嵌入正交权值神经网络在摄像机内外参数标定中的应用

葛动元; 姚锡凡; 向文江

doi:10.3788/OPE.20111911.2782

您当前的位置：

首页 >

文章列表页 >

嵌入正交权值神经网络在摄像机内外参数标定中的应用

信息科学 | 更新时间：2020-08-12

- 嵌入正交权值神经网络在摄像机内外参数标定中的应用
- Application of neural network with embedded orthogonal weight to calibration of camera’s intrinsic and extrinsic parameters
- 光学精密工程 2011年19卷第11期页码：2782-2790
- 作者机构：
  
  1. 华南理工大学机械与汽车工程学院,广东广州,510640
  2. 邵阳学院机械与能源工程系,湖南邵阳,422004
- 作者简介：
- 基金信息：
  
  国家自然科学基金资助项目(No.51175187)();国家863高技术研究发展计划资助项目(No.2007AA04Z111)();湖南省自然科学基金资助项目(No.0
- DOI：10.3788/OPE.20111911.2782
  中图分类号： TP183;TP391
- 收稿日期：2010-09-13，
  
  修回日期：2010-11-18，
  
  网络出版日期：2011-11-25，
  
  纸质出版日期：2011-11-25
- 稿件说明：
移动端阅览
葛动元, 姚锡凡, 向文江. 嵌入正交权值神经网络在摄像机内外参数标定中的应用[J]. 光学精密工程, 2011,19(11): 2782-2790

GE Dong-yuan, YAO Xi-fan, XIANG Wen-jiang. Application of neural network with embedded orthogonal weight to calibration of camera’s intrinsic and extrinsic parameters[J]. Editorial Office of Optics and Precision Engineering, 2011,19(11): 2782-2790
葛动元, 姚锡凡, 向文江. 嵌入正交权值神经网络在摄像机内外参数标定中的应用[J]. 光学精密工程, 2011,19(11): 2782-2790 DOI： 10.3788/OPE.20111911.2782.

GE Dong-yuan, YAO Xi-fan, XIANG Wen-jiang. Application of neural network with embedded orthogonal weight to calibration of camera’s intrinsic and extrinsic parameters[J]. Editorial Office of Optics and Precision Engineering, 2011,19(11): 2782-2790 DOI： 10.3788/OPE.20111911.2782.

摘要

针对机器视觉在某些应用场合对摄像机内外参数的需要

提出了一种基于内嵌正交权值矩阵神经网络的摄像机标定方法。首先

使设计的神经网络的权值分别与摄像机的外参数和内参数相对应

即使得所设计的神经网络与摄像机的物理模型一致。正交权值矩阵的生成在迭代中相当于遗传算法的上一次变异

系统的性能指标为由网络输出组成的矢量与对应特征点投影于图像平面的齐次坐标差值2-范数的平方。同时引进混合遗传-模拟退火算法

使得系统在达到平衡状态时可根据神经网络的权值完成摄像机内外参数的标定。实验结果表明

该方法具有较好的鲁棒性

且标定精度高

算法简洁

是摄像机内外参数标定的有效解决方案。

Abstract

According to the intrinsic and extrinsic parameters of a camera to be needed in some applications for machine vision

a method based on the neural network with embedded orthogonal weight matrix is proposed for camera calibration. Firstly

the neural network with the embedded orthogonal weight matrix is structured

whose weights are corresponding to the camera extrinsic parameters and intrinsic parameters. Thus

the structured neural network coincides with the model of camera. The generation of orthogonal weight matrix is served as the last mutation operator in the iteration of genetic algorithm

and the performance index is the square of 2-norm of the difference between vector consisting of network's outputs and homogeneous coordinate of corresponding feature point projected in image plane. Meanwhile

a hybrid genetic-simulation annealing algorithm is introduced into the solving-programming. When the system comes to the equilibrium

the intrinsic and extrinsic parameters of the camera can be obtained in the light of network's weights. The simulated experiments illustrate that the proposed algorithm is robust

and has the advantages of high calibration precision

simplicity and clarity according to real experiments. It provides an effective solution scheme for camera calibration of intrinsic and extrinsic parameters.

关键词

Keywords

references

MA S D. A Self-calibration technique for active vision system [J]. IEEE Trans. Robotics and Automation, 1996, 12(1): 114-120.[2] ZHANG Z. A flexible new technique for camera calibration [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(11):1330-1334.[3] 吴福朝, 李华, 胡占义. 基于主动视觉系统的摄像机自标定方法研究[J]. 自动化学报, 2001, 27(6):752-762. WU F CH, LI H, HU ZH Y. New active vision based camera self-calibration technique [J]. Acta Automatica Sinica. 2001, 27(6):752-762. (in Chinese)[4] 吕朝辉, 张兆杨, 安平. 基于神经网络的立体视觉摄像机标定[J]. 机械工程学报, 2003, 39(9):93-96. LV CH H, ZHANG Z Y, AN P. Camera calibration for stereo vision based on neural network [J]. Chinese Journal of Mechanical Engineering, 2003, 39(9):93-96. (in Chinese)[5] 胡峰,胡春生, 王省书,等. 成像激光雷达与摄像机外部位置关系的标定 [J]. 光学精密工程, 2011, 19(4): 938-943. HU F, HU CH SH, WANG X SH, et al.. Study of binocular vision system calibration in laser remanufacturing robot [J]. Opt. Precision Eng., 2011, 19(4): 938-943. (in Chinese)[6] 葛动元, 姚锡凡. 基于瑞利原理摄像机标定的精度研究 [J]. 计量学报, 2009, 30(1):11-15. GE D Y, YAO X F. Precision research of camera calibration based on rayleigh principle [J]. 2009, 30(1):11-15. (in Chinese)[7] 葛动元, 姚锡凡. 有侧向连接的Sanger算子在摄像机标定中的应用[J]. 光电子激光, 2010, 21(11):1722-1726. GE D Y, YAO X F. Application of the Sanger operator with lateral connection in camera calibration [J]. Journal of Optoelectronics Laser, 2010, 21(11):1722-1726. (in Chinese)[8] 霍炬, 杨卫, 杨明. 基于消隐点几何特性的摄像机自标定方法 [J]. 光学学报, 2010, 30(2):465-472. HUO J, YANG W, YANG M. A self-calibration technique based on the geometry property of the vanish point [J]. Acta Optica Sinica, 2010, 30(2):465-472. (in Chinese)[9] 许海霞, 王耀南, 袁小芳, 等. 基于矢量差分的未标定摄像机P5P 问题的求解 [J]. 自动化学报, 2009, 35(8): 1140-1144. XU H X, WANG Y N, YUAN X F, et al.. Solution to the P5P problem with un-calibrated camera based on vector difference [J]. Acta Automatica Sinica, 2009, 35(8): 1140-1144. (in Chinese)[10] 唐正宗, 梁晋, 肖振中, 等. 用于三维变形测量的数字图像相关系统 [J]. 光学精密工程, 2010, 18(10): 2244-2253. TANG Z Z, LIANG J, XIAO ZH ZH, et al.. Digital image correlation system for three-dimensional deformation measurement [J]. Opt. Precision Eng., 2010, 18(10): 2244-2253. (in Chinese)[11] YAO X F, LIAN ZH T, GE D Y,et al.. Approaches to model and control non-linear systems by RBF neural network [J]. International Journal of Innovative Computing, Information and Control, 2011, 7(2): 941-954.[12] 蔡良伟, 李霞. 基于距离测度的改进自适应遗传退火算法[J]. 模式识别与人工智能, 2005, 18(3):380-384. CAI L W, LI X. Improved adaptive genetic annealing algorithm based on distance measurement [J]. Pattern Recognition and Artificial Intelligence, 2005, 18(3):380-384. (in Chinese)[13] CHATTERJEE C. A nonlinear gauss-seidel algorithm for noncoplanar and coplanar camera calibration with convergence analysis [J]. Computer Vision and Imagine Understanding, 1997, 67 (1): 58-80.[14] 孙军华, 刘震,张广军,等. 基于柔性立体靶标的摄像机标定[J]. 光学学报, 2009, 29(12): 3433-3439. SUN J H, L ZH, ZH G J, et al.. Camera calibration based on flexible 3D target [J]. Acta Optica Sinica, 2009, 29(12): 3433- 3439.(in Chinese)

浏览量

545

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

机器视觉在工业测量中的应用与研究

摄像机标定中一种约束条件选择方法

基于机器视觉的透明软管内微量液体体积测量

基于圆卷积神经网络的粘连导电粒子检测

融合多尺度特征的蜗杆表面缺陷检测