首页 | 本学科首页   官方微博 | 高级检索  
     

GPS双目摄像机标定及空间坐标重建
引用本文:孔筱芳,陈钱,顾国华,钱惟贤,任侃.GPS双目摄像机标定及空间坐标重建[J].光学精密工程,2017,25(2):485-492.
作者姓名:孔筱芳  陈钱  顾国华  钱惟贤  任侃
作者单位:南京理工大学 电子工程与光电技术学院, 江苏 南京 210094
基金项目:江苏省基础研究计划(自然科学基金)青年基金项目
摘    要:为减轻双目摄像机标定过程中对高精度靶标的依赖,实现摄像机参数的精确标定,并对空间坐标进行高精度重建,提出一种GPS双目摄像机标定及空间坐标重建方法,采用GPS代替2D或3D靶标进行双目摄像机标定。将GPS的位置在视场中任意移动,由被标摄像机拍摄多组含有GPS的图像,利用空间三维坐标与图像二维坐标间的映射关系,结合摄像机成像模型和双目摄像机标定原理,标定出双目摄像机参数,并对空间坐标进行精确重建。通过空间重建坐标与GPS实际测量值之间的相对距离误差,对重建精度进行检验。实验证明,该方法能够克服双目摄像机标定过程中对高精度靶标的依赖,空间重建坐标具有较高的精度,相对距离误差从1.56%减小到0.52%。

关 键 词:计算机视觉  双目摄像机标定  摄像机成像模型  GPS坐标转换  空间坐标重建
收稿时间:2016-07-02

GPS-based binocular camera calibration and space coordinate reconstruction
KONG Xiao-fang,CHEN Qian,GU Guo-hua,QIAN Wei-xian,REN Kan.GPS-based binocular camera calibration and space coordinate reconstruction[J].Optics and Precision Engineering,2017,25(2):485-492.
Authors:KONG Xiao-fang  CHEN Qian  GU Guo-hua  QIAN Wei-xian  REN Kan
Affiliation:School of Electronic and Optical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China
Abstract:To alleviate the dependence of binocular camera on high-accuracy target during calibration,achieve precise calibration of camera parameters and implement high-accuracy reconstruction of space coordinates,a GPS-based method for binocular camera calibration and space coordinate reconstruction is introduced,in which GPS instead of 2D or 3D targets is used to carry out the camera calibration.Through arbitrarily moving the GPS at the shooting site,multiple images with GPS are shot by calibrated camera;based on the mapping relation between spatial 3D coordinates and 2D image coordinates,the binocular camera parameters are calibrated and precise reconstruction of space coordinates is carried out by combining camera imaging model with calibration theory of binocular camera;finally,the reconstruction accuracy is tested in accordance with relative distance error between the space reconstruction coordinates and actual measurement values.The reconstruction results of proposed method are compared with the true data obtained by the GPS,and the results indicate that the method can help binocular camera to overcome the dependence on high-accuracy targets and equip the space reconstruction coordinates with relatively high accuracy,with a relative distance error decreased from 1.56% to 0.52%.
Keywords:computer vision  binocular camera calibration  camera model  GPS coordinate transformation  space coordinate reconstruction
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《光学精密工程》浏览原始摘要信息
点击此处可从《光学精密工程》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号