大尺寸工件视觉测量中的图像拼接方法  

Image Mosaics for Vision Measurement on Large-scale Workpiece

作　　者：陈海林[1] 熊芝[1] 宋小春[1] 夏清沛 

机构地区：[1]湖北工业大学机械工程学院,武汉430068

出　　处：《计算机测量与控制》2015年第2期523-525,541,共4页Computer Measurement &Control

基　　金：国家自然科学基金资助项目(51305130);湖北省自然科学基金资助项目(2013CFB033);湖北工业大学博士启动金资助项目(BSQD12123)

摘　　要：为提高大尺寸工件机器视觉测量精度,提出一种基于坐标变换的图像拼接方法,并给出该坐标变换方法的解算模型及求解方法,分析了局部测量误差、重叠区域公共点个数等对坐标变换精度的影响,得到了最佳拼接条件,并通过实验对其进行了验证;实验结果表明,该方法简单易行、结果可靠,对存在平移、旋转以及尺度缩放的图像都具有良好的效果,能有效地解决采用机器视觉进行大尺寸工件精确测量时的坐标归一化问题。In order to improve the measurement precision of machine vision on the large-scale workpiece,a image mosaic method based on coordinate transformation was presented in this paper.Firstly,the calculation model and solving method of this coordinate transformation means was given,then the relationship between the accuracy of coordinate transformation and factors such as local measurement errors and the number of public points in overlapping areas were analyzed.Finally,optimal stitching conditions were found and further verified by experiments.Experimental results show that this coordinate transformation method is simple,reliable and has a good effect of images that has translation,rotation and scaling,which can effectively solve the problem of coordinate normalization problem in machine vision on the precision measurement of large-scale workpiece.

关 键 词：汽车大梁 视觉测量 坐标变换 图像拼接 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]