首页 | 官方网站   微博 | 高级检索  
     

机器视觉系统中图像误差补偿方法
引用本文:王跃轩,易红,倪中华.机器视觉系统中图像误差补偿方法[J].数据采集与处理,2005,20(4):472-477.
作者姓名:王跃轩  易红  倪中华
作者单位:东南大学机械工程系,南京,210096
基金项目:国家自然科学基金(50475074)资助项目;江苏省科技基金(BK2005072)资助项目.
摘    要:在分析和比较常用图像畸变校正算法的基础上,提出将BP神经网络原理应用于图像的畸变校正,从而在对图像进行检测和定量分析之前,消除畸变误差对测量结果的影响;利用在标准栅格样板图像上选择的控制点对神经网络进行训练,并将训练后的神经网络以动态链接库的形式提供给LabVIEW调用,在LabVIEW环境中实现了具有图像误差补偿功能的机器视觉测量系统.

关 键 词:机器视觉  图像误差补偿  BP神经网络  LabVIEW  动态链接库
文章编号:1004-9037(2005)04-0472-06
收稿时间:2004-12-27
修稿时间:2005-05-25

Method for Image Error Compensation in Machine Vision System
WANG Yue-xuan,YI Hong,NI Zhong-hua.Method for Image Error Compensation in Machine Vision System[J].Journal of Data Acquisition & Processing,2005,20(4):472-477.
Authors:WANG Yue-xuan  YI Hong  NI Zhong-hua
Affiliation:Department of Mechanical Engineering, Southeast University, Nanjing, 210096, China
Abstract:In the machine vision system,the nonlinear and perspective distortion errors generally exist in acquired images due to the severe geometrical distortion induced by optical lens and camera position.Based on the analysis and the comparison of current research methods,the BP neural network is applied in image correction to eliminate the immediate effect on measurements before extracting quantitative measurements.The neural network for the correction is trained by control points selected from standard grid templet as a dynamic link library(DLL) file for LabVIEW call,and the machine vision system with the image error compensation function is realized in LabVIEW.
Keywords:machine vision  image error compensation  BP neural network  LabVIEW  dynamic link library
本文献已被 CNKI 维普 万方数据 等数据库收录!
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-23

京公网安备 11010802026262号