become a symbol of scientific and technological level of a country. Although CNC machining technology of aircraft parts and dies which contain free-form surfaces have made remarkable progress, automation degree of detection and quality assurance of above them has always been very low. In this paper, digital detection technology of aircraft parts and components is discussed, detection scheme which integrate the line structured light sensor and machine vision under CMM control is put forward, and key technologies of above technique are researched in depth.
Main works of us focus on following parts:
1.According to structure feature and detecting requirement of the aircraft parts and components, a multiple sensor detection model is set up, which distributes detection tasks among CMM contact measurement, line structured light visual detection system and binocular vision measurement system in order to get complete part appearance data. Quality evaluation method and indicator of complicated shape is also established, which can be used as guidelines to compare measuring data with digital produce model, thus quality detection of typical aircraft components and parts can be realize.
2. Connection between devices such as CCD camera, IEEE1394 card, line-structure laser senor, power supply with computer system is established, interface programs used to getting image is coded, which can set parameters, customize model of shoot and frame-ratio, and save the image files.
3. Image process algorithms are studied, which include eliminating random noise of image data, devising laser-stripes image and getting center points from these stripes. An advanced median-filter arithmetic based on 5×5 windows is used to eliminate random noise, thus simplify latter image process, reduces misjudgments in the process of getting center points and increasing the reliability of measurement data.
4. Mathematical model of CCD camera based on pin-hole imaging theory set up by us has considered the influence of radial distortion error towards measurement data, based on above theory a calibration method which have two main steps is provided. Only one shoot to calibration model is needed to get all the inner and outer parameters of the camera. Equation of projection plane is acquired by three non-linear points, combined with above results the calibration of line structured light senor can be deduced.
5. An experimental inspection system is developed with VC++6.0. Through this system, 3D digital information of measuring data is derived from image information, and then precision of this system is verified by these data.
主要研究内容包括:
1.根据飞机零部件的结构和检测要求,构建了多传感器检测规划模型,分配CMM接触式测量、线结构光视觉测量系统和双目视觉测量系统的检测任务,获得完备的零件外形数据。建立复杂形状质量评价方法与指标,将所获测量数据与零件数字化模型进行比较,来完成飞机典型部件与零件的质量检测。
2.实现了CCD摄像机、图像采集卡、线结构光激光传感器以及供电电源等设备与计算机系统的连接,编制了获取图像数据的接口程序,实现了摄像机的拍摄参数的配置,图像拍摄模式、帧速率以及图像文件的保存等功能。
3.研究了图像处理的算法,包括图像数据的随机噪声的剔除,激光光条图像的分割以及光条中心点的提取。在随机噪声的剔除算法中,使用一种改进的基于5×5滤波窗口的中值滤波技术,大大简化了后续的图像处理过程,并降低了条纹中心提取的误判率,从而提高了测量数据的可靠性。
4.在建立的基于小孔成像的摄像机成像数学模型中,考虑了径向畸变误差对测量数据的影响,在此基础上实现了一种基于两步法的摄像机标定方法。该方法只需对标定模板进行一次拍摄,即可获得摄像机内外参数。通过投影面上三个不共线点的坐标来确定结构光平面方程,结合前面获得的标定结果,即可完成线结构激光传感器的标定。
5.使用VC+6.0开发了实验系统。通过获取的计算机图像信息得到了测量数据的三维信息,以此验证系统的测量精度。
In aircraft manufacturing, techniques of processing and detecting complex surfaces have become a symbol of scientific and technological level of a country. Although CNC machining technology of aircraft parts and dies which contain free-form surfaces have made remarkable progress, automation degree of detection and quality assurance of above them has always been very low. In this paper, digital detection technology of aircraft parts and components is discussed, detection scheme which integrate the line structured light sensor and machine vision under CMM control is put forward, and key technologies of above technique are researched in depth.
Main works of us focus on following parts:
1.According to structure feature and detecting requirement of the aircraft parts and components, a multiple sensor detection model is set up, which distributes detection tasks among CMM contact measurement, line structured light visual detection system and binocular vision measurement system in order to get complete part appearance data. Quality evaluation method and indicator of complicated shape is also established, which can be used as guidelines to compare measuring data with digital produce model, thus quality detection of typical aircraft components and parts can be realize.
2. Connection between devices such as CCD camera, IEEE1394 card, line-structure laser senor, power supply with computer system is established, interface programs used to getting image is coded, which can set parameters, customize model of shoot and frame-ratio, and save the image files.
3. Image process algorithms are studied, which include eliminating random noise of image data, devising laser-stripes image and getting center points from these stripes. An advanced median-filter arithmetic based on 5×5 windows is used to eliminate random noise, thus simplify latter image process, reduces misjudgments in the process of getting center points and increasing the reliability of measurement data.
4. Mathematical model of CCD camera based on pin-hole imaging theory set up by us has considered the influence of radial distortion error towards measurement data, based on above theory a calibration method which have two main steps is provided. Only one shoot to calibration model is needed to get all the inner and outer parameters of the camera. Equation of projection plane is acquired by three non-linear points, combined with above results the calibration of line structured light senor can be deduced.
5. An experimental inspection system is developed with VC++6.0. Through this system, 3D digital information of measuring data is derived from image information, and then precision of this system is verified by these data.
引文
[1]范玉清.现代飞机制造技术[M].北京:北京航空航天大学出版社. 2001, 6
[2]金涛,城建良,童水光.逆向工程技术研究进展[J].中国机械工程, 2002, 13(16): 1430-1437
[3] C. K. Song, S.W. Kim. Reverse Engineering: Autonomous Digitization of Free-formed Surfaces on a CNC Coordinate Measuring Machine[J]. International Journal of Machine Tools and Manufacture, 1997, 37(7): 1041-1051
[4] Varady T., Martin R., Cox J. Reverse Engineering of Geometric Model-an Introduction[J]. Computer Aided Design, 1997, 29(4): 255-268
[5] Seiler A., Balendram V., Sivayoganathan K, et al. Reverse Engineering from Uni-Directional CMM Scan Data[J]. International Journal of Advanced Manufacturing Technology, 1996, 11: 276-284
[6] S. F.EI Hakim, Fujiyoshi H., Patil R. Moving target classification and tracking from real-time video[C]. In: Proc IEEE Workshopon Applictions of Computer Vision, Princeton. 1998, 12: 8-14
[7]闵新力,万德安,张剑.CCD双目视觉测量系统结构参数设置的理论研究[J].机械设计与制造,2001, 25(3): 53-56
[8] Wen-Chin Tai and Ming Chang.Noncontact Profilometric Measurement of Large Form Parts[J]. Optical Engineering, 1996, 35(9): 2730-2735
[9]王学军.激光自动变位测量汽车三维形状的研究[J].中国激光, 1998, 25(1): 1051-1055
[10] Jianlin Yang, Nien-Lung Lee, C.HMenq.Application of Computer Visionin Reverse Engineering for 3D Coordinate Acquisition[J]. Concurrent Product and Process Engineering, MSE, 1995, 1(85): 43-156
[11]贾波,苏显渝,郭履容.采用激光光刀的叶片三维面型测量方法[J].中国激光, 1992, 19(4): 271-275
[12] Colln Bradley, Geofrey W. Vickers. Free-form Surface Reconstruction for Machine Vision Rapid Prototyping[J]. Optical Engineering, 1993, 2(9): 2191-2200
[13] Milroy M.J., Geffrey W.Vickers, et al.Automated Laser Scanning Based Orthogonal CrossSection[J] .Machine vision and Application, 1996, 9: 106-118
[14] Chenggang Che, JunNi.A Generic Coordinate Transformation Uncertainty Assessment Approach and Its Application in Machine Vision Metrology[J]. International Journal of Machine Tools & Manufacture, 1998, 38: 1241-1256
[15]王春和,邹定海,叶声华.三维视觉检测与结构光视觉标定[J].仪器仪表学报, 1994, 15(2): 119-124
[16]赵明淘.层析三维成像测量中图象处理技术的研究[D].西安:西安交通大学机械学院, 1998
[17]李礼夫,钟先信,陈愚.三维曲面轮廓的非接触式现代测量[J].实用测试技术, 1996, 1: 22-26
[18]王平江,陈吉红,李作清,周济.参数曲面形状误差计算迭代逼近法[J].华中理工大学学报, 1997, 25(3): 1-4
[19]周长发.精通Visual C++图象处理编程[M].北京:电子工业出版社, 2006.
[20]徐静.基于非线性模型的摄像机标定技术研究[J].现代电子技术, 2008, 31(12): 153-155
[21]李凯,王鉴,韩焱.基于空间透视相关点的摄像机标定方法[J].中北大学学报, 2007, 28: 135-138
[22]田原嫄,张云辉,谭庆昌.CCD摄像机标定的研究[J] .微计算机信息, 2008, 24(15): 206-207
[23]李中伟,王从军,史玉升.3D测量系统中的高精度摄像机标定算法[J].光电工程, 2008, 35(4): 58-63
[24]胡占义,吴福朝.基于主动视觉的计算机标定方法[J].计算机学报, 2002, 25(11): 1149-1156
[25]雷成,吴福朝.Kruppa方程与摄像机标定[J].自动化学报, 2001, 7(5): 621-630
[26]陈恪.机器视觉技术在测量中的应用[J].中国计量,2008,9: 61-62
[27]罗明.多传感器机器视觉测量系统的研究及应用[D].天津:天津大学, 1996
[28] Voshida, K. and Hirose, S. Laser Triangulation Range Finder Available under Direct Sunlight, Proc. of the 1988 IEEE Int. Cord, on Robotics and Automation, 1988, Vol.3: 1702-1708
[29] Zhang, J. X. and Djordjevich, A Study on Laser Stripe Sensor[J]. Sensors and Actuators, 1999, Vol 72: 224-228
[30] W. Faig, Calibration of close-range photo grammetry system: mathematical formulation. Photogrammetric Eng. Remote Sensing, 1975, 41912: 1479-1486
[31] Tsai R. Y., Aversatile camera calibration technique for high-accuracy 3D machine vision metrology using off-the-shelf TV camera and lenses[J]. IEEE Journal of Automation, 1987, V3(4): 323-334.
[32] Tsai R.Y. An efficient and accurate camera calibration technique for 3D machine vision, Proc. CVPR'86. New York: IEEE 1986, 364-374
[33] Ju Yang Weng, Paul C. Camera calibration with distortion models and accuracy evaluation[J]. IEEE Transactions on PAMI, 1992, 14(10): 965-980
[34] ZHANG Zheng-you. A flexible new technique for camera calibration [R]. USA: Microsoft Corporation, 1998.
[35]胡占义,吴福朝.基于主动视觉摄像机标定方法[J] .计算机学报,2002, 25(11): 1149-1156
[36]雷成,胡占义,吴福朝.一种新的基于Kruppa方程的摄像机自标定方法[J].计算机学报, 2003, 26(5): 587-597
[37]胡占义,吴福朝.基于主动视觉摄像机标定方法[J].计算机学报,2002, 26(5): 1149-1156
[38] Faugers O. D., Toscani G., The Calibration Problem for 3D computer vision[C]. Proc of IEEE Conf. of Computer Vision and Pattern Recognition
[39] Faugeras O.D., Maybank S. Motion from point matches: multiplicity of solutions Int. [J] .Computer Vision, 1990, 4: 225-246
[40] Maybank SJ, Faugeras O.D.A theory of self-calibration of a moving camera[J]. Computer Vision, 1992, 8(2): 123-151
[41] S.D.Ma. A Self-calibration technique for active vision system[R]. IEEE trans. Robotics and Automation, 1996, 12(1): 114-120
[42] R. I. Hartley.Estimation of relative camera position for uncalibrated camera[M]. Proc of the ECCV'92.Italy: Santa Margherita Ligure, 1992: 379-387
[43]周士侃,刑渊.用于反求工程中的CCD摄像机图像中心、比例因子的标定[J].计测技术, 2004,(6): 11-13
[44]高立志,方勇,林志航.立体视觉测量中摄像机标定的新技术[J].机械科学与技术, 1998, 17(5): 808-811
[45]张健新,段发阶,叶声华.简便的高精度摄像机标定技术[J].仪器仪表学报,1999, 20(2): 193-196
[2]金涛,城建良,童水光.逆向工程技术研究进展[J].中国机械工程, 2002, 13(16): 1430-1437
[3] C. K. Song, S.W. Kim. Reverse Engineering: Autonomous Digitization of Free-formed Surfaces on a CNC Coordinate Measuring Machine[J]. International Journal of Machine Tools and Manufacture, 1997, 37(7): 1041-1051
[4] Varady T., Martin R., Cox J. Reverse Engineering of Geometric Model-an Introduction[J]. Computer Aided Design, 1997, 29(4): 255-268
[5] Seiler A., Balendram V., Sivayoganathan K, et al. Reverse Engineering from Uni-Directional CMM Scan Data[J]. International Journal of Advanced Manufacturing Technology, 1996, 11: 276-284
[6] S. F.EI Hakim, Fujiyoshi H., Patil R. Moving target classification and tracking from real-time video[C]. In: Proc IEEE Workshopon Applictions of Computer Vision, Princeton. 1998, 12: 8-14
[7]闵新力,万德安,张剑.CCD双目视觉测量系统结构参数设置的理论研究[J].机械设计与制造,2001, 25(3): 53-56
[8] Wen-Chin Tai and Ming Chang.Noncontact Profilometric Measurement of Large Form Parts[J]. Optical Engineering, 1996, 35(9): 2730-2735
[9]王学军.激光自动变位测量汽车三维形状的研究[J].中国激光, 1998, 25(1): 1051-1055
[10] Jianlin Yang, Nien-Lung Lee, C.HMenq.Application of Computer Visionin Reverse Engineering for 3D Coordinate Acquisition[J]. Concurrent Product and Process Engineering, MSE, 1995, 1(85): 43-156
[11]贾波,苏显渝,郭履容.采用激光光刀的叶片三维面型测量方法[J].中国激光, 1992, 19(4): 271-275
[12] Colln Bradley, Geofrey W. Vickers. Free-form Surface Reconstruction for Machine Vision Rapid Prototyping[J]. Optical Engineering, 1993, 2(9): 2191-2200
[13] Milroy M.J., Geffrey W.Vickers, et al.Automated Laser Scanning Based Orthogonal CrossSection[J] .Machine vision and Application, 1996, 9: 106-118
[14] Chenggang Che, JunNi.A Generic Coordinate Transformation Uncertainty Assessment Approach and Its Application in Machine Vision Metrology[J]. International Journal of Machine Tools & Manufacture, 1998, 38: 1241-1256
[15]王春和,邹定海,叶声华.三维视觉检测与结构光视觉标定[J].仪器仪表学报, 1994, 15(2): 119-124
[16]赵明淘.层析三维成像测量中图象处理技术的研究[D].西安:西安交通大学机械学院, 1998
[17]李礼夫,钟先信,陈愚.三维曲面轮廓的非接触式现代测量[J].实用测试技术, 1996, 1: 22-26
[18]王平江,陈吉红,李作清,周济.参数曲面形状误差计算迭代逼近法[J].华中理工大学学报, 1997, 25(3): 1-4
[19]周长发.精通Visual C++图象处理编程[M].北京:电子工业出版社, 2006.
[20]徐静.基于非线性模型的摄像机标定技术研究[J].现代电子技术, 2008, 31(12): 153-155
[21]李凯,王鉴,韩焱.基于空间透视相关点的摄像机标定方法[J].中北大学学报, 2007, 28: 135-138
[22]田原嫄,张云辉,谭庆昌.CCD摄像机标定的研究[J] .微计算机信息, 2008, 24(15): 206-207
[23]李中伟,王从军,史玉升.3D测量系统中的高精度摄像机标定算法[J].光电工程, 2008, 35(4): 58-63
[24]胡占义,吴福朝.基于主动视觉的计算机标定方法[J].计算机学报, 2002, 25(11): 1149-1156
[25]雷成,吴福朝.Kruppa方程与摄像机标定[J].自动化学报, 2001, 7(5): 621-630
[26]陈恪.机器视觉技术在测量中的应用[J].中国计量,2008,9: 61-62
[27]罗明.多传感器机器视觉测量系统的研究及应用[D].天津:天津大学, 1996
[28] Voshida, K. and Hirose, S. Laser Triangulation Range Finder Available under Direct Sunlight, Proc. of the 1988 IEEE Int. Cord, on Robotics and Automation, 1988, Vol.3: 1702-1708
[29] Zhang, J. X. and Djordjevich, A Study on Laser Stripe Sensor[J]. Sensors and Actuators, 1999, Vol 72: 224-228
[30] W. Faig, Calibration of close-range photo grammetry system: mathematical formulation. Photogrammetric Eng. Remote Sensing, 1975, 41912: 1479-1486
[31] Tsai R. Y., Aversatile camera calibration technique for high-accuracy 3D machine vision metrology using off-the-shelf TV camera and lenses[J]. IEEE Journal of Automation, 1987, V3(4): 323-334.
[32] Tsai R.Y. An efficient and accurate camera calibration technique for 3D machine vision, Proc. CVPR'86. New York: IEEE 1986, 364-374
[33] Ju Yang Weng, Paul C. Camera calibration with distortion models and accuracy evaluation[J]. IEEE Transactions on PAMI, 1992, 14(10): 965-980
[34] ZHANG Zheng-you. A flexible new technique for camera calibration [R]. USA: Microsoft Corporation, 1998.
[35]胡占义,吴福朝.基于主动视觉摄像机标定方法[J] .计算机学报,2002, 25(11): 1149-1156
[36]雷成,胡占义,吴福朝.一种新的基于Kruppa方程的摄像机自标定方法[J].计算机学报, 2003, 26(5): 587-597
[37]胡占义,吴福朝.基于主动视觉摄像机标定方法[J].计算机学报,2002, 26(5): 1149-1156
[38] Faugers O. D., Toscani G., The Calibration Problem for 3D computer vision[C]. Proc of IEEE Conf. of Computer Vision and Pattern Recognition
[39] Faugeras O.D., Maybank S. Motion from point matches: multiplicity of solutions Int. [J] .Computer Vision, 1990, 4: 225-246
[40] Maybank SJ, Faugeras O.D.A theory of self-calibration of a moving camera[J]. Computer Vision, 1992, 8(2): 123-151
[41] S.D.Ma. A Self-calibration technique for active vision system[R]. IEEE trans. Robotics and Automation, 1996, 12(1): 114-120
[42] R. I. Hartley.Estimation of relative camera position for uncalibrated camera[M]. Proc of the ECCV'92.Italy: Santa Margherita Ligure, 1992: 379-387
[43]周士侃,刑渊.用于反求工程中的CCD摄像机图像中心、比例因子的标定[J].计测技术, 2004,(6): 11-13
[44]高立志,方勇,林志航.立体视觉测量中摄像机标定的新技术[J].机械科学与技术, 1998, 17(5): 808-811
[45]张健新,段发阶,叶声华.简便的高精度摄像机标定技术[J].仪器仪表学报,1999, 20(2): 193-196