预览加载中,请您耐心等待几秒...
1/4
2/4
3/4
4/4

在线预览结束,喜欢就下载吧,查找使用更方便

如果您无法下载资料,请参考说明:

1、部分资料下载需要金币,请确保您的账户上有足够的金币

2、已购买过的文档,再次下载不重复扣费

3、资料包下载后请先用软件解压,在使用对应软件打开

基于灰度共生矩阵的高效积分图像计算(英文) Abstract: Inthispaper,weproposeahigh-efficiencyalgorithmforthecomputationofintegralimagesbasedonthegray-levelco-occurrencematrix.Theintegralimage,alsoknownasthesummedareatable,iswidelyusedinimageprocessingtospeedupoperationssuchasconvolutionandfeatureextraction.Ourmethodusesthegray-levelco-occurrencematrixtocomputetheintegralimageinamoreefficientmannerthanthetraditionalapproach.Experimentresultsshowthatouralgorithmperformsbetterintermsofbothtimeconsumptionandmemoryusagecomparedtothetraditionalmethod. Introduction: Theintegralimageisanimportanttoolinimageprocessingthatenablesfastcomputationofvariousimageoperationssuchasconvolution,edgedetection,andfeatureextraction.Itisa2Dmatrixthatcontainsthesumofallthepixelsintheimageuptoacertainpoint,calculatedbyaddingupthepixelsalongthecolumnandrowdirections.Traditionally,theintegralimageiscomputedusingabrute-forcemethodthatinvolvesiteratingthrougheachpixelintheimage.Thismethodiscomputationallyexpensiveandmemory-intensive,especiallyforlargeimages. Toovercomethesechallenges,manyresearchershaveproposedvariousalgorithmstocomputetheintegralimagemoreefficiently.Forexample,theBoxfiltermethodandtheFastintegralimagemethodusedifferentconvolutiontechniquestospeedupthecomputation.However,thesemethodsstillhavelimitationsintermsoftimeconsumptionandmemoryusage. Inthispaper,weproposeanovelalgorithmforthecomputationofintegralimagesbasedonthegray-levelco-occurrencematrix.Thegray-levelco-occurrencematrixisastatisticalmeasurethatquantifiesthedistributionofgray-levelpairsinanimage.Itiswidelyusedinimageanalysisandtextureclassification.Ourmethodutilizesthegray-levelco-occurrencematrixtocomputeintegralimagesinamoreefficientmannercomparedtothetraditionalapproach. Methods: Ourmethodconsistsofthefollowingsteps: 1.Computethegray-levelco-occurrencematrixfortheinputimage. 2.Computetheintegralimageofthegray-levelco-occurrencematrixusingthetraditionalmethod. 3.Computetheintegralimageoftheinputimageusingthegray-levelco-occurrencematrix. St