首页 | 本学科首页   官方微博 | 高级检索  
     

印品检测过程中基于SIFT 算法缩小匹配范围的方法
引用本文:赵立辉,杨红喆,郭栋,霍春宝.印品检测过程中基于SIFT 算法缩小匹配范围的方法[J].包装工程,2013,34(17):104-107.
作者姓名:赵立辉  杨红喆  郭栋  霍春宝
作者单位:辽宁工业大学, 锦州121001,北京理工大学, 北京100083,辽宁工业大学, 锦州121001,辽宁工业大学, 锦州121001
基金项目:辽宁省科技攻关项目(2011201035);辽宁省教育厅项目(L2012230)
摘    要:针对多个CCD 采集多幅图像会产生一定的重叠区域,为了实现印品在线检测的要求,提出了一种结合SIFT 和缩小匹配范围的图像检测方法。该方法基于SIFT 提取特征点,改进了局部搜索范围,利用RANSAC 算法计算图像坐标变换矩阵,采用多分辨率融合方法对拼接图像进行融合处理。结果表明,采用该方法可以减少匹配和图像检验时间,降低估算概率,完成检测图像拼接。

关 键 词:SIFT    CCD    图像拼接    印品检测
收稿时间:7/6/2013 12:00:00 AM
修稿时间:2013/9/10 0:00:00

Method to Reduce Matching Range in Print Detection Based on SIFT Algorithm
ZHAO Li-hui,YANG Hong-zhe,GUO Dong and HUO Chun-bao.Method to Reduce Matching Range in Print Detection Based on SIFT Algorithm[J].Packaging Engineering,2013,34(17):104-107.
Authors:ZHAO Li-hui  YANG Hong-zhe  GUO Dong and HUO Chun-bao
Affiliation:Liaoning University of Technology, Jinzhou 121001, China,Beijing Institute of Technology, Beijing 100083,China,Liaoning University of Technology, Jinzhou 121001, China and Liaoning University of Technology, Jinzhou 121001, China
Abstract:A print detection method based on SIFT to reduce matching range was proposed to solve image overlap region when acquiring images with the multiple CCD for online printing detection requirement. The method improve local searching range using image extraction feature of SIFT. RANSAC algorithm was used to calculate the image coordinate transform matrix and the multi resolution fusion method was used to fuse stitching images. The results showed that the method can effectively reduce the matching and test time of image, reduce the error probability, and complete image stitching.
Keywords:SIFT  CCD  image stitching  print detection
点击此处可从《包装工程》浏览原始摘要信息
点击此处可从《包装工程》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