Descripteur
Documents disponibles dans cette catégorie (9)
Ajouter le résultat dans votre panier
Visionner les documents numériques
Affiner la recherche Interroger des sources externesEtendre la recherche sur niveau(x) vers le bas
Fast forward feature selection of hyperspectral images for classification with gaussian mixture models / Mathieu Fauvel in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol 8 n° 6 (June 2015)
Titre : Fast forward feature selection of hyperspectral images for classification with gaussian mixture models Type de document : Article/Communication Auteurs : Mathieu Fauvel, Auteur ; Clément Dechesne Année de publication : 2015 Article en page(s) : pp 2824 - 2831 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] algorithme de Gauss
[Termes IGN] apprentissage automatique
[Termes IGN] classificateur
[Termes IGN] classification
[Termes IGN] extraction de traits caractéristiques
[Termes IGN] image hyperspectrale
[Termes IGN] itérationRésumé : (auteur) A fast forward feature selection algorithm is presented in this paper. It is based on a Gaussian mixture model (GMM) classifier. GMM are used for classifying hyperspectral images. The algorithm selects iteratively spectral features that maximizes an estimation of the classification rate. The estimation is done using the k-fold cross validation (k-CV). In order to perform fast in terms of computing time, an efficient implementation is proposed. First, the GMM can be updated when the estimation of the classification rate is computed, rather than re-estimate the full model. Secondly, using marginalization of the GMM, submodels can be directly obtained from the full model learned with all the spectral features. Experimental results for two real hyperspectral data sets show that the method performs very well in terms of classification accuracy and processing time. Furthermore, the extracted model contains very few spectral channels. Numéro de notice : A2015--068 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/JSTARS.2015.2441771 En ligne : http://dx.doi.org/10.1109/JSTARS.2015.2441771 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=83227
in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing > vol 8 n° 6 (June 2015) . - pp 2824 - 2831[article]
Titre : Gaussian processes uncertainty estimates in experimental Sentinel-2 LAI and leaf chlorophyll content retrieval Type de document : Article/Communication Auteurs : Jochem Verrlest, Auteur ; Juan Pablo Rivera, Auteur ; José Moreno, Auteur ; Gustavo Camps-Valls, Auteur Année de publication : 2013 Article en page(s) : pp 157 - 167 Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] algorithme de Gauss
[Termes IGN] apprentissage automatique
[Termes IGN] chlorophylle
[Termes IGN] image Sentinel-MSI
[Termes IGN] incertitude des données
[Termes IGN] indice foliaire
[Termes IGN] Leaf Area Index
[Termes IGN] régression
[Termes IGN] surveillance de la végétation
[Termes IGN] teneur en chlorophylle des feuilles
[Termes IGN] variable biophysique (végétation)Résumé : (Auteur) ESA’s upcoming Sentinel-2 (S2) Multispectral Instrument (MSI) foresees to provide continuity to land monitoring services by relying on optical payload with visible, near infrared and shortwave infrared sensors with high spectral, spatial and temporal resolution. This unprecedented data availability leads to an urgent need for developing robust and accurate retrieval methods, which ideally should provide uncertainty intervals for the predictions. Statistical learning regression algorithms are powerful candidats for the estimation of biophysical parameters from satellite reflectance measurements because of their ability to perform adaptive, nonlinear data fitting. In this paper, we focus on a new emerging technique in the field of Bayesian nonparametric modeling. We exploit Gaussian process regression (GPR) for retrieval, which is an accurate method that also provides uncertainty intervals along with the mean estimates. This distinct feature is not shared by other machine learning approaches. In view of implementing the regressor into operational monitoring applications, here the portability of locally trained GPR models was evaluated. Experimental data came from the ESA-led field campaign SPARC (Barrax, Spain). For various simulated S2 configurations (S2-10m, S2-20m and S2-60m) two important biophysical parameters were estimated: leaf chlorophyll content (LCC) and leaf area index (LAI). Local evaluation of an extended training dataset with more variation over bare soil sites led to improved LCC and LAI mapping with reduced uncertainties. GPR reached the 10% precision required by end users, with for LCC a NRMSE of 3.5–9.2% (r2: 0.95–0.99) and for LAI a NRMSE of 6.5–7.3% (r2: 0.95–0.96). The developed GPR models were subsequently applied to simulated Sentinel images over various sites. The associated uncertainty maps proved to be a good indicator for evaluating the robustness of the retrieval performance. The generally low uncertainty intervals over vegetated surfaces suggest that the locally trained GPR models are portable to other sites and conditions. Numéro de notice : A2013-708 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2013.09.012 En ligne : https://doi.org/10.1016/j.isprsjprs.2013.09.012 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=32844
in ISPRS Journal of photogrammetry and remote sensing > vol 86 (December 2013) . - pp 157 - 167[article]
Titre : Active learning methods for biophysical parameter estimation Type de document : Article/Communication Auteurs : Edoardo Pasolli, Auteur ; F. Melgani, Auteur ; N. Alajlan, Auteur ; B. Yakoub, Auteur Année de publication : 2012 Article en page(s) : pp 4071 - 4084 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Intelligence artificielle
[Termes IGN] algorithme de Gauss
[Termes IGN] apprentissage automatique
[Termes IGN] chlorophylle
[Termes IGN] régression
[Termes IGN] séparateur à vaste marge
[Termes IGN] variable biophysique (végétation)Résumé : (Auteur) In this paper, we face the problem of collecting training samples for regression problems under an active learning perspective. In particular, we propose various active learning strategies specifically developed for regression approaches based on Gaussian processes (GPs) and support vector machines (SVMs). For GP regression, the first two strategies are based on the idea of adding samples that are dissimilar from the current training samples in terms of covariance measure, while the third one uses a pool of regressors in order to select the samples with the greater disagreements between the different regressors. Finally, the last strategy exploits an intrinsic GP regression outcome to pick up the most difficult and hence interesting samples to label. For SVM regression, the method based on the pool of regressors and two additional strategies based on the selection of the samples distant from the current support vectors in the kernel-induced feature space are proposed. The experimental results obtained on simulated and real data sets show that the proposed strategies exhibit a good capability to select samples that are significant for the regression process, thus opening the way to the active learning approach for remote-sensing regression problems. Numéro de notice : A2012-528 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE/INFORMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2012.2187906 Date de publication en ligne : 17/04/2012 En ligne : https://doi.org/10.1109/TGRS.2012.2187906 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=31974
in IEEE Transactions on geoscience and remote sensing > vol 50 n° 10 Tome 2 (October 2012) . - pp 4071 - 4084[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 065-2012101B MANQUANT Revue Centre de documentation Indéterminé Disponible
Titre : An unbiased algorithm for detection of curvilinear structures in urban remote sensing images Type de document : Article/Communication Auteurs : Jinzheng Peng, Auteur ; Ya-Qiu Jin, Auteur Année de publication : 2007 Article en page(s) : pp 5377 - 5395 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image
[Termes IGN] algorithme de Gauss
[Termes IGN] courbe
[Termes IGN] extraction automatique
[Termes IGN] extraction du réseau routier
[Termes IGN] image aérienne
[Termes IGN] objet géographique linéaireRésumé : (Auteur) Based on the Gaussian scale-space, a Gaussian comparison function is presented for extracting the linearly road features in aerial remote sensing image. Combining the geometric and radiometric features, the curvilinear structures of the roads are extracted based on locally oriented energy in continuous scale-space. Curvilinear features of roads are verified, grouped and extracted by using both topologic and geometric methods. This algorithm is applicable to extracting the road features in different scale such as rural roads or urban highways, and significantly reduces the computation complexity of line tracing. Some discussions on the zero drift of the Gaussian comparison function are also presented. Copyright Taylor & Francis Numéro de notice : A2007-537 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1080/01431160601075574 En ligne : https://doi.org/10.1080/01431160601075574 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=28900
in International Journal of Remote Sensing IJRS > vol 28 n°23-24 (December 2007) . - pp 5377 - 5395[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 080-07131 RAB Revue Centre de documentation En réserve L003 Disponible
Titre : A sub-pixel location method for interest points by means of the Harris interest strength Type de document : Article/Communication Auteurs : Q. Zhu, Auteur ; B. Wu, Auteur Année de publication : 2007 Article en page(s) : pp 321 - 335 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Photogrammétrie numérique
[Termes IGN] algorithme de Gauss
[Termes IGN] analyse infrapixellaire
[Termes IGN] géoréférencement
[Termes IGN] méthode des moindres carrés
[Termes IGN] niveau de gris (image)
[Termes IGN] point d'intérêt
[Termes IGN] reconstruction 3DRésumé : (Auteur) The sub-pixel location of interest points is one of the most important tasks in refined image-based 3D reconstruction in digital photogrammetry. The interest point detectors based on the Harris principles are generally used for stereoscopic image matching and subsequent 3D reconstruction. However, the locations of the interest points detected in this way can only be obtained to 1 pixel accuracy. The Harris detector has the following characteristics: (1) the Harris interest strength, which denotes the distinctiveness of an interest point, is a grey scale descriptor which computes the gradient at each sample point in a region around the point, and (2) the Harris interest strengths of the pixels in a template window centred on the interest point exhibit an approximately paraboloid distribution. This paper proposes a precise location method to improve the precision of the interest points on the basis of these characteristics of the Harris interest strength. Firstly, a least squares fit of a paraboloid function to the image grey scale surface using the Harris interest strength is designed in a template window and a Gaussian-distance algorithm is employed to determine the weight. Then, the precise coordinates of this interest point are obtained by calculating the extremities of the fitting surface. The location accuracy of this method is studied both from the theoretical and the practical point of view. Experimental analysis is illustrated with synthetic images as well as actual images, which yielded a location accuracy of 0·15 pixels. Furthermore, experimental results also indicate that this method has the desired anti-image-noise and efficiency characteristics. Numéro de notice : A2007-568 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article DOI : 10.1111/j.1477-9730.2007.00450.x En ligne : https://doi.org/10.1111/j.1477-9730.2007.00450.x Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=28931
in Photogrammetric record > vol 22 n° 120 (December 2007 - February 2008) . - pp 321 - 335[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 106-07041 Revue Centre de documentation Revues en salle Disponible PermalinkPermalinkPermalinkPermalink
