首页 | 本学科首页   官方微博 | 高级检索  
     

基于角点定位的印刷品缺陷检测
引用本文:王诗琴,程耀瑜,师惠萍,刘永林. 基于角点定位的印刷品缺陷检测[J]. 传感器世界, 2011, 17(4): 30-33
作者姓名:王诗琴  程耀瑜  师惠萍  刘永林
作者单位:中北大学山西省现代无损检测工程技术研究中心,山西太原,030051
摘    要:通过CCD采集到的图像受光照及环境干扰存在大量噪声,本文首先用中值滤波去除噪声、保留图像边缘细节.缺陷检测中选择了基于角点的定位方法,该算法计算量小、速度快,能够满足在线印刷品检测的要求.定位后待检测图像需要与标准图像配准,通过图像减法得到差分图像,将其各象素与预先设定好的阈值进行比较,如果大于阈值,则记为缺陷点.本文...

关 键 词:角点  图像配准  缺陷检测  连通域

Print defect detection based on corner location
WANG Shi-qin,CENG Yao-yu,SHI Hui-ping,LIU Yong-lin. Print defect detection based on corner location[J]. Sensor World, 2011, 17(4): 30-33
Authors:WANG Shi-qin  CENG Yao-yu  SHI Hui-ping  LIU Yong-lin
Affiliation:WANG Shi-qin CENG Yao-yu SHI Hui-ping LIU Yong-lin(Modern Nondestructive Testing Technology Research Center of Shanxi,North University of China,Taiyuan,030051,China)
Abstract:There are lots of noises in images collected by CCD because of the light and environmental disturbances.Median filter is used to remove noises and preserve image edge details.To detect the defects,corner-based positioning method is adopted in this paper,which has small arithmetic operations,fast arithmetic speed and capability to meet the requirements of online print detection.After positioning,the detected images need registration with the standard image and differencing image is obtained by image subtraction,then pixel values are compared with a predetermined threshold value.If certain pixel value is more than the threshold,this pixel point is marked as defect point.At last connected component analysis(CCA)and blob analysis algorithm are used to extract and classify defects of defect-images.
Keywords:corner  image registration  defect detection  connected component analysis(CCA)
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