Centre dedocumentation
scientifique

An automatic approach for tree species detection and profile estimation of urban street trees using deep learning and Google street view images / Kwanghun Choi in ISPRS Journal of photogrammetry and remote sensing, vol 190 (August 2022)  

Public [article]  inISPRS Journal of photogrammetry and remote sensing > vol 190 (August 2022) . - pp 165 - 180 Titre : An automatic approach for tree species detection and profile estimation of urban street trees using deep learning and Google street view images Type de document : Article/Communication Auteurs : Kwanghun Choi, Auteur ; Wontaek LIM, Auteur ; Byungwoo Chang, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : pp 165 - 180 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] apprentissage profond [Termes IGN] arbre urbain [Termes IGN] détection automatique [Termes IGN] détection d'arbres [Termes IGN] diamètre à hauteur de poitrine [Termes IGN] gestion forestière durable [Termes IGN] image Streetview [Termes IGN] inventaire de la végétation [Termes IGN] segmentation sémantique [Termes IGN] Séoul Résumé : (auteur) Tree species and canopy structural profile (‘tree profile’) are among the most critical environmental factors in determining urban ecosystem services such as climate and air quality control from urban trees. To accurately characterize a tree profile, the tree diameter, height, crown width, and height to the lowest live branch must be all measured, which is an expensive and time-consuming procedure. Recent advances in artificial intelligence aids to efficiently and accurately measure the aforementioned tree profile parameters. This can be particularly helpful if spatially extensive and accurate street-level images provided by Google (‘streetview’) or Kakao (‘roadview’) are utilized. We focused on street trees in Seoul, the capital city of South Korea, and suggested a novel approach to create a tree profile and inventory based on deep learning algorithms. We classified urban tree species using the YOLO (You Only Look Once), one of the most popular deep learning object detection algorithms, which provides an uncomplicated method of creating datasets with custom classes. We further utilized semantic segmentation algorithm and graphical analysis to estimate tree profile parameters by determining the relative location of the interface of tree and ground surface. We evaluated the performance of the model by comparing the estimated tree heights, diameters, and locations from the model with the field measurements as ground truth. The results are promising and demonstrate the potential of the method for creating urban street tree profile inventory. In terms of tree species classification, the method showed the mean average precision (mAP) of 0.564. When we used the ideal tree images, the method also reported the normalized root mean squared error (NRMSE) for the tree height, diameter at breast height (DBH), and distances from the camera to the trees as 0.24, 0.44, and 0.41. Numéro de notice : A2022-503 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2022.06.004 Date de publication en ligne : 22/06/2022 En ligne : https://doi.org/10.1016/j.isprsjprs.2022.06.004 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101001 [article]

Five decades of ground flora changes in a temperate forest: The good, the bad and the ambiguous in biodiversity terms / K.J. Kirby in Forest ecology and management, vol 505 (1 February 2022)  

Public [article]  inForest ecology and management > vol 505 (1 February 2022) . - n° 119896 Titre : Five decades of ground flora changes in a temperate forest: The good, the bad and the ambiguous in biodiversity terms Type de document : Article/Communication Auteurs : K.J. Kirby, Auteur ; D.R. Bazely, Auteur ; E.A. Goldberg, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° 119896 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] analyse diachronique [Termes IGN] biomasse forestière [Termes IGN] Brachypodium (genre) [Termes IGN] Cervidae [Termes IGN] composition floristique [Termes IGN] dépérissement [Termes IGN] détection de changement [Termes IGN] eutrophisation [Termes IGN] flore forestière [Termes IGN] forêt tempérée [Termes IGN] Fraxinus excelsior [Termes IGN] gestion forestière [Termes IGN] maladie phytosanitaire [Termes IGN] richesse floristique [Termes IGN] Royaume-Uni [Termes IGN] Tracheophyta [Vedettes matières IGN] Foresterie Résumé : (auteur) We explore how the ground flora of a temperate woodland (Wytham Woods, southern England) changed in terms of species-richness, cover and biomass over five decades; what the drivers of change were; and possible future change as a consequence of the decline in Fraxinus excelsior as a canopy dominant. Vascular plants were recorded from 164 permanent, 10x10 m plots, distributed as a 141 m grid, in 1974, 1991, 1999, 2012, and 2018. Species presence and frequency/abundance in each plot were estimated and used to model biomass changes. Changes in species-richness, vegetation composition and structure were analysed. Stands opened out by thinning or which became denser through tree growth gained or lost species respectively, particularly non-woodland species. Deer pressure favoured the spread of Brachypodium sylvaticum and reduced Rubus fruticosus. No obvious impacts of climate change, eutrophication or of invasive species were detected in the plot records although other signs suggest these are starting to affect the flora. Just 12 out of 235 species contributed 47% of all species occurrences, 82% of the vegetation cover and 87% of the modelled biomass. We conclude that the ground flora is highly variable over decadal timescales, but the patterns of change observed differ according to the measures used (species richness, cover, biomass, etc). Site level drivers in the short-term swamped effects of slower acting regional/global drivers. Legacy effects were seen in the greater richness of specialists in the older woodland. While some impacts can be mitigated by management, others are largely beyond control at the site level. Numéro de notice : A2022-041 Affiliation des auteurs : non IGN Thématique : BIODIVERSITE/FORET Nature : Article DOI : 10.1016/j.foreco.2021.119896 Date de publication en ligne : 02/12/2021 En ligne : https://doi.org/10.1016/j.foreco.2021.119896 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99389 [article]

Progress on incorporating biodiversity monitoring in REDD+ through national forest inventories / Loïc Gillerot in Global ecology and conservation, vol 32 (December 2021)  

Public [article]  inGlobal ecology and conservation > vol 32 (December 2021) . - n° e01901 Titre : Progress on incorporating biodiversity monitoring in REDD+ through national forest inventories Type de document : Article/Communication Auteurs : Loïc Gillerot, Auteur ; Giorgio Grussu, Auteur ; Rocio Condor-Golec, Auteur ; et al., Auteur Année de publication : 2021 Article en page(s) : n° e01901 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] déboisement [Termes IGN] indicateur de biodiversité [Termes IGN] inventaire de la végétation [Termes IGN] inventaire forestier étranger (données) [Termes IGN] inventaire forestier national (données France) [Termes IGN] puits de carbone [Termes IGN] Réduction des émissions dues à la déforestation et la dégradation des forêts, REDD [Termes IGN] surveillance forestière [Termes IGN] télédétection [Vedettes matières IGN] Inventaire forestier Résumé : (auteur) There is a well-documented opportunity and need to incorporate biodiversity conservation priorities into REDD+ (Reducing Emissions from Deforestation and Forest Degradation) initiatives. This requires thorough monitoring of changes to biodiversity at appropriate temporal and spatial scales. A national forest inventory is one of the essential tools used to monitor carbon stock changes but can also be expanded to include biodiversity indicators. Here we analyse the progress and potential of 70 countries in monitoring primarily non-tree biodiversity using national forest inventories. Progress on national forest inventories among countries participating in REDD+ is variable: 11 countries have not started; 26 have started but do not include non-tree biodiversity indicators; the remaining 33 countries do include non-tree biodiversity indicators but use various methodological approaches, levels of detail and taxonomic groups. Very few of these provide comprehensive and accessible manuals or results, highlighting a need for greater transparency. The capacity of countries to fund ongoing national forest inventories is a constraining factor. Remote sensing technologies can help reduce costs for countries with limited monitoring capacity but the need to understand biodiversity variation at finer scales often limits the utility of such methods. Numéro de notice : A2021-866 Affiliation des auteurs : non IGN Thématique : BIODIVERSITE/FORET Nature : Article DOI : 10.1016/j.gecco.2021.e01901 Date de publication en ligne : 02/11/2021 En ligne : https://doi.org/10.1016/j.gecco.2021.e01901 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99096 [article]

Performance evaluation of artificial neural networks for natural terrain classification / Perpetual Hope Akwensi in Applied geomatics, vol 13 n° 1 (May 2021)  

Public [article]  inApplied geomatics > vol 13 n° 1 (May 2021) Titre : Performance evaluation of artificial neural networks for natural terrain classification Type de document : Article/Communication Auteurs : Perpetual Hope Akwensi, Auteur ; Eric Thompson Brantson, Auteur ; Johanna Ngula Niipele, Auteur ; et al., Auteur Année de publication : 2021 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de télédétection [Termes IGN] Afrique occidentale [Termes IGN] classification par nuées dynamiques [Termes IGN] échantillonnage [Termes IGN] fonction de base radiale [Termes IGN] image Landsat-OLI [Termes IGN] image multibande [Termes IGN] inventaire de la végétation [Termes IGN] réalité de terrain [Termes IGN] regroupement de données [Termes IGN] réseau neuronal artificiel [Termes IGN] segmentation d'image Résumé : (auteur) Remotely sensed image segmentation and classification form a very important part of remote sensing which involves geo-data processing and analysis. Artificial neural networks (ANNs) are powerful machine learning approaches that have been successfully implemented in numerous fields of study. There exist many kinds of neural networks and there is no single efficient approach for resolving all geospatial problems. Therefore, this research aims at investigating and evaluating the efficiency of three ANN approaches, namely, backpropagation neural network (BPNN), radial basis function neural network (RBFNN), and Elman backpropagation recurrent neural network (EBPRNN) using multi-spectral satellite images for terrain feature classification. Additionally, there has been close to no application of EBPRNN in modeling multi-spectral satellite images even though they also contain patterns. The efficiency of the three tested approaches is presented using the kappa coefficient, user’s accuracy, producer’s accuracy, overall accuracy, classification error, and computational simulation time. The study demonstrated that all the three ANN models achieved the aim of pattern identification, segmentation, and classification. This paper also discusses the observations of increasing sample sizes as inputs in the various ANN models. It was concluded that RBFNN’s computational time increases with increasing sample size and consequently increasing the number of hidden neurons; BPNN on overall attained the highest accuracy compared to the other models; EBPRNN’s accuracy increases with increasing sample size, hence a promising and perhaps an alternative choice to BPNN and RBFNN if very large datasets are involved. Based on the performance metrics used in this study, BPNN is the best model out of the three evaluated ANN models. Numéro de notice : A2021-223 Affiliation des auteurs : non IGN Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1007/s12518-021-00360-9 Date de publication en ligne : 13/02/2021 En ligne : https://doi.org/10.1007/s12518-021-00360-9 Format de la ressource électronique : url article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=97194 [article]

Déterminants de la composition floristique et estimations des stocks de carbone des peuplements forestiers matures de Uma (Tshopo, RDC) / John Katembo Mukirania (2021)  

Public Titre : Déterminants de la composition floristique et estimations des stocks de carbone des peuplements forestiers matures de Uma (Tshopo, RDC) Titre original : Determinants of floristic composition and estimates of carbon stocks in mature forest stands in Uma (Tshopo, DRC) Type de document : Thèse/HDR Auteurs : John Katembo Mukirania, Auteur ; Faustin Boyemba Bosela, Directeur de thèse ; Nicolas Barbier, Directeur de thèse Editeur : Montpellier : Centre de Coopération Internationale en recherche agronomique pour le Développement CIRAD Année de publication : 2021 Note générale : bibliographie thèse soutenue le 30 mars 2021 Langues : Français (fre) Descripteur : [Vedettes matières IGN] Applications de télédétection [Termes IGN] biomasse forestière [Termes IGN] composition floristique [Termes IGN] Congo [Termes IGN] dynamique de la végétation [Termes IGN] image Geoeye [Termes IGN] puits de carbone Résumé : (auteur) The study of tree assemblages in tropical forests is gaining new impetus with the need to assess carbon emissions at high precision and resolution, while limiting the erosion of diversity and promoting sustainable forest management. The objective of this study was to (i) investigate the respective roles of topographic / soil gradients and endogenous dynamics in shaping local variations in dominance; (ii) demonstrate the feasibility of studying canopy texture by harmonizing Fourier-based Textural Ordination (FOTO) indices of two GeoEye - 50 cm images, acquired from different phenologic seasons, to calibrate AGB inversion model using inventory plots. The study was conducted in Uma forest, East of Kisangani, Democratic Republic of Congo. Dataset of 30 1-ha plots, in which all trees above 10 cm diameter at 1.30 m height (DBH) were measured and identified. Standard physical and chemical properties of soil samples were determined (macro-nutrients, textural classes and pH) and a digital elevation model (SRTM 30 m) was used to infer relevant topographical features (altitude and hydromorphy). The forest in the study area is characterized by variations in the abundance of three dominant species: Petersianthus macrocarpus (P. BEAUV.) LIBEN, Gilbertiodendron dewevrei (De Wild.) J. Léonard and Julbernardia seretii (DE WILD.) TROUPIN, one non-pioneer, light demanding species and two late successional, shade tolerant species respectively. These variations occur nearly independently of variations in the substratum or topography, despite important gradients of the range of considered variables. Analyzing differential relative abundance of the three dominant species in the lower strata and in the canopy, did not provide evidence of shifts in dominance, in which a species would obviously tend to replace another through time in any of the three floristic groups. This suggests that in this study area the states of dominance in the vegetation are stable across generations, that successional dynamics are very slow or that they are localized to peculiar locations. Using FOTO method, this study documents a strong relation between observed and predicted AGBs, without cross validation (R² of the linear regression reached 0.82 (mean square error = 27.24 T/ha). This correlation was still present, although weaker, with cross validation (R² of the linear regression between observed and predicted AGBs = 0.64). The mean square error increases to 46.68 T/ha after cross validation for a mean of 450 T/ha. This result confirms the potential of FOTO indices of optical very high resolution satellite images to quantify aboveground biomass without no signal saturation in high AGB tropical forests. Numéro de notice : 17670 Affiliation des auteurs : non IGN Thématique : FORET Nature : Thèse étrangère Note de thèse : thèse : Ecologie et gestion des ressources végétales : Kisangani (République Démocratique du Congo) : 2021 Organisme de stage : UMR AMAP - Botanique et Modélisation de l'Architecture des Plantes et des Végétations nature-HAL : Thèse DOI : sans En ligne : https://hal.archives-ouvertes.fr/tel-03268307 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=97979

Apport des SIG dans le suivi des activités de reboisement dans la Forêt école de l’ENEF Mbalmayo / Daniele Roseline Bikie Mindang in Revue Internationale de Géomatique, Aménagement et Gestion des Ressources, RIGAGER, n° 7-8 (décembre 2020)  

Permalink

A convolutional neural network approach for counting and geolocating citrus-trees in UAV multispectral imagery / Lucas Prado Osco in ISPRS Journal of photogrammetry and remote sensing, vol 160 (February 2020)  

Permalink

Estimation et suivi de la ressource en bois en France métropolitaine par valorisation des séries multi-temporelles à haute résolution spatiale d'images optiques (Sentinel-2) et radar (Sentinel-1, ALOS-PALSAR) / David Morin (2020)  

Permalink

Faut-il des relevés de flore exhaustifs pour caractériser et cartographier l'acidité et les propriétés nutritionnelles des sols ? / Paulina E. Pinto in Rendez-vous techniques, n° 61-62 (hiver - printemps 2019) 

Permalink

Considering spatiotemporal processes in big data analysis: Insights from remote sensing of land cover and land use / Alexis Comber in Transactions in GIS, Vol 23 n° 5 (October 2019)  

Permalink

n° 58-59-60 - Special RENECOFOR - 25 ans de suivi des écosystèmes forestiers, bilan et perspectives (Bulletin de Rendez-vous techniques, n° 58-59-60 [01/09/2019])  

Permalink

A general method for the classification of forest stands using species composition and vertical and horizontal structure / Miquel de Cáceres in Annals of Forest Science [en ligne], vol 76 n° 2 (June 2019)  

Permalink

Changements du stock de bois sur pied des forêts françaises : description, analyse et simulation sur des horizons temporels pluri-décennal (1975 - 2015) et séculaire à partir des données de l'inventaire forestier national et de statistiques anciennes / Anaïs Denardou-Tisserand (2019)  

Permalink

Fuzzy modelling of growth potential in forest development simulation / Damjan Strnad in Ecological Informatics, vol 48 (November 2018)  

Permalink

How to calibrate historical aerial photographs : a change analysis of naturally dynamic boreal forest landscapes / Niko Kulha in Forests, vol 9 n° 10 (October 2018)  

Permalink