Tree Species Classification Based on 3D Bark Texture Analysis - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2014

Tree Species Classification Based on 3D Bark Texture Analysis

Résumé

TerrestrialLaserScanning(TLS)techniqueistodaywidelyusedingroundplotstoacquire3Dpointcloudsfromwhichforestin-ventoryattributesarecalculated.Inthecaseofmixedplantingswherethe3Dpointcloudscontaindatafromseveraldifferenttreespecies,itisimportanttobeabletoautomaticallyrecognizethetreespeciesinordertoanalyzethedataofeachofthespeciesseparately.AlthoughautomatictreespeciesrecognitionfromTLSdataisanimportantproblem,ithasreceivedverylittleattentionfromthescientificcommunity.InthispaperweproposeamethodforclassifyingfivedifferenttreespeciesusingTLSdata.Ourmethodisbasedontheanalysisofthe3Dgeometrictextureofthebarkinordertocomputeroughnessmeasuresandshapecharacteris-ticsthatarefedasinputtoaRandomForestclassifiertoclassifythetreespecies.Themethodhasbeenevaluatedonatestsetcomposedof265samples(53samplesofeachofthe5species)andtheresultsobtainedareveryencouraging.
Fichier principal
Vignette du fichier
OTHMANI_PSIVT_2013.pdf (1.2 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01116701 , version 1 (14-02-2015)

Identifiants

Citer

Ahlem Othmani, Alexandre Piboule, Oscar Dalmau, Nicolas Lomenie, Said Mokrani, et al.. Tree Species Classification Based on 3D Bark Texture Analysis. 6th Pacific-Rim Symposium on Image and Video Technology (PSIVT 2013), Oct 2013, Guanajuato, Mexico. pp.279 - 289, ⟨10.1007/978-3-642-53842-1_24⟩. ⟨hal-01116701⟩
1173 Consultations
746 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More