Centre dedocumentation
scientifique

A deep learning framework based on generative adversarial networks and vision transformer for complex wetland classification using limited training samples / Ali Jamali in International journal of applied Earth observation and geoinformation, vol 115 (December 2022)  

Public [article]  inInternational journal of applied Earth observation and geoinformation > vol 115 (December 2022) . - n° 103095 Titre : A deep learning framework based on generative adversarial networks and vision transformer for complex wetland classification using limited training samples Type de document : Article/Communication Auteurs : Ali Jamali, Auteur ; Masoud Mahdianpari, Auteur ; fariba Mohammadimanesh, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° 103095 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de télédétection [Termes IGN] apprentissage profond [Termes IGN] Canada [Termes IGN] carte thématique [Termes IGN] données d'entrainement (apprentissage automatique) [Termes IGN] image Sentinel-MSI [Termes IGN] image Sentinel-SAR [Termes IGN] réseau antagoniste génératif [Termes IGN] zone humide Résumé : (auteur) Wetlands have long been recognized among the most critical ecosystems globally, yet their numbers quickly diminish due to human activities and climate change. Thus, large-scale wetland monitoring is essential to provide efficient spatial and temporal insights for resource management and conservation plans. However, the main challenge is the lack of enough reference data for accurate large-scale wetland mapping. As such, the main objective of this study was to investigate the efficient deep-learning models for generating high-resolution and temporally rich training datasets for wetland mapping. The Sentinel-1 and Sentinel-2 satellites from the European Copernicus program deliver radar and optical data at a high temporal and spatial resolution. These Earth observations provide a unique source of information for more precise wetland mapping from space. The second objective was to investigate the efficiency of vision transformers for complex landscape mapping. As such, we proposed a 3D Generative Adversarial Network (3D GAN) to best achieve these two objectives of synthesizing training data and a Vision Transformer model for large-scale wetland classification. The proposed approach was tested in three different study areas of Saint John, Sussex, and Fredericton, New Brunswick, Canada. The results showed the ability of the 3D GAN to stimulate and increase the number of training data and, as a result, increase the accuracy of wetland classification. The quantitative results also demonstrated the capability of jointly using data augmentation, 3D GAN, and Vision Transformer models with overall accuracy, average accuracy, and Kappa index of 75.61%, 73.4%, and 71.87%, respectively, using a disjoint data sampling strategy. Therefore, the proposed deep learning method opens a new window for large-scale remote sensing wetland classification. Numéro de notice : A2022-828 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article DOI : 10.1016/j.jag.2022.103095 Date de publication en ligne : 08/11/2022 En ligne : https://doi.org/10.1016/j.jag.2022.103095 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102012 [article]

Beyond topo-climatic predictors: Does habitats distribution and remote sensing information improve predictions of species distribution models? / Arthur Sanguet in Global ecology and conservation, vol 39 (November 2022)  

Public [article]  inGlobal ecology and conservation > vol 39 (November 2022) . - n° e02286 Titre : Beyond topo-climatic predictors: Does habitats distribution and remote sensing information improve predictions of species distribution models? Type de document : Article/Communication Auteurs : Arthur Sanguet, Auteur ; Nicolas Wyler, Auteur ; Blaise Petitpierre, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° e02286 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] carte d'occupation du sol [Termes IGN] changement climatique [Termes IGN] distribution spatiale [Termes IGN] échantillonnage de données [Termes IGN] habitat (nature) [Termes IGN] modèle de simulation [Termes IGN] montagne [Termes IGN] pédologie locale [Termes IGN] Suisse [Termes IGN] télédétection [Termes IGN] topographie locale [Termes IGN] zone humide [Vedettes matières IGN] Ecologie forestière Résumé : (auteur) Species Distribution Models (SDM) represent a powerful tool to predict species’ habitat suitability on a landscape and fill the gap between truncated observation data and all possible locations. SDMs have been widely used in theoretical studies of species niches as well as in conservation applications. Here, we evaluated the impacts of predictors’ type on models’ performances and spatial predictions using 72 plant species belonging to six ecological groups at a regional scale in the area of Geneva (Switzerland). Twelve models were created using various combinations of high-resolution (25 m) explanatory variables including topography, pedology, climate, habitats and remote sensing data. Models integrating a combination of habitats and topopedo-climatic predictors had significantly higher performances, while remote sensing predictors showed low performances. Our results suggest that the number and the level of details of habitat predictors (broad or very precise) do not fundamentally affect prediction maps. However, selecting too few, overly simplified or exceedingly complex habitat predictors tend to lower models’ performances. The use of eight habitat categories complemented with eight topopedo-climatic predictors produced models with the highest performances. Ecological groups of species responded differently to models and while alpine and ruderal species have greater average performances due to a high affinity with topopedo-climatic predictors, wetlands’ species were less performant on average. These results underline the necessity of developing or having access to habitats distribution data especially in a conservation context. Numéro de notice : A2022-815 Affiliation des auteurs : non IGN Thématique : BIODIVERSITE/FORET Nature : Article DOI : 10.1016/j.gecco.2022.e02286 Date de publication en ligne : 13/09/2022 En ligne : https://doi.org/10.1016/j.gecco.2022.e02286 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101977 [article]

Heat wave-induced augmentation of surface urban heat islands strongly regulated by rural background / Shiqi Miao in Sustainable Cities and Society, vol 82 (July 2022)  

Public [article]  inSustainable Cities and Society > vol 82 (July 2022) . - n° 103874 Titre : Heat wave-induced augmentation of surface urban heat islands strongly regulated by rural background Type de document : Article/Communication Auteurs : Shiqi Miao, Auteur ; Wenfeng Zhan, Auteur ; Jiameng Lai, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° 103874 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de télédétection [Termes IGN] changement climatique [Termes IGN] Chine [Termes IGN] climat tropical [Termes IGN] couvert végétal [Termes IGN] densité de la végétation [Termes IGN] données environnementales [Termes IGN] forêt [Termes IGN] humidité de l'air [Termes IGN] ilot thermique urbain [Termes IGN] image Terra-MODIS [Termes IGN] nuit [Termes IGN] température au sol [Termes IGN] zone humide [Termes IGN] zone rurale Résumé : (auteur) The impact of heat waves (HWs) on surface urban heat islands (SUHIs) has been widely studied, but the spatial pattern of SUHI responsiveness to HWs across various climates remains unclear, and the influence of HW intensity on SUHI responsiveness has not been systematically quantified. Using MODIS land surface temperature data, here we investigated the responsiveness of SUHI to HWs (quantified as ∆I) as well as its variations with HW intensity in 354 cities in seven climate zones across China. We find that during HW periods, the SUHI and surface urban cool island are augmented in the humid and arid regions of China, respectively. The inter-city heterogeneity in rural vegetation coverage accounts for such a spatial pattern. In eastern China, the ∆I peaks in the north subtropical climate (0.72 ± 0.54 K for daytime and 0.29 ± 0.23 K for the nighttime) probably for its specific rural farming method. With the intensification of HWs, the augmentation effect can be further enhanced for the north subtropical, warm temperate, and arid temperate climates during the day and for almost all the climates at night. These findings can help advance the understanding of the responsiveness of SUHI to extreme climatic events. Numéro de notice : A2022-375 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article DOI : 10.1016/j.scs.2022.103874 Date de publication en ligne : 13/04/2022 En ligne : https://doi.org/10.1016/j.scs.2022.103874 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100624 [article]

The integration of multi-source remotely sensed data with hierarchically based classification approaches in support of the classification of wetlands / Aaron Judah in Canadian journal of remote sensing, vol 48 n° 2 (April 2022)  

Public [article]  inCanadian journal of remote sensing > vol 48 n° 2 (April 2022) . - pp 158 - 181 Titre : The integration of multi-source remotely sensed data with hierarchically based classification approaches in support of the classification of wetlands Type de document : Article/Communication Auteurs : Aaron Judah, Auteur ; Baoxin Hu, Auteur Année de publication : 2022 Article en page(s) : pp 158 - 181 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image mixte [Termes IGN] classification bayesienne [Termes IGN] classification par forêts d'arbres décisionnels [Termes IGN] image Landsat-8 [Termes IGN] image Landsat-TM [Termes IGN] image Sentinel-SAR [Termes IGN] intégration de données [Termes IGN] modèle numérique de terrain [Termes IGN] précision de la classification [Termes IGN] tourbière [Termes IGN] utilisation du sol [Termes IGN] zone humide Résumé : (auteur) Methodologies were developed to classify wetlands (Open Bog, Treed Bog, Open Fen, Treed Fen, and Swamps) from remotely sensed data using advanced classification algorithms through two hierarchical approaches. The data utilized included multispectral optical and thermal data (Landsat-5, and Landsat-8), radar imagery (Sentinel-1), and a digital elevation model. Goals were to determine the best way to combine imagery to classify wetlands through hierarchically based classification approaches to produce more accurate and efficient maps compared to standard classification. Algorithms used were Random Forest (RF), and Naïve Bayes. A hierarchically based RF classification methodology produced the most accurate classification result (91.94%). The hierarchically based approaches also improved classification accuracies for low-quality data, as defined through feature analysis, when compared to a nonhierarchical classifier. The hierarchical approaches also produced a significant increase in classification accuracy for the Naïve Bayes classifier versus the standard approach (∼12% increase) while not significantly increasing computation time – comparable in accuracy to the RF tests for around 20% the computational effort. Preselection of spectral bands, polarizations and other input parameters (Normalized Difference Vegetation Index, Normalized Difference Water Index, albedo, slope, etc.) using log-normal or RF variable importance analysis was very effective at identifying low-quality features and features which were of higher quality. Numéro de notice : A2022-372 Affiliation des auteurs : non IGN Thématique : IMAGERIE/INFORMATIQUE Nature : Article DOI : 10.1080/07038992.2021.1967732 Date de publication en ligne : 13/11/2021 En ligne : https://doi.org/10.1080/07038992.2021.1967732 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100614 [article]

Non-linear GNSS signal processing applied to land observation with high-rate airborne reflectometry / Hamza Issa (2022)  

Public Titre : Non-linear GNSS signal processing applied to land observation with high-rate airborne reflectometry Type de document : Thèse/HDR Auteurs : Hamza Issa, Auteur ; Serge Reboul, Directeur de thèse ; Ghaleb Faour, Directeur de thèse Editeur : Dunkerque : Université du Littoral-Côte-d'Opale Année de publication : 2022 Importance : 213 p. Format : 21 x 30 cm Note générale : Bibliographie Thèse en vue de l'obtention du grade de Docteur de l’Université du Littoral Côte d’Opale Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image radar et applications [Termes IGN] bande L [Termes IGN] capteur aérien [Termes IGN] données radar [Termes IGN] humidité du sol [Termes IGN] modèle statistique [Termes IGN] précision métrique [Termes IGN] rapport signal sur bruit [Termes IGN] réflectométrie par GNSS [Termes IGN] signal GNSS [Termes IGN] zone humide Index. décimale : THESE Thèses et HDR Résumé : (Auteur) Soil moisture remote sensing has been an active area of research over the past few decades due to its essential role in agriculture and in the prediction of some natural disasters. GNSS-Reflectometry (GNSS-R) is an emerging bistatic remote sensing technique that uses the L-band GNSS signals as sources of opportunity to characterize Earth surface. In this passive radar system, the amplitudes of the GNSS signal reflected by soil and the GNSS signal received directly from the GNSS satellites can be used to derive measurements of reflectivity from which the soil moisture content of the surface is determined.The study of soil moisture content using reflectivity measurements can also be applied for the detection of in-land water body surfaces. In this dissertation, we propose in the first step a non-linear estimate of the GNSS signal amplitude. This estimate is based on a statistical model that we develop for the coherent detection of a GNSS signal quantized on 1 bit. We show with experimentations on synthetic and real data that the proposed estimator is more accurate than reference approaches and provide measurements of the Signal-to-Noise Ratio (SNR) at a higher rate. When the reflected GNSS signal is obtained in an airborne experiment, its evolution as a function of time is piecewise stationary. The different stationary parts are associatedto different kinds of reflecting surfaces. We propose in a second step a change point detector that takes into account the radar signal characteristics in order to segment the signal. We show on synthetic data that the proposed change point detector can detect and localize changes more accurately than reference approaches present in the literature. This work is applied to airborne GNSSR observation of Earth. We propose in the third step, a new GNSS-R sensor with its implementation on a lightweight airborne carrier. We also propose a new front-end receiver architecture, a software radio implementation of thereceiver, and the complete instrumentation of the airborne carrier. A real flight experimentation has taken place in the North of France obtaining reflections from different landforms. We show using the airborne GNSS measurements obtained, that the proposed radar technique detects different surfaces along the flight trajectory, and in particular in-land water bodies, with high temporal and spatial resolution. We also show that we can localize the edges of the detected water body surfaces at meter accuracy. Note de contenu : General Introduction 1. Remote Sensing of Soil Moisture 1.1 Introduction 1.2 L-band emissions of land covers 1.3 Soil moisture remote sensing techniques 1.4 Remote sensing using GNSS-R 1.5 Conclusion 2. Carrier-to-Noise Estimation : Application to Soil Moisture Retrieval using GNSS-R 2.1 Introduction 2.2 Signal and system model 2.3 C/N0 estimators 2.4 Soil moisture retrieval from GNSS-R 2.5 Conclusion 3. A Probabilistic Model for On-line Estimation of the GNSS Carrier?to-Noise Ratio 3.1 Introduction 3.2 1-bit coherent detection principle 3.3 GNSS front end 3.4 Estimation of the GNSS signal amplitude 3.5 Experimentation 3.6 Conclusion 4. Segmentation of the GNSS Signal Amplitudes 4.1 Introduction 4.2 Change point detection principle 4.3 On-line/Off-line change detection system 4.4 Experimentation 4.5 Conclusion 5. Airborne GNSS Reflectometry for Water Body Detection 5.1 Introduction 5.2 Airborne GNSS system 5.3 Airborne experimental setup 5.4 GNSS-R software receiver 5.5 Flight Experimentation 5.6 Data analysis 5.7 Conclusion General Conclusion Numéro de notice : 26837 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse française Note de thèse : Thèse de Doctorat : Traitement du signal et des images : Université du Littoral Côte d’Opale : 2022 Organisme de stage : Laboratoire d'Informatique Signal et Image de la Côte d'Opale LISIC nature-HAL : Thèse DOI : sans Date de publication en ligne : 03/06/2022 En ligne : https://tel.archives-ouvertes.fr/tel-03687353/document Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101094

Age-dependence of stand biomass in managed boreal forests based on the Finnish National Forest Inventory data / Anna Repo in Forest ecology and management, vol 498 (October-15 2021)  

Permalink

Automatic detection of inland water bodies along altimetry tracks for estimating surface water storage variations in the Congo basin / Frédéric Frappart in Remote sensing, vol 13 n° 19 (October-1 2021)  

Permalink

Application of feature selection methods and machine learning algorithms for saltmarsh biomass estimation using Worldview-2 imagery / Sikdar M. M. Rasel in Geocarto international, vol 36 n° 10 ([01/06/2021])  

Permalink

Identifying the effects of chronic saltwater intrusion in coastal floodplain swamps using remote sensing / Elliott White Jr in Remote sensing of environment, vol 258 (June 2021)  

Permalink

A GIS- and AHP-based approach to map fire risk: a case study of Kuan Kreng peat swamp forest, Thailand / Narissara Nuthammachot in Geocarto international, vol 36 n° 2 ([01/02/2021])  

Permalink

Drought propagation and its impact on groundwater hydrology of wetlands: a case study on the Doode Bemde nature reserve (Belgium) / Buruk Kitachew Wossenyeleh in Natural Hazards and Earth System Sciences, vol 21 n° 1 (January 2021)  

Permalink

Polarimetric Synthetic Aperture Radar / Irena Hajnsek (2021)  

Permalink

Les synergies de la télédétection optique par drone et satellite / Emilien Alvarez-Vanhard (2021)  

Permalink

Boreal peatland forests: ditch network maintenance effort and water protection in a forest rotation framework / Jenny Miettinen in Canadian Journal of Forest Research, vol 50 n° 10 (October 2020)  

Permalink

Mapping wetland using the object-based stacked generalization method based on multi-temporal optical and SAR data / Yaotong Cai in International journal of applied Earth observation and geoinformation, vol 92 (October 2020)  

Permalink