基于紋理分析的目標(biāo)圖像識別技術(shù)研究.pdf

1、分類號UDC注]密級鱉Ⅱ——學(xué)、位論文基于紋理分析的目標(biāo)圖像識別技術(shù)研究(題名和剮題名)孫棟(作者姓名)指導(dǎo)教師姓名型一睦型塾撞甲請學(xué)位級別塑專業(yè)名稱蕉劍堡迨量蕉型三堡論文提交日期2QQ5魚——論文答辯FJ期學(xué)位授予單位干口日期——由立堡三盤堂答辯委員會主席評閱人2005年06月日注l：注明《國際十進(jìn)分類法UDCi)的類號。ABSTRACT碩士論文ABSTRACTItisbackgroundonTVguidanceandregardst

2、herunwayastherecognitionobjectinthispaperWedosomeresearchonthetechnologyofthetargetimagerecognitionbasedontextureanalysisThepaperanalyzesandcomparesseveralkindsofthealgorithmoftheedgedetectionsuchasRobert、Krischalgodthma

3、ndKrischischosentoextracttheedgeofthetargetItdescribestheshapeoftherunwaywiththerelative—momentmethodanddoessomerevisionstotheresultWedosomeresearchonprincipleandfeatureofthetwotextureanalysismethods，Theproblemofthesensi

4、tivitytotheimagerotationiswellsolvedbyadoptingthegray—gradientCOoccurrencemethodAgeneralframeoftherunwayrecognitionalgorithmisconstructedhereAfterthepre—processing，thetargetrunwayissegmentedfromthesourceimageandtheshapef

5、eatureisextractedthenInsuccessionthetexturefeatureisextractedbythetextureanalysismethodTherunwaymodelisbuiltonthebasisofthetwomainfeaturesextractedbeforeIntheprocessofrecognition，thereal—timeinputimageisobjectsegmentedwi

6、ththesamemethodasbefore，thenextractedtheobject’SshapefeatureandcomparedthefeaturewiththatofthemodelExtractthetexturefeatureonconditionthattheshapematchesthemodelcorrectlyLatermatchthemodelwitllthetexturefeatureIfthetwofe

7、aturesmatchthemodelcorrectlyweconsidertheobjectasthetargetrunwayLastdevelopandrealizethegeneralMgodtbanundertheVCenvironmentanddotheprimarytestonthealgorithmTheresultilluminatesthatthisalgorithmhasthebeaereffectwiththelo