1163.基于sift特征的巖石圖像拼接研究

1、學位論文創(chuàng)新性聲明本人聲明所呈交的學位論文是我個人在導師指導下進行的研究工作及取得的研究成果。盡我所知，除了文中特別加以標注和致謝中所羅列的內容以外，論文中不包含其他人已經(jīng)發(fā)表或撰寫過的研究成果；也不包含為獲得西安石油大學或其它教育機構的學位或證書而使用過的材料。與我一同工作的同志對本研究所做的任何貢獻均已在論文中做了明確的說明并表示了謝意。申請學位論文與資料若有不實之處，本人承擔一切相關責任。論文作者簽名：日期：儼哆r(，護學位論文使

2、用授權的說明本人完全了解西安石油大學有關保留和使用學位論文的規(guī)定，即：研究生在校攻讀學位期間論文工作的知識產(chǎn)權單位屬西安石油大學。學校享有以任何方法發(fā)表、復制、公開閱覽、借閱以及申請專利等權利。本人離校后發(fā)表或使用學位論文或與該論文直接相關的學術論文或成果時，署名單位仍然為西安石油大學。論文作者簽導師簽日期：VU。6I口嗍冽通堵豢英文摘要Subject：ResearchofRockImageStitchingbasedonSIFTFea

3、tureSpeciality：Name：YmInstructor：ABS’l。RAC’I。11linsectionsareanimportantresearchsubjectforreservoirmicrostructuresneform，range，grainsizeandfabricoftherockcanbestudiedindetailunderamicroscopeHoweverwiththelimitationofthes

4、copeofmicroscopehorizontheoverallcharacteristicsandthelocaldetailsoftherockthinsectioncannotbeshowedclearlySIFTalgorithmisafastandeffectivealgorithmforimagefeatureextractionandthefeaturepointsextractedhavegooduniquenessa

5、ndhiglIcompatibilityThisarticleexplainstheprocessofimagestitchingtechnologyanalysestheprocedureofSIFTalgorithmforextractingfeaturepointsWithanemphasisontheapplicationofimagestitchingtocks，theadvantagesanddisadvantagesoft

6、heSIFTalgorithmisanalyzedandimprovedOnonehand，theoverlappingareasoftheimagesalecalculatedandthefeaturepointsoftheoverlappingareasareextractedSOthatmatchesCanreducetheerrorrateOntheotherhand，beforematchingtheimages，Princi

7、palComponentAnalysis(PCA)isusedteducethedimensionofSIFTfeaturevectorsinoverlappingareRExperimentalresultsshowthatthenewmethodreducesthecomputationofmatchshortenstherunningtimeandimproveaccuracyAfterthestitchoftwoimages，t

8、hismethodisexpandedtohugeamountsofhighresolutionmicroscopyinⅡ呼sforstitchingandacomprehensivesuperultrahighresolutionpanoramicviewofthemicro掣aphisgotBecausetheexistingcomputersoftwareisdifficulttofastdisplayoflargeinlages

9、，pyramidmodelisusedtopartitionhierarchyandstorageanddisplayinghierarchicallyUltimatelyresearcherscanobservetheoverviewfeaturesunderthelargescaleandstudythelocaldetailsundersmal】sealeKeywords：Thinsection，SIFTfeature，Image