首页 | 本学科首页   官方微博 | 高级检索  
     

空间圆几何参数视觉测量方法研究
引用本文:崔岸,何大鹏,王丹. 空间圆几何参数视觉测量方法研究[J]. 计算机工程与应用, 2010, 46(32): 55-58. DOI: 10.3778/j.issn.1002-8331.2010.32.015
作者姓名:崔岸  何大鹏  王丹
作者单位:吉林大学 汽车动态模拟国家重点实验室,长春 130025
基金项目:国家重点实验室基金,吉林省科技发展计划基金
摘    要:对空间圆几何参数的非接触在线测量方法进行了研究。通过在区域划分中引入模糊化思想,基于模糊隶属度构建一种神经网络的拓扑结构,采用立体标靶进行摄像机标定。面向局部检测区域小的空间尺寸测量问题,利用图像的梯度相关作为匹配特性,开发一种高效的基于梯度图像的立体匹配算法。在实验研究中,对某车身翼子板安装孔进行了检测,结果显示当空间圆所在平面与图像平面约50°角时,测量空间圆的相对误差优于±0.6%。研究结果表明,提出的方法测量精度高,并适用于在线测量。

关 键 词:模糊神经网络  摄像机标定  空间圆  几何参数  双目立体视觉  
收稿时间:2010-05-10
修稿时间:2010-7-15 

Method for measuring geometric parameters of spatial circle
CUI An,HE Da-peng,WANG Dan. Method for measuring geometric parameters of spatial circle[J]. Computer Engineering and Applications, 2010, 46(32): 55-58. DOI: 10.3778/j.issn.1002-8331.2010.32.015
Authors:CUI An  HE Da-peng  WANG Dan
Affiliation:State Key Laboratory of Automotive Dynamic Simulation,Jilin University,Changchun 130025,China
Abstract:The non-contact inspection method for the geometrical parameters of spatial circle is studied.A neural network topology is constructed based on fuzzy membership degree through introducing fuzzy theory at the regional division.The three-dimensional target is used for camera calibration.A gradient-based stereo image matching algorithm is developed by using the image gradient-related as the matching feature,especially for the space size measurement of a small region.Experimental results show that the relative error of spatial circle measurement is better than ±0.6% by the proposed approach,when the space plane containing spatial circle and the image plane is at 50°.This indicates that the proposed approach can achieve a high precision.It is applied to on-line measurement.
Keywords:fuzzy neural network  camera calibration  spatial circle  geometric parameters  binocular stereo vision
本文献已被 维普 万方数据 等数据库收录!
点击此处可从《计算机工程与应用》浏览原始摘要信息
点击此处可从《计算机工程与应用》下载免费的PDF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