Centre dedocumentation
scientifique

Deep learning for conifer/deciduous classification of airborne LiDAR 3D point clouds representing individual trees / Hamid Hamraz in ISPRS Journal of photogrammetry and remote sensing, Vol 158 (December 2019)  

Public [article]  inISPRS Journal of photogrammetry and remote sensing > Vol 158 (December 2019) . - pp 219 - 230 Titre : Deep learning for conifer/deciduous classification of airborne LiDAR 3D point clouds representing individual trees Type de document : Article/Communication Auteurs : Hamid Hamraz, Auteur ; Nathan B. Jacobs, Auteur ; Marco A. Contreras, Auteur ; Chase H. Clark, Auteur Année de publication : 2019 Article en page(s) : pp 219 - 230 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Lasergrammétrie [Termes descripteurs IGN] apprentissage profond [Termes descripteurs IGN] arbre caducifolié [Termes descripteurs IGN] classification par réseau neuronal convolutif [Termes descripteurs IGN] données d'apprentissage [Termes descripteurs IGN] données lidar [Termes descripteurs IGN] données localisées 3D [Termes descripteurs IGN] houppier [Termes descripteurs IGN] modèle numérique de surface [Termes descripteurs IGN] pinophyta [Termes descripteurs IGN] semis de points Résumé : (auteur) The purpose of this study was to investigate the use of deep learning for coniferous/deciduous classification of individual trees segmented from airborne LiDAR data. To enable processing by a deep convolutional neural network (CNN), we designed two discrete representations using leaf-off and leaf-on LiDAR data: a digital surface model with four channels (DSM × 4) and a set of four 2D views (4 × 2D). A training dataset of tree crowns was generated via segmentation of tree crowns, followed by co-registration with field data. Potential mislabels due to GPS error or tree leaning were corrected using a statistical ensemble filtering procedure. Because the training data was heavily unbalanced (~8% conifers), we trained an ensemble of CNNs on random balanced sub-samples. Benchmarked against multiple traditional shallow learning methods using manually designed features, the CNNs improved accuracies up to 14%. The 4 × 2D representation yielded similar classification accuracies to the DSM × 4 representation (~82% coniferous and ~90% deciduous) while converging faster. Further experimentation showed that early/late fusion of the channels in the representations did not affect the accuracies in a significant way. The data augmentation that was used for the CNN training improved the classification accuracies, but more real training instances (especially coniferous) likely results in much stronger improvements. Leaf-off LiDAR data were the primary source of useful information, which is likely due to the perennial nature of coniferous foliage. LiDAR intensity values also proved to be useful, but normalization yielded no significant improvement. As we observed, large training data may compensate for the lack of a subset of important domain data. Lastly, the classification accuracies of overstory trees (~90%) were more balanced than those of understory trees (~90% deciduous and ~65% coniferous), which is likely due to the incomplete capture of understory tree crowns via airborne LiDAR. In domains like remote sensing and biomedical imaging, where the data contain a large amount of information and are not friendly to human visual system, human-designed features may become suboptimal. As exemplified by this study, automatic, objective derivation of optimal features via deep learning can improve prediction tasks in such domains. Numéro de notice : A2019-547 Thématique : FORET/IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2019.10.011 date de publication en ligne : 03/11/2019 En ligne : https://doi.org/10.1016/j.isprsjprs.2019.10.011 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=94192 [article]

Réservation

Réserver ce document

Exemplaires (3)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
081-2019121	SL	Revue	Centre de documentation	Revues en salle	Disponible
081-2019123	DEP-RECP	Revue	MATIS	Dépôt en unité	Exclu du prêt
081-2019122	DEP-RECF	Revue	Nancy	Dépôt en unité	Exclu du prêt

Using a U-net convolutional neural network to map woody vegetation extent from high resolution satellite imagery across Queensland, Australia / Neil Flood in International journal of applied Earth observation and geoinformation, vol 82 (October 2019)  

Public [article]  inInternational journal of applied Earth observation and geoinformation > vol 82 (October 2019) . - 15 p. Titre : Using a U-net convolutional neural network to map woody vegetation extent from high resolution satellite imagery across Queensland, Australia Type de document : Article/Communication Auteurs : Neil Flood, Auteur ; Fiona Watson, Auteur ; Lisa Collett, Auteur Année de publication : 2019 Article en page(s) : 15 p. Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes descripteurs IGN] apprentissage profond [Termes descripteurs IGN] arbre (flore) [Termes descripteurs IGN] bois sur pied [Termes descripteurs IGN] Buisson [Termes descripteurs IGN] carte de la végétation [Termes descripteurs IGN] données d'apprentissage [Termes descripteurs IGN] image à haute résolution [Termes descripteurs IGN] image satellite [Termes descripteurs IGN] méthode de Monte-Carlo [Termes descripteurs IGN] mosaïque d'images [Termes descripteurs IGN] Queensland (Australie) [Termes descripteurs IGN] réseau neuronal convolutif [Termes descripteurs IGN] texture d'image Résumé : (auteur) Convolutional neural networks offer a new approach to classifying high resolution imagery. We use the U-net neural network architecture to map the presence or absence of trees and large shrubs across the Australian state of Queensland. From a state-wide mosaic of 1 m resolution 3-band Earth-i imagery, a selection of 827 squares (1 km2) are manually labeled for the presence of trees or large shrubs, and these are used to train the neural network. The training is intended to capture the textures which are primary visual cues of such vegetation. The trained neural network has an accuracy on independent data of around 90%. The resulting map over the whole of Queensland (1.73 million km2) is intended to be manually checked, and edited where necessary, to provide a high quality map of woody vegetation extent to serve a range of government policy objectives. Numéro de notice : A2019-474 Thématique : FORET/IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.jag.2019.101897 date de publication en ligne : 28/06/2019 En ligne : https://doi.org/10.1016/j.jag.2019.101897 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=93635 [article]

RoofN3D: a database for 3D building reconstruction with deep learning / Andreas Wichmann in Photogrammetric Engineering & Remote Sensing, PERS, vol 85 n° 6 (June 2019)  

Public [article]  inPhotogrammetric Engineering & Remote Sensing, PERS > vol 85 n° 6 (June 2019) . - pp 435 - 443 Titre : RoofN3D: a database for 3D building reconstruction with deep learning Type de document : Article/Communication Auteurs : Andreas Wichmann, Auteur ; Amgad Agoub, Auteur ; Valentina Schmidt, Auteur Année de publication : 2019 Article en page(s) : pp 435 - 443 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image [Termes descripteurs IGN] .Net [Termes descripteurs IGN] apprentissage profond [Termes descripteurs IGN] base de données localisées 3D [Termes descripteurs IGN] données d'apprentissage [Termes descripteurs IGN] reconstruction 3D du bâti [Termes descripteurs IGN] toit Résumé : (Auteur) Machine learning methods, in particular those based on deep learning, have gained in importance through the latest development of artificial intelligence and computer hardware. However, the direct application of deep learning methods to improve the results of 3D building reconstruction is often not possible due, for example, to the lack of suitable training data. To address this issue, we present RoofN3D which provides a three-dimensional (3D) point cloud training dataset that can be used to train machine learning models for different tasks in the context of 3D building reconstruction. The details about RoofN3D and the developed framework to automatically derive such training data are described in this paper. Furthermore, we provide an overview of other available 3D point cloud training data and approaches from current literature in which solutions for the application of deep learning to 3D point cloud data are presented. Finally, we exemplarily demonstrate how the provided data can be used to classify building roofs with the PointNet framework. Numéro de notice : A2019-248 Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.14358/PERS.85.6.435 En ligne : https://doi.org/10.14358/PERS.85.6.435 Format de la ressource électronique : URL Article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=93004 [article]

Réservation

Réserver ce document

Exemplaires (1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
105-2019061	SL	Revue	Centre de documentation	Revues en salle	Disponible