Centre dedocumentation
scientifique

PPD: Pyramid Patch Descriptor via convolutional neural network / Jie Wan in Photogrammetric Engineering & Remote Sensing, PERS, vol 85 n° 9 (September 2019)  

Public [article]  inPhotogrammetric Engineering & Remote Sensing, PERS > vol 85 n° 9 (September 2019) . - pp 673 - 686 Titre : PPD: Pyramid Patch Descriptor via convolutional neural network Type de document : Article/Communication Auteurs : Jie Wan, Auteur ; Alper Yilmaz, Auteur ; Lei Yan, Auteur Année de publication : 2019 Article en page(s) : pp 673 - 686 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes descripteurs IGN] analyse comparative [Termes descripteurs IGN] appariement d'images [Termes descripteurs IGN] benchmark spatial [Termes descripteurs IGN] données localisées de référence [Termes descripteurs IGN] échantillonnage [Termes descripteurs IGN] état de l'art [Termes descripteurs IGN] image aérienne [Termes descripteurs IGN] image satellite [Termes descripteurs IGN] réseau neuronal convolutif Résumé : (Auteur) Local features play an important role in remote sensing image matching, and handcrafted features have been excessively used in this area for a long time. This article proposes a pyramid convolutional neural triplet network that extracts a 128-dimensional deep descriptor that significantly improves the matching performance. The proposed approach first extracts deep descriptors of the anchor patches and corresponding positive patches in a batch using the proposed pyramid convolutional neural network. Following this step, the approaches chooses the closest negative patch for each anchor patch and corresponding positive patch pair to form the triplet sample based on the descriptor distances among all other image patches in the batch. These triplets are used to optimize the parameters of the network using a new loss function. We evaluated the proposed deep descriptors on two benchmark data sets (Brown and HPatches) as well as real image data sets. The results reveal that the proposed descriptor achieves the state-of-the-art performance on the Brown data set and a comparatively very high performance on the HPatches data set. The proposed approach finds more correct matches than the classical handcrafted feature descriptors on aerial image pairs and is observed to be robust to variations in the viewpoint and illumination. Numéro de notice : A2019-416 Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.14358/PERS.85.9.673 date de publication en ligne : 01/09/2019 En ligne : https://doi.org/10.14358/PERS.85.9.673 Format de la ressource électronique : URL Article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=93543 [article]

Two contemporary and efficient two-stage sampling methods for estimating the volume of forest stands: a brief overview and unified mathematical description / Aristeidis Georgakis in Open journal of forestry, vol 9 n° 3 (July 2019)  

Public [article]  inOpen journal of forestry > vol 9 n° 3 (July 2019) . - pp 241 - 254 Titre : Two contemporary and efficient two-stage sampling methods for estimating the volume of forest stands: a brief overview and unified mathematical description Type de document : Article/Communication Auteurs : Aristeidis Georgakis, Auteur ; Georgios Stamatellos, Auteur Année de publication : 2019 Article en page(s) : pp 241 - 254 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes descripteurs IGN] échantillonnage [Termes descripteurs IGN] estimation statistique [Termes descripteurs IGN] inventaire forestier (techniques et méthodes) [Termes descripteurs IGN] peuplement forestier [Termes descripteurs IGN] surface terrière [Termes descripteurs IGN] volume en bois [Vedettes matières IGN] Inventaire forestier Résumé : (auteur) Big Basal Area Factor (Big BAF) and Point-3P are two-stage sampling methods. In the first stage the sampling units, in both methods, are Bitterlich points where the selection of the trees is proportional to their basal area. In the second stage, sampling units are trees which are a subset of the first stage trees. In the Big BAF method, the probability of selecting trees in the second stage is made proportional to the two BAFs’ ratio, with a basal area factor larger than that of the first stage. In the Point-3P method the probability of selecting trees, in the second stage, is based on the height prediction and use of a specific random number table. Estimates of the forest stands’ volume and their sampling errors are based on the theory of the product of two random variables. The increasing error in the second stage is small, but the total cost of measuring the trees is much smaller than simply using the first stage, with all the trees measured. In general, the two sampling methods are modern and cost-effective approaches that can be applied in forest stand inventories for forest management purposes and are receiving the growing interest of researchers in the current decade. Numéro de notice : A2019-381 Thématique : FORET/MATHEMATIQUE Nature : Article DOI : 10.4236/ojf.2019.93013 En ligne : https://doi.org/10.4236/ojf.2019.93013 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=93460 [article]

Virtual Support Vector Machines with self-learning strategy for classification of multispectral remote sensing imagery / Christian Geiss in ISPRS Journal of photogrammetry and remote sensing, vol 151 (May 2019)  

Public [article]  inISPRS Journal of photogrammetry and remote sensing > vol 151 (May 2019) . - pp 42 - 58 Titre : Virtual Support Vector Machines with self-learning strategy for classification of multispectral remote sensing imagery Type de document : Article/Communication Auteurs : Christian Geiss, Auteur ; Patrick Aravena Pelizari, Auteur ; Lukas Blickensdörfer, Auteur ; Hannes Taubenböck, Auteur Année de publication : 2019 Article en page(s) : pp 42 - 58 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes descripteurs IGN] analyse d'image orientée objet [Termes descripteurs IGN] apprentissage automatique [Termes descripteurs IGN] classification [Termes descripteurs IGN] classification par séparateurs à vaste marge [Termes descripteurs IGN] Cologne [Termes descripteurs IGN] échantillon [Termes descripteurs IGN] échantillonnage [Termes descripteurs IGN] image à très haute résolution [Termes descripteurs IGN] image multibande [Termes descripteurs IGN] invariant [Termes descripteurs IGN] Kenya [Termes descripteurs IGN] séparateur à vaste marge Résumé : (Auteur) We follow the idea of learning invariant decision functions for remote sensing image classification with Support Vector Machines (SVM). To do so, we generate artificially transformed samples (i.e., virtual samples) from available prior knowledge. Labeled samples closest to the separating hyperplane with maximum margin (i.e., the Support Vectors) are identified by learning an initial SVM model. The Support Vectors are used for generating virtual samples by perturbing the features to which the model should be invariant. Subsequently, the model is relearned using the Support Vectors and the virtual samples to eventually alter the hyperplane with maximum margin and enhance generalization capabilities of decision functions. In contrast to existing approaches, we establish a self-learning procedure to ultimately prune non-informative virtual samples from a possibly arbitrary invariance generation process to allow for robust and sparse model solutions. The self-learning strategy jointly considers a similarity and margin sampling constraint. In addition, we innovatively explore the invariance generation process in the context of an object-based image analysis framework. Image elements (i.e., pixels) are aggregated to image objects (as represented by segments/superpixels) with a segmentation algorithm. From an initial singular segmentation level, invariances are encoded by varying hyperparameters of the segmentation algorithm in terms of scale and shape. Experimental results are obtained from two very high spatial resolution multispectral data sets acquired over the city of Cologne, Germany, and the Hagadera Refugee Camp, Kenya. Comparative model accuracy evaluations underline the favorable performance properties of the proposed methods especially in settings with very few labeled samples. Numéro de notice : A2019-203 Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2019.03.001 date de publication en ligne : 12/03/2019 En ligne : https://doi.org/10.1016/j.isprsjprs.2019.03.001 Format de la ressource électronique : URL Article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=92666 [article]

Réservation

Réserver ce document

Exemplaires (3)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
081-2019051	SL	Revue	Centre de documentation	Revues en salle	Disponible
081-2019053	DEP-RECP	Revue	MATIS	Dépôt en unité	Exclu du prêt
081-2019052	DEP-RECF	Revue	Nancy	Dépôt en unité	Exclu du prêt

Is tree age or tree size reducing height increment in Abies alba Mill. at its southernmost distribution limit? / Pasquale A. Marziliano in Annals of Forest Science [en ligne], vol 76 n° 1 (March 2019)  

Public [article]  inAnnals of Forest Science [en ligne] > vol 76 n° 1 (March 2019) Titre : Is tree age or tree size reducing height increment in Abies alba Mill. at its southernmost distribution limit? Type de document : Article/Communication Auteurs : Pasquale A. Marziliano, Auteur ; Roberto Tognetti, Auteur ; Fabio Lombardi, Auteur Année de publication : 2019 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes descripteurs IGN] Abies alba [Termes descripteurs IGN] allométrie [Termes descripteurs IGN] Apennins [Termes descripteurs IGN] croissance végétale [Termes descripteurs IGN] dendrométrie [Termes descripteurs IGN] diamètre à hauteur de poitrine [Termes descripteurs IGN] échantillonnage [Termes descripteurs IGN] hauteur des arbres [Vedettes matières IGN] Sylviculture Résumé : (auteur) Key message : Conventional methods for estimating the current annual increment of stand volume are based on the uncertain assumption that height increment decreases with tree age. Conversely, size, rather than age, should be accounted for the observed senescence-related declines in relative growth rate and, consequently, implemented in silvicultural manuals. Results stem from a study on Abies alba Mill. at its southern limit of distribution. Context : Many factors limit height increment when age and size increase in large-statured tree species. Height–diameter allometric relationships are commonly used measures of tree growth. Aims : In this study, we tested if tree age was the main factor affecting the reduction in height increment of silver fir trees (Abies alba Mill.), verifying also whether tree size had a significant role in ecophysiological-biomechanical limitations to tree growth. Methods : The study was carried out in a silver fir forest located in Southern Italy, at the southernmost distribution limit for this species. Through a stratified random sampling, 100 trees were selected. All the selected trees were then felled and the total tree height, height increments (internode distances), diameter at breast height, and diameter increments (ring widths) were measured. Results : The analyses of allometric models and scaling coefficients showed that the correlation between tree age and height increment was not always significant. Conclusion : We may conclude that tree age did not statistically explain the decrease in height increment in older trees. Instead, the increase in tree size and related physiological processes (expressed as product between diameter at breast height and tree height) explained the reduction in height increment in older trees and was the main factor limiting height growth trends in marginal population of silver fir. Numéro de notice : A2019-028 Thématique : FORET Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1007/s13595-019-0803-5 date de publication en ligne : 13/02/2019 En ligne : https://doi.org/10.1007/s13595-019-0803-5 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=92036 [article]

A multi-faceted CNN architecture for automatic classification of mobile LiDAR data and an algorithm to reproduce point cloud samples for enhanced training / Bhavesh Kumar in ISPRS Journal of photogrammetry and remote sensing, vol 147 (January 2019)  

Public [article]  inISPRS Journal of photogrammetry and remote sensing > vol 147 (January 2019) . - pp 80 - 89 Titre : A multi-faceted CNN architecture for automatic classification of mobile LiDAR data and an algorithm to reproduce point cloud samples for enhanced training Type de document : Article/Communication Auteurs : Bhavesh Kumar, Auteur ; Gaurav Pandey, Auteur ; Bharat Lohani, Auteur ; Subhas C. Misra, Auteur Année de publication : 2019 Article en page(s) : pp 80 - 89 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Lasergrammétrie [Termes descripteurs IGN] apprentissage profond [Termes descripteurs IGN] classification automatique [Termes descripteurs IGN] données lidar [Termes descripteurs IGN] données localisées 3D [Termes descripteurs IGN] échantillonnage [Termes descripteurs IGN] image à haute résolution [Termes descripteurs IGN] précision de la classification [Termes descripteurs IGN] réseau neuronal convolutif Résumé : (auteur) Mobile Laser Scanning (MLS) data of outdoor environment are typically characterised by occlusion, noise, clutter, large data size and high quantum of information which makes their classification a challenging problem. This paper presents three deep Convolutional Neural Network (CNN) architectures in three dimension (3D), namely single CNN (SCN), multi-faceted CNN (MFC) and MFC with reproduction (MFCR) for automatic classification of MLS data. The MFC uses multiple facets of an MLS sample as inputs to different SCNs, thus providing additional information during classification. The MFC, once trained, is used to reproduce additional samples with the help of existing samples. The reproduced samples are employed to further refine the MFC training parameters, thus giving a new method called MFCR. The three architectures are evaluated on an ensemble of 3D outdoor MLS data consisting of four classes, i.e. tree, pole, house and ground covered with low vegetation along with car samples from KITTI dataset. The total accuracy and kappa values of classifications reached up to (i) 86.0% and 81.3% for the SCN (ii) 94.3% and 92.4% for the MFC and (iii) 96.0% and 94.6% for the MFCR, respectively. The paper has demonstrated the use of multiple facets to significantly improve classification accuracy over the SCN. Finally, a unique approach has been developed for reproduction of samples which has shown potential to improve the accuracy of classification. Unlike previous works on the use of CNN for structured point cloud of indoor objects, this work shows the utility of different proposed CNN architectures for classification of varieties of outdoor objects, viz., tree, pole, house and ground which are captured as unstructured point cloud by MLS. Numéro de notice : A2019-034 Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2018.11.006 date de publication en ligne : 20/11/2018 En ligne : https://doi.org/10.1016/j.isprsjprs.2018.11.006 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=91971 [article]

Réservation

Réserver ce document

Exemplaires (3)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
081-2019011	SL	Revue	Centre de documentation	Revues en salle	Disponible
081-2019013	DEP-EXM	Revue	MATIS	Dépôt en unité	Exclu du prêt
081-2019012	DEP-EAF	Revue	Nancy	Dépôt en unité	Exclu du prêt

Désambiguïsation des entités spatiales par apprentissage actif / Amal Chihaoui in Revue internationale de géomatique, vol 28 n° 2 (avril - juin 2018)  

Permalink

An (almost) automated process to track the Martians dunes : ac.GetPreciseShifts / Arthur Coqué (2018) 

Permalink

Tree size thresholds produce biased estimates of forest biomass dynamics / Eric B. Searle in Forest ecology and management, vol 400 (15 September 2017)  

Permalink

A review of sampling effects and response bias in internet participatory mapping (PPGIS/PGIS/VGI) / Greg Brown in Transactions in GIS, vol 21 n° 1 (February 2017)  

Permalink

Estimates of stem wood increments in forest resources: comparison of different approaches in forest inventory: consequences for international reporting: case study of European forests / Andrius Kuliesis in Annals of Forest Science [en ligne], vol 73 n° 4 (December 2016)  

Permalink

How many samples are needed? An investigation of binary logistic regression for selective omission in a road network / Qi Zhou in Cartography and Geographic Information Science, vol 43 n° 5 (November 2016)  

Permalink

Broad scale forest cover reconstruction from historical topographic maps / Dominik Kaim in Applied Geography, vol 67 (February 2016)  

Permalink

Statistical rigor in LiDAR-assisted estimation of aboveground forest biomass / Timothy G. Gregoire in Remote sensing of environment, vol 173 (February 2016)  

Permalink

Use of models in large-area forest surveys: comparing model-assisted, model-based and hybrid estimation / Göran Stahl in Forest ecosystems, vol 3 (2016)  

Permalink

How much do we know about the endangered Atlantic Forest? Reviewing nearly 70 years of information on tree community surveys / Renato A.F. de Lima in Biodiversity & Conservation, vol 24 n° 9 (September 2015)  

Permalink