单目立体视觉系统的棱镜位置估计

doi:10.12068/j.issn.1005-3026.2015.06.002

东北大学学报:自然科学版 ›› 2015, Vol. 36 ›› Issue (6): 765-768.DOI: 10.12068/j.issn.1005-3026.2015.06.002

单目立体视觉系统的棱镜位置估计

崔笑宇，赵越，范群安，魏永涛

(东北大学中荷生物医学与信息工程学院，辽宁沈阳110819)

收稿日期:2014-04-15 修回日期:2014-04-15 出版日期:2015-06-15 发布日期:2015-06-11
通讯作者: 崔笑宇
作者简介:崔笑宇(1984-)，男，辽宁丹东人，东北大学讲师，博士；赵越(1966-)，女，辽宁沈阳人，东北大学教授.
基金资助:
国家自然科学基金青年基金资助项目(61100022)；辽宁省教育厅一般项目(L2014086)；中央高校基本科研业务费青年教师科研启动基金资助项目(N130319002)；沈阳市科技基金资助项目(F13-317-6-17)

Position Estimation for Single-Lens Stereovision System Using Prism

CUI Xiao-yu， ZHAO Yue， FAN Qun-an， WEI Yong-tao

Sino-Dutch Biomedical and Information Engineering School， Northeastern University， Shenyang 110819， China.

Received:2014-04-15 Revised:2014-04-15 Online:2015-06-15 Published:2015-06-11
Contact: CUI Xiao-yu
About author:-
Supported by:
-

摘要/Abstract

摘要： 为了完善基于棱镜的单相机立体视觉系统的理论模型，提高该系统三维重建精度，提出一种基于几何光学的棱镜参数化简方法和棱镜位置估计方法.首先推导出基于棱镜的单相机立体视觉系统透视投影变换模型；之后利用向量的旋转变换和点的位置关系将表示棱镜位置的参数化简为6个；最后使用非线性最优化方法，通过单相机标定过程中得到的一组点的对应关系，解出这6个参数，继而得到较为精确的棱镜位置.实验结果表明:提出的方法重投影平均误差小于0.5个像素，三维点重建平均误差为2.22mm.本方法可以应用于匹配过程中极线的计算，与现有方法相比，三维重建精度提高2至3倍.

关键词: 计算机视觉, 位置估计, 几何光学, 棱镜, 相机标定

Abstract: In order to improve the theoretical model of prism based single-lens stereovision system， and enhance the precision of three-dimensional reconstruction， a parameter simplifying method for prism as well as a position estimation method that based on geometrical optics were proposed. First， a perspective projection model for prism based single-lens stereovision system was derived. Then， the parameters of prism which could express its position were decreased to six by making use of rotational vector and the relationship between points. At last， these six parameters could be solved out as a result of nonlinear optimization， and an accurate position of prism could be obtained. Experimental results indicated that the average reprojection error of proposed method was less than 0.5 pixel， the average error for three dimensional reconstruction was 2.22mm. This method could be used for calculating the epipolar geometry of stereo matching. It could increase the precision of three-dimensional reconstruction up to 2~3 times.

Key words: computer vision, position estimation, geometrical optics, prism, camera calibration

中图分类号:

TP11

崔笑宇，赵越，范群安，魏永涛. 单目立体视觉系统的棱镜位置估计[J]. 东北大学学报:自然科学版, 2015, 36(6): 765-768.

CUI Xiao-yu， ZHAO Yue， FAN Qun-an， WEI Yong-tao. Position Estimation for Single-Lens Stereovision System Using Prism[J]. Journal of Northeastern University Natural Science, 2015, 36(6): 765-768.

参考文献

[1]Gao Q，Wang H P，Wang J J.A single camera volumetric particle image velocimetry and its application［J］.Science China Technological Science，2012，55(9):2501-2510.
[2]Genovese K，Casaletto L，Rayas J A，et al.Stereo-digital image correlation measurements with a single camera using a biprism［J］.Lasers Engineer，2013，51(2):78-85.
[3]Cui X Y，Lim K B.A prism-based single-lens stereovision system:from trinocular to multi-ocular［J］.Image and Vision Computing，2007，25(11):1725-1736.
[4]Wu L F，Zhu J G，Xie H M.A modified virtual point model of the 3D DIC technique using a single camera and a bi-prism［J］.Measurement Science & Technology，2014，25(11):283-287.
[5]Erol A，Bebis G，Nicoescu M，et al.Vision-based hand pose estimation:a review［J］.Computer Vision and Image Understanding，2007，108(1/2):52-73.
[6]Zhao M J，Lim K B，Wei L K.Geometrical-analysis-based algorithm for stereo matching of single-lens binocular and multi-ocular stereovision system［J］.Journal of Electronic Science and Technology，2012，10(2):107-112.
[7]Cui X Y，Lim K B，Guo Q Y，et al.Accurate geometrical optics model for single-lens stereovision system using a prism［J］.Journal of Optical Society of America A，2012，29(9):1828-1837.
[8]Zhang Z Y.A flexible new technique for camera calibration［J］.Applied Optics，2013，52(1):182-187.
[9]Chen C Y，Deng Q L，Sun W S.Panoramic stereo photography based on single-lens with a double-symmetric prism ［J］.Optical Express，2013，21(7):8474-8482.
[10]郭方，王克逸，闫佩正，等.用于大视场目标定位的复眼系统标定［J］.光学精密工程，2012，20(5):913-920.(Guo Fang，Wang Ke-yi，Yan Pei-zheng，et al.Calibration of compound eye system for target positioning with large field of view［J］.Optical Precision Engineering，2012，20(5):913-920.)(上接第764页)
[10]Lee J M，Yoo C K，Choi S W，et al.Nonlinear process monitoring using kernel principal component analysis［J］.Chemical Engineering Science，2004，59:223-234.
[11]Yu J，Qin S J.Multimode process monitoring with Bayesian inference-based finite Gaussian mixture models［J］.AICHE Journal，2008，54:1811-1829.
[12]Lee J，Kang B，Kang S.Integrating independent component analysis and local outlier factor for plant-wide process monitoring［J］.Journal of Process Control，2011，21:1011-1021.

单目立体视觉系统的棱镜位置估计

Position Estimation for Single-Lens Stereovision System Using Prism

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 6

编辑推荐

Metrics

本文评价

[1]	顾德英，张松，孟范伟. 基于单目视觉的车辆3D空间检测方法[J]. 东北大学学报（自然科学版）, 2022, 43(3): 328-334.
[2]	张田，田勇，王子，王昭东. 基于清晰度评价的自适应阈值图像分割法[J]. 东北大学学报:自然科学版, 2020, 41(9): 1231-1238.
[3]	王大志，李汶浍，袁天清，周迎宾. 基于高频信号注入的永磁同步电机转子位置估计[J]. 东北大学学报:自然科学版, 2019, 40(9): 1234-1239.
[4]	王生生，王萌，王光耀. 基于深度神经网络剪枝的两阶段遥感图像目标检测[J]. 东北大学学报:自然科学版, 2019, 40(2): 174-179.
[5]	赵越，魏永涛，崔笑宇. 棱镜单目立体视觉系统的参数最优化[J]. 东北大学学报(自然科学版), 2013, 34(6): 802-805.
[6]	王文学;赵姝颖;孙萍;徐心和. 多智能足球机器人系统的关键技术[J]. 东北大学学报(自然科学版), 2001, 22(2): 192-195.