基于模式識別的鞋底花紋自動識別與分類.pdf

1、學(xué)位論文基于模式識別的鞋底花紋自動識別與分類(題目)楊曉勝(作者姓名)指導(dǎo)教師梁德群教授申請學(xué)位級別工學(xué)碩士專業(yè)名稱信號與信息處理學(xué)位授予單位大連海事大學(xué)2005年3月AbstraotThefootprIntandshoeprintisveryimportantevidancewhiehIseasytobegottennthecase，Itisveryimportantinseriespara]1elcaseBetbecauseofev

2、erylocalpolice’sfootprintmanagementmethodisdifferentfromeachotherandtheformissingle，ItmakesfootprintcannotbeusedlikefingerprintincriminalcaseSothesoletexCure’sautomaticidentificationandclas8ificationhasagreateconomicsign

3、／fieanceandsocialsignificanceInthispaper，basedontheanalysisofmanyfootprintimagesandtheresearchworkinedgedetectionfieldthathavebeendoneinrecentyearsbyImageProcessingInstituteofXi’anJiaotongUniversityandDalJanMaritimeUnive

4、rsity，thealgorithmsarebroughtforwardasfollowing：tthepunctuatetextureisidentifiedbythei柚ageprobabilitymethodinertiamoment2，theinterlacedtextureisidentifiedbydetectingthenumberofedge’Scrossdot，basedOnthewaveletedgedetect3，

5、TheshapeanalysisisusedtoanalyzethemainshapeoftheedgeimageSomecharacterisfoundlikelength，width，theratiooflengthandwidth，acreage，andSOon4theflexuoustextureand1inetextureareidentifiedbyusingheratiooflengthandwidthanddetecti

6、ngtheinflexion’snumberinthethinningimageofthe1abelimage5thenubblytextureisidentifiedbyusingtheratiooflengthandwidth，thewidthanddetectingthec：rossdot’snumberinthethinningimageofthelabelimage6thecirculartexture，theelliptic