基于灰度图像的答题卡识别技术

28320096Vol.28No.3Jun.200999JournalofShandongUniversityofScienceandTechnologyNaturalScience基于灰度图像的答题卡识别技术,(山东科技大学信息科学与工程学院,山东青岛266510):针对目前电脑阅卷技术中存在的一些问题,提出了一种基于灰度图像识别技术的答题卡判别方法首先对答题卡图像进行高斯滤波,依据模板定位填涂区域,然后针对正常填涂未填涂和涂改等情况做出正确的判断,最后,将答案以及考生信息读入数据库此系统为不同答题卡定义相应的模板,可应用于各类答题卡的阅卷工作实验结果表明了此方法的有效性:灰度图像;高斯滤波;图像识别;特征提取;答题卡:TP391.41:A:1672-3767(2009)03-0099-04TheExamCardReadingTechnologyBasedonGreyImageYANGQing-yan,PENGYan-jun(CollegeofInformationScienceandEngineering,SUST,Qingdao,Shandong266510,China)Abstract:Inviewoftheshortageexistedincomputerscoringtechnologynowadays,thispaperintroducesanewmethodforexamcardjudgmentbasedongrayimagerecognitiontechnology.Firstly,itmakestheexamcardimagebeGaussianfilteredtofocusonthemarkedareasbasedontemplate;then,itmakesthecorrectjudgmentsuponnormalmark,unwrittenmarkandchangedmark,andlastly,itinputssuchinformationasanswersandstudentsintothedatabase.Thissystemcandefinerelevantformatsupondifferentexamcards,thus,itissuitableforanswerscoringindifferentkindsofexaminations.Andthetestresultsprovethatthemethodiseffective.Keywords:grayimage;Gaussianfilter;imageidentification;characterextracting;examcard:2008-06-04:(J05C10):(1983)),,,,.(1971)),,,,,.1Fig.1Theflowchartofwholeframe[1],,,,,,,,1DIB,,,,,,,1,10028320096Vol.28No.3Jun.200922.1,,[a,b][Zl,Zk],,[a,b][Zl,Zk][2],zz:Zc=Zk-Zlb-a(Z-a)+Z(1),:Zc=Zk-Zlb-a(Z-a)+Za[z[bZlzaZkzb(2),,,232.2,,,:Q(x,y)=12PR2e-x2+y22R2(3)2fcolor(x,y),Gcolor(x,y)=63i=063j=0g(xi,yj)*fcolor(xi,yj)16(4)g(xi,yj),,,[3]33.1,[4],,43.2,4,(),[5],,(5)~(8)X=x6w/5i=0fcolor(xi,y)20,(5)Y=y6h/5j=0fcolor(x,yj)20,(6)基于灰度图像的答题卡识别技术101JournalofShandongUniversityofScienceandTechnologyNaturalScienceX=X-X,(7)Y=Y-Y(8)h,w,1/5,,4Fig.4Theimageoftemplate3.3,,:,,T,,X=xfcolor(x,y)T,(9)Y=yfcolor(x,y)T(10),,,,;,,14,5,10,11,nR,,150,6,5644.1,,:count=6h,wi=0,j=0fcolor(xi,yj)(11)countB=6h,wi=0,j=0fcolor(xi,yj),fcolor(xi,yj)T(12)10228320096Vol.28No.3Jun.2009Ratio=count/countB(13),count,countB,Ratio,4.2,,[6],,,,,,,,,5:P4,512MB,Windows2003,,2000,120,:,2/s,,100%;0.5%;0.4%,,,6[7],,,,,:[1]高育鹏,杨俊,何广军.基于图像识别的自动阅卷系统研究[J].现代电子技术,2006,29(22):119-120.GAOYu-peng,YANGJun,HEGuang-jun.Researchonauto-gradingsystembasedonimageidentificationtechnology[J].ModernElectronicsTechnique,2006,29(22):119-120.[2]岳永哲,李然.图像处理技术在阅卷系统的应用[J].电脑知识与技术,2007(14):521-522.YUEYong-zhe,LIRan.Imageprocessingtechnologyinthecorrectpaperssystem[J].ComputerKnowledgeandTechnology,2007(14):521-522.[3]陈万海,赵春晖.基于高斯低通滤波的超光谱遥感图像分类研究[J].黑龙江大学自然科学学报,2007,12:251-156.CHENWan-hai,ZHAOChun-hui.HyperspectralremotesensingclassificationbasedonGaussianlowpassfilter[J].JournalofNaturalScienceofHeilongjiangUniversity,2007,12:251-156.[4]周万珍,郑广,王健霞.数字图像处理技术在客观中的应用[J].数学的实践与认识,2006(8):243-246.ZHOUWan-zhen,ZHENGGuang,WANGJian-xia.Digitalimageprocessingapplicationinrecognitionofimpersonalexamcardreader[J].MathematicsinPracticeandTheory,2006(8):243-246.[5]张香让,崔喆.击中击不中变换在标准答题卡分割中的应用[J].计算机应用,2004,24(10):141-143.ZHANGXiang-rang,CUIZhe.Applicationofhit/misstransforminseparationofstandardanswercard[J].ComputerApplications,2004,24(10):141-143.[6]郑广,秦敏,付冬,等.基于图像识别的客观题阅卷研究[J].仪器仪表学报,2006,27(6):273-274,791.ZHENGGuang,QINMin,FUDong,etal.Impersonalexamcardreadingbasedondigitalimagerecognitionresearch[J].ChineseJournalofScientificInstrument,2006,27(6):273-274,791.[7]田春谨.数字图像处理技术在网上评卷系统中的应用[D].昆明:昆明理工大学,2007.