|
OALib Journal期刊
ISSN: 2333-9721
费用：99美元
|
|
|
|
基于主动轮廓技术的植物叶图像提取方法
DOI: 10.3969/j.jssn.1000-2006.2009.03.034, PP. 146-150
Keywords: 图像提取,可变形模板,叶脉络,细胞神经网络
Abstract:
叶是植物的重要特征信息，叶片图像提取是植物器官建模和生鲜植物识别的关键步骤。在植物自动识别和叶建模领域具有重要价值。笔者提出了一种基于主动轮廓技术和细胞神经网络的叶图像提取方法，实践表明基于细胞神经网络的可变模板技术实现了对植物叶片轮廓的灵活控制，同时结合了隐含模型和参数模型的特征，提高了提取的精度和鲁棒性。提取结果表明，采用该算法可以有效提取叶脉络。
| [1] | wangd,kerbysondj.realisticimagesynthesisofplantstructuresforgeneticanalysis[j].imageandvisioncomputing,2001,19:517-522.
|
| [2] | wangz,chiz,fengd.shapebasedleafimageretrieval,ieeproc？vis[j].imagesignalprocess,2003,150(1):34-43.
|
| [3] | fuhong.atwo？stageapproachforleafveinextraction,ieeeint.conf[j].neuralnetworks&signalprocessing,2003,208-211.
|
| [4] | kassm,witkina.snakes:activecontourmodels[j].internationaljournalofcomputervision,1988(1):321-331.
|
| [5] | vilarinodl,cabellod,pardoxm,etal.cellularneuralnetworksandactivecontours:atoolforimagesegmentation[j].imageandvisioncomputing,2003,21:189-204.
|
| [6] | chualo,yangl.cellularneuralnetworks:theory[j].ieeetransactionsoncircuitsandsystems,1988,35:1257-1273.
|
| [7] | zhuy,yanh.computerizedtumorboundarydetectionusingahopfieldneuralnetwork[j].ieeetransactionsonmedicalimaging,1997,16(1):55-67.
|
| [8] | casellesv,cattef,dibosf.ageometricmodelforactivecontoursinimageprocessing[j].numericalmaths,1993,66:1-31.
|
| [9] | vilarinodl.discrete？timecnnforimagesegmentationbyactivecontours[j].patternrecognitionletters,1998,19(8):721-734.
|
| [10] | markshammel,przemyslawprusinkiewicz.modellingcompoundleavesusingimplicitcontours[r].proceedingsofcomputergraphicsinternational.japan:tokyouniversity,1992.
|
| [11] | manhag,rabatelg,assematl,etal.weedleafimagesegmentationbydeformabletemplatesj[j].agricengngres,2001,80(2):139-146.
|
| [12] | mcinerneyt,terzopoulosd.adynamicfiniteelementsurfacemodelforsegmentationandtrackinginmultidimensionalmedicalimageswithapplicationtocardiac4dimageanalysis[j].journalofcomputersonmedicalimagingandgraphics,1995,19(1):69-83.
|
Full-Text
|
|
Contact Us
service@oalib.com QQ:3279437679 
WhatsApp +8615387084133
|
|
