欢迎来到天天文库
浏览记录
ID:53022895
大小:1.67 MB
页数:11页
时间:2020-04-12
《多光谱掌脉和掌纹离焦图像融合方法.pdf》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、ISSN10049037,CODENSCYCE4http://sjcj.nuaa.edu.cnJournalofDataAcquisitionandProcessingVo1.30,No.1,Jan.2015,PP.175—185E—mail:sjcj@nuaa.edu.cnDOI:10.16337/i.1004—9037.2015.01.017Tel/Fax:+86—025-84892742◎2015byJournalofDataAcquisitionandProcessing多光谱掌脉和掌纹离焦图像融合方
2、法汤永华苑玮琦(沈阳工业大学视觉检测技术研究所,沈阳,110870)摘要:非接触条件下单模态手部图像纹理信息有限,不利于识别。本文提出利用同一装置在多光谱下分时快速获取掌纹和掌脉图像。通过掌纹图像在静脉图像中进行掌纹主线增强处理,达到增加纹理信息、改善识别效果的目的。控制掌纹、掌脉两幅图像获取间隔小于0.1S,忽略手掌细微空间移位,对获取的掌纹图像进行感兴趣区域(Regionofinterest,ROI)提取,并利用掌纹R0I坐标参数进行静脉图像的ROI提取。然后对两个R0I区域进行预处理,通过多层小波分解得
3、到的掌纹高频信息快速定位掌纹主线。最后根据掌纹主线归一化后的高频分量,进行掌纹和掌脉小波系数的融合,最终得到在静脉纹理图像中增强掌纹主线纹理的融合图像。实验结果表明,融合后的图像中在保持静脉纹理信息原有状态下,掌纹主线纹理信息明显增加,识别效果得到改善。关键词:掌脉和掌纹数据融合;非接触成像;高低频混合;多光谱中图分类号:TP391文献标志码:AMultispectralPalm—VeinandPalmprintDefoeusedImageFusionMethodTangYonghua,YuanWeiqi(C
4、omputerVisionGroup,ShenyangUniversityofTechnology,Shenyang,110870,China)Abstract:Textureinformationofthesinglemodehandsimageisnotsufficientforrecognitioninnon—con—tactcondition.Asingledeviceunderthemulti—spectrumisputforwardtorapidlycapturetheimagesofpalmpr
5、intandpalmvein.Thetextureinformationisenhancedandtherecognitionperformanceisim—provedthroughenhancingthepalmprintmainlinesoftheveinimage.Firstly,theintervalofacquiringthepalmprintandpalmveinimagesarelimitedtolessthan0.1S,andthetinydisplacementofpalmisignore
6、d.AndtheROIofthepalmprintimageisdetermined,whosecoordinatesareusedtodeterminetheROIofpalmveinimage.Then,thepreprocessingofthetwoROIregionsareperformed,andthepalmprintmainlinesarerapidlydeterminedviathepalmprinthigh—frequeneeinformationfrommulti—lay—erwavele
7、tdecomposition.Finally,waveletcoefficientsofpalmprintandpalmveinarefusedaccordingtothehighfrequencycomponentofpalmprintmainlinenormalized,andtheblendedtextureimagewiththeenhancedpalmprintmainlineintheveintextureimageisacquired.Theexperimentresultsshowthatth
8、einformationofpalmprintmainlinetextureissignificantlyenhancedinthefusionimagesintheconditionofkeepingtheoriginalstateoftheveintexture,andtherecognitionperformanceisimproved.Keywords:palm—veinandpalmpri
此文档下载收益归作者所有