偏振成像技术提取目标纹理的研究毕业

《偏振成像技术提取目标纹理的研究毕业》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、偏振成像技术提取目标纹理的研究摘要本文主要通过实验研究偏振图像与强度图像在目标探测方面上的优劣。实验基于偏振光特性结和Stokes矢量表示法，采用激光主动成像与自然光被动成像两种方式，分别在室内、外进行实验，用CCD采集获得各个目标3幅原始图像，再用Matlab软件对3幅原始图像进行处理，得到多组距离、光强不同的偏振度图像。获得目标物体的强度图像与偏振度图像比较两种图像的优劣。实验结果表明：偏振图像与强度图像相比，偏振图像对低照度（阴影等）部分、边缘部分信息得到增强，偏振信息对物体的纹理探测效果明显。所以偏振成像技术在军事侦查、目标识别、海洋搜救等领域有广阔应用前景。关键词

2、：偏振成像；斯托克斯矢量；图像处理；偏振度图像；强度图AbstractThispaperthroughtheexperimentalstudyofpolarizationimageandimageintensityintargetdetectiononthemerits.ExperimentsbasedonpolarizedcharacteristicsandStokesvectorrepresentation,usinglaseractiveimagingwithnaturallightpassiveimagingintwoways,respectively,andint

3、heindoorexperiment,usingCCDdatafromeachtarget3pieceoftheoriginalimage,andthenMatlabsoftwareonthe3pieceoftheoriginalimageprocessing,resultinginmultiplesetsofdistance,intensityofdifferentdegreeofpolarizationimage.Toobtainthetargetobjectintensityimageanddegreeofpolarizationimagecomparisonoftw

4、oimagequality.Atthesametime,weuselinearpolarizationlaserasthelightsource,makingtargetdetectionandrecognitionexperimentforseveralconventionalman-madeobjectsandthenaturaltargetindoorandoutdoor.Weuseimageacquisitionsystemtogetstrengthimageandpolarizationimageofdifferentroughnessmetaltargetand

5、nonmetallictargets,andwritingmatlabprogramtocalculategrayvaluesofimageandenhanceprocessing.Resultsshowthatinlinearpolarizationlaserirradiation,man-madeobjectsandthenaturaltargethaveverybigdifferenceinrefundpartialnature.Technologyoflinearpolarizationlaseractiveimagingcanimprovemetaltargeta

6、ndbackgroundcontrasttothebenefitoftargetdetectionandrecognition.Inactiveimagingofpolarizedactiveimaging,smoothsurfacetargetsaremoreconducivetorecognize.Theexperimentalresultsshowthat:polarizationimageandstrengthcomparedtoimages,polarizationimageonlowintensity(shadowpart),edgeinformationise

7、nhanced,thepolarizationinformationofobjecttexturedetectioneffectisobvious.Sothepolarizationimagingtechnologyinmilitaryreconnaissance,targetidentification,marinesearchandrescueandotherfieldshavebroadapplicationprospects.Keywords:Polarizationimaging；Stokesvector