基于线结构光的系统标定技术研究
|
摘要
基于线结构光的系统标定技术是线结构光三维测量系统在利用非接触方法快速准确获得被测三维物体的几何构型的关键技术之一。由线激光器生成,投射于被测物体表面的线结构光受到被测物体表面的调制,从而形成带有被测物体表面调制信息的线结构光。CCD摄像机捕获到带有调制信息的光条图像并对其进行相关处理,通过三维测量平台的运动关系获得CCD摄像机的数学模型。
本文对基于线结构光的系统标定技术进行了深入的研究,主要完成了以下工作:
首先,分析、讨论了结构光的选取,介绍了线结构光三维测量的基本原理并结合传统的摄像标定方法详细论证了世界坐标系、摄像机坐标系、图像坐标系和像素坐标系之间的坐标转换关系;
其次,依据线结构光三维测量方法的基本原理重点论述了本文实验用的三维测量系统的构成、参数设置和调整步骤,由此系统可以获取摄像机的内外参数及被测物体的表面信息;
最后,论述了基于线结构光系统的摄像机标定的软件设计,并进一步论述了利用Matlab软件对CCD工业相机获取的图像进行的灰度化、二值化、区域检测等图像处理过程和从图像特征点的二维坐标到实物的三维世界坐标的数据获取过程。
本论文完成了基于线结构光系统的标定研究中的一些基础性工作,论文中所作的理论分析和测量实验为今后的三维重建的进一步研究提供基础。
The technology of system calibration which based on line-structured light is one of the cruxes of line-structured light 3D measurement system when obtaining the geometry of the 3D object with the non-contact way. The line-structured light which is generated by linear laser generator projects onto the surface of object is modulated, and it carries the information of the surface of object. The light stripe which carries the information is taken in by CCD industry camera and then through the motion relationships of the 3D measurement platform,the CCD camera mode is obtained finally.
The technology of system calibration is deeply researched in this paper. The main work completed is following:
Firstly, the selection of structured light is analyzed and discussed and the 3D measuring principle is introduced. It is expounded that the coordinator transformat relationship between the world coordinate system, camera coordinate system, the image coordinates and pixel coordinates in detail with traditional camera calibration methods.
Secondly, the composition, rameter settings and the adjustment steps according to the 3D measuring principle based on the line-structured light are discussed, through which the internal parameters, external parameters and the information of object’s surface could be obtained.
Finally, software design of camera calibration system based on the line-structure light is elaborated, then the images, which are gained from CCD industry camera, are processed by using the Matlab software, such as grayscale, binaryzation, area detection and so on and the data gaining process of the two-dimensional coordinates of the image points and theirs’three dimensional world coordinate of the characteristic points is also discussed.
The present paper has completed some foundational research work of system's calibration based on line-structured light, the theoretical analysis and the measurement experiment in this paper provides a foundation for the research of 3D reconstruction.
本文对基于线结构光的系统标定技术进行了深入的研究,主要完成了以下工作:
首先,分析、讨论了结构光的选取,介绍了线结构光三维测量的基本原理并结合传统的摄像标定方法详细论证了世界坐标系、摄像机坐标系、图像坐标系和像素坐标系之间的坐标转换关系;
其次,依据线结构光三维测量方法的基本原理重点论述了本文实验用的三维测量系统的构成、参数设置和调整步骤,由此系统可以获取摄像机的内外参数及被测物体的表面信息;
最后,论述了基于线结构光系统的摄像机标定的软件设计,并进一步论述了利用Matlab软件对CCD工业相机获取的图像进行的灰度化、二值化、区域检测等图像处理过程和从图像特征点的二维坐标到实物的三维世界坐标的数据获取过程。
本论文完成了基于线结构光系统的标定研究中的一些基础性工作,论文中所作的理论分析和测量实验为今后的三维重建的进一步研究提供基础。
The technology of system calibration which based on line-structured light is one of the cruxes of line-structured light 3D measurement system when obtaining the geometry of the 3D object with the non-contact way. The line-structured light which is generated by linear laser generator projects onto the surface of object is modulated, and it carries the information of the surface of object. The light stripe which carries the information is taken in by CCD industry camera and then through the motion relationships of the 3D measurement platform,the CCD camera mode is obtained finally.
The technology of system calibration is deeply researched in this paper. The main work completed is following:
Firstly, the selection of structured light is analyzed and discussed and the 3D measuring principle is introduced. It is expounded that the coordinator transformat relationship between the world coordinate system, camera coordinate system, the image coordinates and pixel coordinates in detail with traditional camera calibration methods.
Secondly, the composition, rameter settings and the adjustment steps according to the 3D measuring principle based on the line-structured light are discussed, through which the internal parameters, external parameters and the information of object’s surface could be obtained.
Finally, software design of camera calibration system based on the line-structure light is elaborated, then the images, which are gained from CCD industry camera, are processed by using the Matlab software, such as grayscale, binaryzation, area detection and so on and the data gaining process of the two-dimensional coordinates of the image points and theirs’three dimensional world coordinate of the characteristic points is also discussed.
The present paper has completed some foundational research work of system's calibration based on line-structured light, the theoretical analysis and the measurement experiment in this paper provides a foundation for the research of 3D reconstruction.
引文
[1] D. Marr, Vision, w H. Freeman and Company, San Francisco, 1982
[2]李瑞峰,李庆喜.机器人双目视觉系统的标定与定位算法[J].哈尔滨工业大学学报,2007, 39(11): 1719-1722
[3]刁灿,王英勋,王金提,苗淼.无人机自动着陆中的机器视觉辅助技术.航空学报, 2008, 29(增): S79-S84
[4]张广军,李秀智.移动机器人足目标定方法[J].机器人, 2007, 29(3): 230-233
[5]王昭,张平.机器视觉导引系统中摄像机的标定方法.航空学报, 2008, 29(增): S204-S208
[6]王志文,郭戈.移动机器人导航技术现状与展望[J].机器人, 2003, 25(5): 470-474
[7]韩立伟,徐德,谭民.机器人无标定视觉中摄像机特性的逼近[J].控制与决策, 2007, 22(1): 1-7
[8]邱茂林,马颂德,李毅.计算机视觉中摄像机定标综述[J].自动化学报, 2000, 26(1): 43-55
[9]吴文琪,孙增圻.机器视觉中的摄像机定标方法综述[J].计算机应用研究, 2004, 2: 4-6
[10]李鹏,王军宁.摄像机标定方法综述[J].山西电子技术, 2007, 4: 77-79
[11]姜大志,郁倩,王冰洋,等.计算机视觉中的设备标定和三维图形重构综述[J].计算机工程与应用, 2001, 13: 53-55
[12]贺俊吉,张广军,杨宪铭.基于交比不变性的镜头畸变参数标定方法[J].仪器仪表学报, 2004, 25(5): 597-599
[13]郑榜贵,田炳香,段建民.基于交比不变量的摄像机标定方法[J].北京工业大学学报, 2008, 34(5): 476-480
[14]伍雪冬,蒋新华,李建兴,等.计算机视觉中传统摄像机定标方法综述[J].福建工程学院学报, 2007, 5(1): 57-61
[15]胡培成,黎宁,周建江.一种改进的基于圆环点的摄像机自标定方法[J].光电工程, 2007, 34(12): 54-60
[16]张小苗,刘肖琳,于起峰.基于等价图像的摄像机自标定方法[J].光学技术, 2008, 34(2): 217-220
[17]许海霞,王耀南,万琴,等.一种机器人手眼关系自标定方法[J].机器人, 2008, 4: 373-378
[18]傅丹,冯卫东,于起峰,等.一种摄像机自标定的线性方法[J].光电工程, 2008, 35(1): 71-74
[19]王晓华,傅卫平.一种改进的摄像机标定方法[J].东北林业大学学报, 2007, 35(6): 51-53
[20]刘金颂,原思聪,张庆阳,刘道华.双目立体视觉中的摄像机标定技术研究[J].计算机工程与应用, 2008, 44(6): 237-239
[21]孙瑾,顾宏斌.矩形模板下摄像机标定和目标定位方法研究[J].小型微型计算机系统, 2008, 9(9): 1740-1744
[22]胡钊政,谈正.利用二次曲线拟合和圆环点进行摄像机标定[J].西安交通大学学报, 2006, 40(10): 1065-1068
[23]刘元坤,苏显渝,吴庆阳.基于傅里叶条纹分析的多摄像机标定方法[J].光子学报, 2007, 36(9): 1734-1737
[24]李欣菊,祝海江,吴福朝.基于平面相似图形的摄像机标定方法[J].模式识别与人工智能, 2004, 17(4): 457-461
[25]袁野,欧宗瑛.一种基于单个神经元的摄像机标定自适应算法[J].大连理工大学学报, 2005, 45(6): 823-826
[26]蔡盛,李清安,乔彦峰.基于BP神经网络的姿态测量系统摄像机标定[J].光电子?激光, 2007, 18(7): 832-834
[27]阎平凡,张长水.人工神经网络与模拟进化计算(第二版)[M].北京:清华大学出版社, 2005
[28]Hagan M T, Demuth H B, Beale M H.神经网络设计[M].戴葵,译.北京:机械工业出版社, 2002
[29]赵清杰,孙增圻,兰丽.摄像机神经网络标定技术[J].控制与决策, 2002, 17(3): 336-339
[30]李红霞,赵新华,迟海燕.基于改进BP神经网络模型的地面沉降预测及分析[J].天津大学学报, 2009, 1(42): 60-64
[31]曹承志,张坤,郑海英.基于人工鱼群算法的BP神经网络速度辨识器[J].系统仿真学报, 2009, 4(21): 1047-1050
[32] Fu Li, Liu Zhenzhong. Camera Calibration for Linear Structured Light 3D Digital Measurement, 2010 International Conference on Intelligent Control and Information Processing (ICICIP), 352-355
[33]李小峰,李峰.单目摄像机标定方法的研究[J].计算机工程与应用, 2009, 45(15): 229-232
[34]张宇,黄亚博,焦建彬.一种基于机器视觉的圆形零件检测技术.计算机工程, 2008, 34(19): 185-186
[35]杜立彬,高晓辉,夏进军,等.一种新型激光三维扫描测量仪的研制[J].激光技术, 2005, 29(4): 366-369
[2]李瑞峰,李庆喜.机器人双目视觉系统的标定与定位算法[J].哈尔滨工业大学学报,2007, 39(11): 1719-1722
[3]刁灿,王英勋,王金提,苗淼.无人机自动着陆中的机器视觉辅助技术.航空学报, 2008, 29(增): S79-S84
[4]张广军,李秀智.移动机器人足目标定方法[J].机器人, 2007, 29(3): 230-233
[5]王昭,张平.机器视觉导引系统中摄像机的标定方法.航空学报, 2008, 29(增): S204-S208
[6]王志文,郭戈.移动机器人导航技术现状与展望[J].机器人, 2003, 25(5): 470-474
[7]韩立伟,徐德,谭民.机器人无标定视觉中摄像机特性的逼近[J].控制与决策, 2007, 22(1): 1-7
[8]邱茂林,马颂德,李毅.计算机视觉中摄像机定标综述[J].自动化学报, 2000, 26(1): 43-55
[9]吴文琪,孙增圻.机器视觉中的摄像机定标方法综述[J].计算机应用研究, 2004, 2: 4-6
[10]李鹏,王军宁.摄像机标定方法综述[J].山西电子技术, 2007, 4: 77-79
[11]姜大志,郁倩,王冰洋,等.计算机视觉中的设备标定和三维图形重构综述[J].计算机工程与应用, 2001, 13: 53-55
[12]贺俊吉,张广军,杨宪铭.基于交比不变性的镜头畸变参数标定方法[J].仪器仪表学报, 2004, 25(5): 597-599
[13]郑榜贵,田炳香,段建民.基于交比不变量的摄像机标定方法[J].北京工业大学学报, 2008, 34(5): 476-480
[14]伍雪冬,蒋新华,李建兴,等.计算机视觉中传统摄像机定标方法综述[J].福建工程学院学报, 2007, 5(1): 57-61
[15]胡培成,黎宁,周建江.一种改进的基于圆环点的摄像机自标定方法[J].光电工程, 2007, 34(12): 54-60
[16]张小苗,刘肖琳,于起峰.基于等价图像的摄像机自标定方法[J].光学技术, 2008, 34(2): 217-220
[17]许海霞,王耀南,万琴,等.一种机器人手眼关系自标定方法[J].机器人, 2008, 4: 373-378
[18]傅丹,冯卫东,于起峰,等.一种摄像机自标定的线性方法[J].光电工程, 2008, 35(1): 71-74
[19]王晓华,傅卫平.一种改进的摄像机标定方法[J].东北林业大学学报, 2007, 35(6): 51-53
[20]刘金颂,原思聪,张庆阳,刘道华.双目立体视觉中的摄像机标定技术研究[J].计算机工程与应用, 2008, 44(6): 237-239
[21]孙瑾,顾宏斌.矩形模板下摄像机标定和目标定位方法研究[J].小型微型计算机系统, 2008, 9(9): 1740-1744
[22]胡钊政,谈正.利用二次曲线拟合和圆环点进行摄像机标定[J].西安交通大学学报, 2006, 40(10): 1065-1068
[23]刘元坤,苏显渝,吴庆阳.基于傅里叶条纹分析的多摄像机标定方法[J].光子学报, 2007, 36(9): 1734-1737
[24]李欣菊,祝海江,吴福朝.基于平面相似图形的摄像机标定方法[J].模式识别与人工智能, 2004, 17(4): 457-461
[25]袁野,欧宗瑛.一种基于单个神经元的摄像机标定自适应算法[J].大连理工大学学报, 2005, 45(6): 823-826
[26]蔡盛,李清安,乔彦峰.基于BP神经网络的姿态测量系统摄像机标定[J].光电子?激光, 2007, 18(7): 832-834
[27]阎平凡,张长水.人工神经网络与模拟进化计算(第二版)[M].北京:清华大学出版社, 2005
[28]Hagan M T, Demuth H B, Beale M H.神经网络设计[M].戴葵,译.北京:机械工业出版社, 2002
[29]赵清杰,孙增圻,兰丽.摄像机神经网络标定技术[J].控制与决策, 2002, 17(3): 336-339
[30]李红霞,赵新华,迟海燕.基于改进BP神经网络模型的地面沉降预测及分析[J].天津大学学报, 2009, 1(42): 60-64
[31]曹承志,张坤,郑海英.基于人工鱼群算法的BP神经网络速度辨识器[J].系统仿真学报, 2009, 4(21): 1047-1050
[32] Fu Li, Liu Zhenzhong. Camera Calibration for Linear Structured Light 3D Digital Measurement, 2010 International Conference on Intelligent Control and Information Processing (ICICIP), 352-355
[33]李小峰,李峰.单目摄像机标定方法的研究[J].计算机工程与应用, 2009, 45(15): 229-232
[34]张宇,黄亚博,焦建彬.一种基于机器视觉的圆形零件检测技术.计算机工程, 2008, 34(19): 185-186
[35]杜立彬,高晓辉,夏进军,等.一种新型激光三维扫描测量仪的研制[J].激光技术, 2005, 29(4): 366-369