(ebook - pdf) Image Processing - Efficient Focusing and Face Detection

(ebook - pdf) Image Processing - Efficient Focusing and Face Detection

ID:37748837

大?。?74.74 KB

頁(yè)數(shù):16頁(yè)

時(shí)間:2019-05-30

(ebook - pdf) Image Processing - Efficient Focusing and Face Detection_第1頁(yè)
(ebook - pdf) Image Processing - Efficient Focusing and Face Detection_第2頁(yè)
(ebook - pdf) Image Processing - Efficient Focusing and Face Detection_第3頁(yè)
(ebook - pdf) Image Processing - Efficient Focusing and Face Detection_第4頁(yè)
(ebook - pdf) Image Processing - Efficient Focusing and Face Detection_第5頁(yè)
資源描述:

《(ebook - pdf) Image Processing - Efficient Focusing and Face Detection》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在行業(yè)資料-天天文庫(kù)。

1、EcientFocusingandFaceDetectionyzYaliAmit,DonaldGemanandBrunoJedynakOctober1997TechnicalReportno.459DepartmentofStatisticsUniversityofChicagoAbstractWepresentanalgorithmforshapedetectionandapplyittofrontalviewsoffacesinstillgreylevelimageswitharbitrarybackgrounds.Detectionisdoneintwostages:(

2、i)focusing,"duringwhicharelativelysmallnumberofregions-of-interestareidenti ed,minimizingcom-putationandfalsenegativesatthe(temporary)expenseoffalsepositivesand(ii)intensiveclassi cation,"duringwhichaselectedregion-of-interestislabeledfaceorbackgroundbasedonmultipledecisiontreesandnormalize

3、ddata.Incontrasttomostdetectionalgorithms,theprocessingisthenveryhighlyconcentratedintheregionsnearfacesandnearfalsepositives.Focusingisbasedonspatialarrangementsofedgefragments.We rstde neanenormousfamilyofthese,allinvariantoverawiderangeofphotometricandgeometrictransformations.Then,usingonl

4、yexamplesoffaces,weselectparticulararrangementswhicharemorecomoninfacesthaningeneralbackgrounds.Thesecondphaseistexture-basedwerecursivelypartitionatrainingsetconsistingofregisteredandstandardizedregions-of-interestofbothfacesandnon-faces.Thefacetrainingdataconsistof30individuals,10imagesperp

5、erson,obtainedfromtheOllivettidatabase.Theprocessingtime(onaUltraSparc2)isunderasecondforatestimageofsize100100,andscaleslinearlywiththesizeoftheimage.Weachieveafalsepositiverateofabout.2per10000pixelsweestimateafalsenegativerateof10%.DepartmentofStatistics,UniversityofChicago,Chicago,IL,60

6、637Email:amit@galton.uchicago.edu.SupportedinpartbytheArmyResearchOceundergrantDAAH04-96-1-0061andtheDepartmentofDefensegrantDAAH04-96-1-0445.yDepartmentofMathematicsandStatistics,UniversityofMassachusetts,Amherst,MA01003Email:geman@math.umass.edu.SupportedinpartbytheNSFundergrantDMS-92176

7、55,ONRundercon-tractN00014-91-J-1021,andtheArmyResearchOce(MURI)undergrantDAAH04-96-1-0445.zDepartmentofStatistics,UniversityofChicago,Chicago,IL,60637Email:bruno@galton.uchicago.edu.Sup-portedinpartbytheArmyResearchOceundergrantDAAH04-96-

當(dāng)前文檔最多預(yù)覽五頁(yè),下載文檔查看全文

此文檔下載收益歸作者所有

當(dāng)前文檔最多預(yù)覽五頁(yè),下載文檔查看全文
溫馨提示:
1. 部分包含數(shù)學(xué)公式或PPT動(dòng)畫(huà)的文件,查看預(yù)覽時(shí)可能會(huì)顯示錯(cuò)亂或異常,文件下載后無(wú)此問(wèn)題,請(qǐng)放心下載。
2. 本文檔由用戶上傳,版權(quán)歸屬用戶,天天文庫(kù)負(fù)責(zé)整理代發(fā)布。如果您對(duì)本文檔版權(quán)有爭(zhēng)議請(qǐng)及時(shí)聯(lián)系客服。
3. 下載前請(qǐng)仔細(xì)閱讀文檔內(nèi)容,確認(rèn)文檔內(nèi)容符合您的需求后進(jìn)行下載,若出現(xiàn)內(nèi)容與標(biāo)題不符可向本站投訴處理。
4. 下載文檔時(shí)可能由于網(wǎng)絡(luò)波動(dòng)等原因無(wú)法下載或下載錯(cuò)誤,付費(fèi)完成后未能成功下載的用戶請(qǐng)聯(lián)系客服處理。