基于形态学的分水岭算法牙科x射线图像分割

基于形态学的分水岭算法牙科x射线图像分割

ID:875786

大小:231.03 KB

页数:16页

时间:2017-09-22

基于形态学的分水岭算法牙科x射线图像分割_第1页
基于形态学的分水岭算法牙科x射线图像分割_第2页
基于形态学的分水岭算法牙科x射线图像分割_第3页
基于形态学的分水岭算法牙科x射线图像分割_第4页
基于形态学的分水岭算法牙科x射线图像分割_第5页
资源描述:

《基于形态学的分水岭算法牙科x射线图像分割》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库

1、英文原文WatershedAlgorithmBasedonMorphologyforDentalX-RayImagesSegmentationAbstract—Anewwatershedalgorithmbasedonmathematicalmorphology,whichcanbeappliedtodentalX-rayimagessegmentation,isproposedinthepaper.Inordertoseparateeachtoothandimprovetheseriousproblemofover-segmentation

2、oftraditionalwatershedalgorithm,weapplytheprocessingbasedonmorphologytogettheimagesenhancedbeforewatershedalgorithm.First,usingthetop-hat-bottom-hattransformationtoamplifythecontrastofforegroundandbackgroundandremovenoises.Andthenerosionprocedureisusedtoweakenthedegreeofadh

3、esionbetweenteeth.Hole-fillingcanhelptoeliminatesomeunnecessarysplit-linewhichcancausethephenomenonofwrongsegmentation.Finallyweapplythewatershedalgorithmusingdistancetransformationofthebinaryimagetogetthesegmentationresult.Fromtheexperimentresults,wecanseethatthisalgorithm

4、canseparateteethaccuratelyandovercometheover-segmentationefficientlycomparedwithtraditionalmethod.Keywords-watershedalgorithm;mathematicalmorphology;dentalX-rayimage;imagesegmentation;distancetransformationI.INTRODUCTIONDentalbiometricsisanappropriatemethodtoidentifydecease

5、dindividualsofdisasterssuchascriminalcases,blastsandTsunamis,becauseteethwiththeirdentalworksareveryresistanttomodestforceeffectsandhightemperatureandchemicalcorrosionandalsopossessgoodbiometricproperties.Anditalsoprovidesimportantinformationindentalmedicaltreatmentforfurth

6、eranalysistogetthediagnostic.Imagesegmentationasaprecursorhasbeenseenasoneofthemostimportantaspectoftheentireprocesslikepatternrecognition.Itgivesclarityinformationandthelevelstosubdivisiondependontheproblembeingsolved.Imagesegmentationasanindispensablemethodisappliedtoextr

7、actquantitativeinformationofspecifictissues,alsoasthepremiseandasteptorealizetheprocessingofvisualization.Anditisstillachallengingproblemincomputervisionandimageprocessing.Areasobtainedbysegmentationarebothindependentfromeachother.Andalltheareasbelongtooneregionwillsharethe

8、specialconsistency.Theobjectiveofsegmentationistolocalizetheregionofeachtoothinden

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。