Centre dedocumentation
scientifique

Instance segmentation of standing dead trees in dense forest from aerial imagery using deep learning / Aboubakar Sani-Mohammed in ISPRS Open Journal of Photogrammetry and Remote Sensing, vol 6 (December 2022)  

Public [article]  inISPRS Open Journal of Photogrammetry and Remote Sensing > vol 6 (December 2022) . - n° 100024 Titre : Instance segmentation of standing dead trees in dense forest from aerial imagery using deep learning Type de document : Article/Communication Auteurs : Aboubakar Sani-Mohammed, Auteur ; Wei Yao, Auteur ; Marco Heurich, Auteur Année de publication : 2022 Article en page(s) : n° 100024 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] arbre mort [Termes IGN] Bavière (Allemagne) [Termes IGN] bois sur pied [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] détection automatique [Termes IGN] gestion forestière durable [Termes IGN] image à haute résolution [Termes IGN] image aérienne [Termes IGN] image infrarouge couleur [Termes IGN] peuplement mélangé [Termes IGN] puits de carbone [Termes IGN] segmentation sémantique Résumé : (auteur) Mapping standing dead trees, especially, in natural forests is very important for evaluation of the forest's health status, and its capability for storing Carbon, and the conservation of biodiversity. Apparently, natural forests have larger areas which renders the classical field surveying method very challenging, time-consuming, labor-intensive, and unsustainable. Thus, for effective forest management, there is the need for an automated approach that would be cost-effective. With the advent of Machine Learning, Deep Learning has proven to successfully achieve excellent results. This study presents an adjusted Mask R-CNN Deep Learning approach for detecting and segmenting standing dead trees in a mixed dense forest from CIR aerial imagery using a limited (195 images) training dataset. First, transfer learning is considered coupled with the image augmentation technique to leverage the limitation of training datasets. Then, we strategically selected hyperparameters to suit appropriately our model's architecture that fits well with our type of data (dead trees in images). Finally, to assess the generalization capability of our model's performance, a test dataset that was not confronted to the deep neural network was used for comprehensive evaluation. Our model recorded promising results reaching a mean average precision, average recall, and average F1-Score of 0.85, 0.88, and 0.87 respectively, despite our relatively low resolution (20 cm) dataset. Consequently, our model could be used for automation in standing dead tree detection and segmentation for enhanced forest management. This is equally significant for biodiversity conservation, and forest Carbon storage estimation. Numéro de notice : A2022-871 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article DOI : 10.1016/j.ophoto.2022.100024 Date de publication en ligne : 10/11/2022 En ligne : https://doi.org/10.1016/j.ophoto.2022.100024 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102165 [article]

Building detection with convolutional networks trained with transfer learning / Simon Šanca in Geodetski vestnik, vol 65 n° 4 (December 2021 - February 2022)  

Public [article]  inGeodetski vestnik > vol 65 n° 4 (December 2021 - February 2022) . - pp 559 - 576 Titre : Building detection with convolutional networks trained with transfer learning Type de document : Article/Communication Auteurs : Simon Šanca, Auteur ; Krištof Oštir, Auteur ; Alen Mangafić, Auteur Année de publication : 2021 Article en page(s) : pp 559 - 576 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] apprentissage profond [Termes IGN] classification automatique d'objets [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] détection du bâti [Termes IGN] données cadastrales [Termes IGN] image aérienne [Termes IGN] image infrarouge couleur [Termes IGN] image proche infrarouge [Termes IGN] image RVB [Termes IGN] orthoimage couleur [Termes IGN] segmentation d'image [Termes IGN] Slovénie Résumé : (Auteur) Building footprint detection based on orthophotos can be used to update the building cadastre. In recent years deep learning methods using convolutional neural networks have been increasingly used around the world. We present an example of automatic building classification using our datasets made of colour near-infrared orthophotos (NIR-R-G) and colour orthophotos (R-G-B). Building detection using pretrained weights from two large scale datasets Microsoft Common Objects in Context (MS COCO) and ImageNet was performed and tested. We applied the Mask Region Convolutional Neural Network (Mask R-CNN) to detect the building footprints. The purpose of our research is to identify the applicability of pre-trained neural networks on the data of another colour space to build a classification model without re-learning. Numéro de notice : A2021-930 Affiliation des auteurs : non IGN Thématique : IMAGERIE/URBANISME Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.15292/geodetski-vestnik.2021.04.559-593 Date de publication en ligne : 03/11/2021 En ligne : https://doi.org/10.15292/geodetski-vestnik.2021.04.559-593 Format de la ressource électronique : URL Article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99409 [article]

Réservation

Réserver ce document

Exemplaires (1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
139-2021041	SL	Revue	Centre de documentation	Revues en salle	Disponible

Super-resolution of VIIRS-measured ocean color products using deep convolutional neural network / Xiaoming Liu in IEEE Transactions on geoscience and remote sensing, vol 59 n° 1 (January 2021)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 59 n° 1 (January 2021) . - pp 114 - 127 Titre : Super-resolution of VIIRS-measured ocean color products using deep convolutional neural network Type de document : Article/Communication Auteurs : Xiaoming Liu, Auteur ; Menghua Wang, Auteur Année de publication : 2021 Article en page(s) : pp 114 - 127 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] analyse spectrale [Termes IGN] apprentissage profond [Termes IGN] bande infrarouge [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] couleur de l'océan [Termes IGN] image infrarouge couleur [Termes IGN] image multibande [Termes IGN] image NPP-VIIRS [Termes IGN] rayonnement Résumé : (auteur) Since its launch in October 2011, the Visible Infrared Imaging Radiometer Suite (VIIRS) onboard the Suomi National Polar-orbiting Partnership (SNPP) satellite has provided high quality global ocean color products, which include normalized water-leaving radiance spectra nLw ( λ ) of six moderate (M) bands (M1–M6) at the wavelengths of 410, 443, 486, 551, 671, and 745 nm with a spatial resolution of 750-m, and one imagery (I) band at a wavelength of 638 nm with a spatial resolution of 375-m. Because the high-resolution I-band measurements are highly correlated spectrally to those of M-band data, it can be used as a guidance to super-resolve the M-band nLw ( λ ) imagery from 750- to 375-m spatial resolution. Super-resolving images from coarse spatial resolution to finer ones have been a field of very active research in recent years. However, no previous studies have been applied to satellite ocean color remote sensing, in particular, for VIIRS ocean color applications. In this study, we employ the deep convolutional neural network (CNN) technique to glean the high-frequency content from the VIIRS I1 band and transfer to super-resolved M-band ocean color images. The network is trained to super-resolve each of the VIIRS six M-bands nLw ( λ ) separately. In our results, the super-resolved (375-m) nLw ( λ ) images are much sharper and show finer spatial structures than the original images. Quantitative evaluations show that biases between the super-resolved and original nLw ( λ ) images are small for all bands. However, errors in the super-resolved nLw ( λ ) images are wavelength-dependent. The smallest error is found in the super-resolved nLw (551) and nLw (671) images, and error increases as the wavelength decreases from 486 to 410 nm. The results show that the networks have the capability to capture the correlations of the M-band and the I1 band images to super-resolved M-band images. Numéro de notice : A2021-031 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2020.2992912 Date de publication en ligne : 20/05/2020 En ligne : https://doi.org/10.1109/TGRS.2020.2992912 Format de la ressource électronique : url article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=96726 [article]

Bretagne, la végétation cartographiée / Marielle Mayo in Géomètre, n° 2185 (novembre 2020)

Public [article]  inGéomètre > n° 2185 (novembre 2020) . - pp 46 - 49 Titre : Bretagne, la végétation cartographiée Type de document : Article/Communication Auteurs : Marielle Mayo, Auteur Année de publication : 2020 Article en page(s) : pp 46 - 49 Langues : Français (fre) Descripteur : [Vedettes matières IGN] Cartographie thématique [Termes IGN] 1:25.000 [Termes IGN] acquisition d'images [Termes IGN] aménagement régional [Termes IGN] ArcGIS [Termes IGN] BD ortho [Termes IGN] Bretagne [Termes IGN] carte de la végétation [Termes IGN] classification orientée objet [Termes IGN] données localisées [Termes IGN] données publiques [Termes IGN] IGN cité [Termes IGN] image infrarouge couleur [Termes IGN] image proche infrarouge [Termes IGN] modèle orienté objet [Termes IGN] phytoécologie Résumé : (Auteur) Une cartographie inédite de la végétation de Bretagne sera accessible en totalité en ligne en décembre. Produite par télédétection grâce à une méthode semi-automatisée innovante, elle répond aux nouveaux besoins des acteurs de la biodiversité et de l'aménagement du territoire. Numéro de notice : A2020-707 Affiliation des auteurs : non IGN Thématique : BIODIVERSITE/FORET/IMAGERIE Nature : Article nature-HAL : ArtSansCL DOI : sans Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=96281 [article]

Réservation

Réserver ce document

Exemplaires (1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
063-2020101	SL	Revue	Centre de documentation	Revues en salle	Disponible

Remotely sensing the species of individual trees / Yifang Shi (2020)  

Public Titre : Remotely sensing the species of individual trees Type de document : Thèse/HDR Auteurs : Yifang Shi, Auteur ; Andrew K. Skidmore, Directeur de thèse ; Tiejun Wang, Directeur de thèse Editeur : Enschede [Pays Bas] : University of Twente Année de publication : 2020 Collection : ITC Dissertation num. 376 Importance : 163 p. Format : 21 x 30 cm Note générale : bibliographie Doctor of Philosophy, Faculty of Geo-Information Science and Earth Observation, University of Twente Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications photogrammétriques [Termes IGN] Abies alba [Termes IGN] analyse comparative [Termes IGN] analyse diachronique [Termes IGN] Bavière (Allemagne) [Termes IGN] chlorophylle [Termes IGN] données lidar [Termes IGN] données localisées 3D [Termes IGN] forêt tempérée [Termes IGN] image hyperspectrale [Termes IGN] image infrarouge couleur [Termes IGN] inventaire forestier (techniques et méthodes) [Termes IGN] Leaf Mass per Area [Termes IGN] orthoimage Résumé : (auteur) The accurate identification of tree species is critical for the management of forest ecosystems. Mapping of tree species is an important task as it can assist a wide range of environmental applications, such as biodiversity monitoring, ecosystem services assessment, invasive species detection, and sustainable forest management. Compared to the conventional approaches based on labor-intensive field measurements, remote sensing has supplied a large variety of cutting-edge techniques to accomplish forest inventory. However, individual tree species classification in natural mixed forests, as it is typical in central Europe, is still a challenging task. High spectral and structural intra-species variability and inter-species similarity, due to phenological effects, differences in tree age and openness of canopies, shadowing effects, and environment variability, restrict tree species separability. An in-depth understanding of the relationship between species-specific features and remote sensing observations for tree species classification needs further investigation. This thesis aimed to accurately map the species of individual trees using multi-source remotely sensed data, including aerial photographs, airborne LiDAR and hyperspectral data. The research in the thesis firstly evaluated the performance of geometric and radiometric metrics from airborne LiDAR data under leaf-on and leaf-off conditions for individual tree species discrimination. The results empathized the importance of intensity-related LiDAR metrics for tree species identification under both leaf-on and leaf-off conditions. Then, the thesis examined whether multi-temporal digital CIR orthophotos could be used to further increase the accuracy of airborne LiDAR-based individual tree species mapping. The results showed that the texture features generated from multi-temporal digital CIR orthophotos under different view-illumination conditions are species-specific. Combining these texture features with LiDAR metrics significantly improved the accuracy of individual tree species mapping. To explore more valuable species-specific features, the thesis consequently integrated three plant functional traits (i.e. equivalent water thickness, leaf mass per area and leaf chlorophyll) retrieved from hyperspectral data with hyperspectral derived spectral features and airborne LiDAR derived metrics for mapping five tree species. Three selected plant functional traits were accurately retrieved using radiative transfer model and further improved the accuracy of tree species classification. Eventually, the thesis focused on an important tree species silver fir, and accurately mapped individuals of this species based on one-class classifiers using integrated airborne hyperspectral and LiDAR data. The mapping results provided the references locating the areas with a high occurrence probability of silver fir trees and hence increase the efficiency in subsequent field campaigns for forest management and biodiversity monitoring. This thesis explored the potential of various remotely sensed datasets for individual tree species mapping. The methodologies and findings in this thesis can be applied in the mapping of other tree species, which enriches the knowledge of species-specific characteristics and related remotely sensed signatures. The emerging of UAVs and the upcoming hyperspectral missions such as EnMAP and HySPIRI deliver valuable datasets with multi-scale coverage and revisit observations, which can be used for mapping the diversity of tree species at stand or regional level. Note de contenu : - General introduction - Important LiDAR metrics for discriminating tree species - Improving LiDAR-based tree species mapping using multi-temporal CIR orthophotos - Tree species classification using remotely sensed plant functional traits - Mapping individual silver fir trees in a Norway spruce dominated forest - Synthesis: Mapping individual tree species using multi-source remotely sensed data Numéro de notice : 17671 Affiliation des auteurs : non IGN Thématique : FORET Nature : Thèse étrangère Note de thèse : PhD thesis : : University of Twente : 2020 DOI : 10.3990/1.978903654953-0 Date de publication en ligne : 31/01/2020 En ligne : https://doi.org/10.3990/1.978903654953-0 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=97985

Area-based estimation of growing stock volume in Scots pine stands using ALS and airborne image-based point clouds / Paweł Hawryło in Forestry, an international journal of forest research, vol 90 n° 5 (December 2017)  

Permalink

Statistical atmospheric parameter retrieval largely benefits from spatial–spectral image compression / Joaquín García-Sobrino in IEEE Transactions on geoscience and remote sensing, vol 55 n° 4 (April 2017)  

Permalink

An operational high-resolution forest inventory / Julianno Sambatti in GIM international [en ligne], vol 30 n° 10 (October 2016)

Permalink

Wide-area mapping of small-scale features in agricultural landscapes using airborne remote sensing / Jerome O’Connell in ISPRS Journal of photogrammetry and remote sensing, vol 109 (November 2015)  

Permalink

Estimation of the mean tree height of forest stands by photogrammetric measurement using digital aerial images of high spatial resolution / Ivan Balenović in Annals of forest research, vol 58 n° 1 (January 2015)  

Permalink

Etude expérimentale en cartographie de la végétation par télédétection / Vanessa Sellin in Cybergeo, European journal of geography, n° 2015 ([01/01/2015])  

Permalink

Orthophotographie nocturne à haute résolution : la nuit, vue du ciel / Eva Frangiamone in Géomatique suisse, vol 112 n° 12 (décembre 2014)

Permalink

Use of handheld thermal imager data for airborne mapping of fire radiative power and energy and flame front rate of spread / Ronan Paugam in IEEE Transactions on geoscience and remote sensing, vol 51 n° 6 Tome 1 (June 2013)  

Permalink

Learning with transductive SVM for semisupervised pixel classification of remote sensing imagery / Ujjwal Maulik in ISPRS Journal of photogrammetry and remote sensing, vol 77 (March 2013)  

Permalink

Road network extraction in suburban areas / A. Grote in Photogrammetric record, vol 27 n° 137 (March - May 2012)  

Permalink