期刊文献+

基于LCD和改进棋盘格模板的摄像机标定 被引量:8

Camera calibration based on LCD and improved checkerboard pattern
下载PDF
导出
摘要 为了克服传统棋盘格模板标定点检测算法对于遮挡、视角、照明和畸变的局限性,提出一种利用基准点标记代替传统棋盘格黑白方块的改进模板,同时提出模板标定点的相应全自动识别算法。为了提高标定模板的质量,标定模板通过LCD(liquid crystal display)显示。因为LCD具有极高的几何精度和纯平面性,从而提高标定精度。传统摄像机标定同时计算针孔模型和透镜畸变模型,从而2种模型耦合在一起,所得到的标定结果仅对训练数据是有效的,而算法对于新数据的标定误差增加。为了克服2种模型的耦合,提出一种分离地标定2种模型的方法。算法利用模板的射影不变量约束,即交比和直线的射影为直线,求解带有畸变的标定点的校正的坐标,然后线性地求解2种模型。实验结果表明算法稳定且精度高。 In order to overcome the shortcomings that the identification algorithms of traditional checkerboard pattern calibration points are limited by occlusion, viewpoint, illumination and lens distortion, an improved checkerboard pattern template that replaces the black and white squares of traditional checkerboard pattern with fiducial markers is designed, and corresponding fully-automatic identification algorithm of the calibration points is proposed. In order to improve the quality of calibration template, the improved checkerboard pattern is displayed on screen using a LCD (liquid crystal display) display as the calibration template. Because LCD has very high geometric accuracy and pure planarity, so the calibration accuracy is improved. A lot of traditional camera calibration methods compute the pinhole and lens distortion models at the same time, so the two models are coupled; and the computed calibration results are valid only for the training data and their calibration errors increase for new data. In order to overcome the coupling of the two models, a calibration method is proposed, which calibrates the two models separately. The algo- rithm uses the geometric invariant of the calibration template that is the cross-ration invariability, i.e. the projection of a straight line in the space to the camera is a line, to compute the undistorted coordinates of distorted calibration points, and then compute the two models linearly. Experiment results show that the proposed algorithm is stable and has high calibration accuracy.
出处 《仪器仪表学报》 EI CAS CSCD 北大核心 2012年第7期1541-1548,共8页 Chinese Journal of Scientific Instrument
基金 黑龙江省自然科学基金(AF200921)资助项目
关键词 摄像机标定 LCD 改进的棋盘格模板 射影不变量 camera calibration liquid crystal display(LCD) improved checkerboard pattern geometric invariant
  • 相关文献

参考文献19

  • 1HUM C, CHANG M H, WU J L, et al. Robust camera calibration and player tracking in broadcast basketball video[J]. IEEE Transactions on Multimedia, 2011, 13 (2) : 266-279. 被引量:1
  • 2李勃,陈启美.基于监控视频的运动车辆行为分析算法[J].仪器仪表学报,2006,27(z3):2118-2120. 被引量:13
  • 3LEE S B, SHIN I Y, HO Y S. Gaze-corrected view gen- eration using stereo camera system for immersive videoco- nferencing[ J ]. IEEE Transactions on Consumer Elec- tronics, 2011, 57 ( 3 ) : 1033-1040. 被引量:1
  • 4ZHANG J, WU Y, LIU W, et al. Novel approach to po- sition and orientation estimation in vision-based UAV navigation[ J ]. IEEE Transactions on Aerospace and E- lectronic Systems, 2010, 46(2): 687-700. 被引量:1
  • 5ZHANG ZH Y. A flexible new technique for camera cali- bration[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence ,2000,22( 11 ) :1330-1334. 被引量:1
  • 6ZHAN S, RONALD C. Use of LCD panel for calibrating structured-light-based range sensing system [ J ]. IEEE Transactions on Instrumentation and Measurement, 2008, 57( 11 ) : 2623-2630. 被引量:1
  • 7BRADSKI G, KAEHLER A. Learning OpenCV [ M]. Sebastopol: O'Reilly Media, Inc. , 2008: 370-430. 被引量:1
  • 8FIALA M. Designing highly reliable fiducial markers [ J ]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010,32(7 ) : 1317- 1324. 被引量:1
  • 9YOON S R, SEO S, HUANG M L, et al. Multi-proces- sor based CRC computation scheme for high-speed wire- less LAN design [ J ]. Electronics Letters, 2010,46 (11 ) : 800 -802. 被引量:1
  • 10MARIN C, BOUCHIREB K, KIEFFER M, et al. Joint exploitation of residual source information and MAC layer CRC redundancy for robust video decoding [ J ]. IEEE Transactions on Wireless Communications, 2010, 9(7) : 2165-2175. 被引量:1

二级参考文献49

  • 1薛振框,李少远.MIMO非线性系统的多模型建模方法[J].电子学报,2005,33(1):52-56. 被引量:18
  • 2熊伟丽,徐保国.基于PSO的SVR参数优化选择方法研究[J].系统仿真学报,2006,18(9):2442-2445. 被引量:66
  • 3郭玉波,姚郁,遆晓光.一种改进的亚像素算法[J].光电工程,2006,33(10):137-140. 被引量:26
  • 4[1]Yuan Xiaojing,Sun Zehang,Yaakov Varol,and George Bebis.A Distributed Visual Surveillance System[C].in IEEE Conference on Advanced Video and Signal Based Surveillance,2003. 被引量:1
  • 5[2]K.Shunsuke.Classification of traffic events based on the spatio-temporal MRF model and the bayesian network[C].9th World Congress on ITS,Chicago,Oct.2002. 被引量:1
  • 6[3]Yu Bin,A.K.Jain.Lane boundary detection using a multiresolution hough transform[C].1997 International Conference on Image Processing (ICIP ' 97) 3-Volume Set-Volume 2 p.748,1997 Washington,DC. 被引量:1
  • 7[4]M.Bertozzi,A.Broggi,GOLD:A parallel real-time stereo vision system for generic obstacle and lane detection[J].IEEE Trans.Image Processing,1998,7:62-81. 被引量:1
  • 8DUNHAM.数据挖掘教程[M].北京:清华大学出版社,2003. 被引量:1
  • 9牛奔.粒子群优化算法[M].北京:冶金工业出版社,2009. 被引量:1
  • 10CANU S, SMOLA A. Kernel methods and the exponential family[ J ]. Neurocomputing, 2006,69:714-720. 被引量:1

共引文献92

同被引文献107

引证文献8

二级引证文献44

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部