Centre de documentation
scientifique

Dépouillements

Dimensionality reduction and classification of hyperspectral images using ensemble discriminative local metric learning / Yanni Dong in IEEE Transactions on geoscience and remote sensing, vol 55 n° 5 (May 2017)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 55 n° 5 (May 2017) . - pp 2509 - 2524 Titre : Dimensionality reduction and classification of hyperspectral images using ensemble discriminative local metric learning Type de document : Article/Communication Auteurs : Yanni Dong, Auteur ; Bo Du, Auteur ; Liangpei Zhang, Auteur ; Lefei Zhang, Auteur Année de publication : 2017 Article en page(s) : pp 2509 - 2524 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] apprentissage dirigé [Termes IGN] classification [Termes IGN] image hyperspectrale [Termes IGN] réduction Résumé : (Auteur) The high-dimensional data space of hyperspectral images (HSIs) often result in ill-conditioned formulations, which finally leads to many of the high-dimensional feature spaces being empty and the useful data existing primarily in a subspace. To avoid these problems, we use distance metric learning for dimensionality reduction. The goal of distance metric learning is to incorporate abundant discriminative information by reducing the dimensionality of the data. Considering that global metric learning is not appropriate for all training samples, this paper proposes an ensemble discriminative local metric learning (EDLML) algorithm for HSI analysis. The EDLML algorithm learns robust local metrics from both the training samples and the relative neighborhood of them and considers the different local discriminative distance metrics by dealing with the data region by region. It aims to learn a subspace to keep all the samples in the same class are as near as possible, while those from different classes are separated. The learned local metrics are then used to build an ensemble metric. Experiments on a number of different hyperspectral data sets confirm the effectiveness of the proposed EDLML algorithm compared with that of the other dimension reduction methods. Numéro de notice : A2017-465 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2016.2645703 En ligne : http://dx.doi.org/10.1109/TGRS.2016.2645703 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=86388 [article]

Superpixel-based multitask learning framework for hyperspectral image classification / Sen Jia in IEEE Transactions on geoscience and remote sensing, vol 55 n° 5 (May 2017)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 55 n° 5 (May 2017) . - pp 2575 - 2588 Titre : Superpixel-based multitask learning framework for hyperspectral image classification Type de document : Article/Communication Auteurs : Sen Jia, Auteur ; Bin Deng, Auteur ; Jiasong Zhu, Auteur ; Xiuping Jia, Auteur Année de publication : 2017 Article en page(s) : pp 2575 - 2588 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] apprentissage automatique [Termes IGN] classification par séparateurs à vaste marge [Termes IGN] filtre de Gabor [Termes IGN] image hyperspectrale Résumé : (Auteur) Due to the high spectral dimensionality of hyperspectral images as well as the difficult and time-consuming process of collecting sufficient labeled samples in practice, the small sample size scenario is one crucial problem and a challenging issue for hyperspectral image classification. Fortunately, the structure information of materials, reflecting region of homogeneity in the spatial domain, offers an invaluable complement to the spectral information. Assuming some spatial regularity and locality of surface materials, it is reasonable to segment the image into different homogeneous parts in advance, called superpixel, which can be used to improve the classification performance. In this paper, a superpixel-based multitask learning framework has been proposed for hyperspectral image classification. Specifically, a set of 2-D Gabor filters are first applied to hyperspectral images to extract discriminative features. Meanwhile, a superpixel map is generated from the hyperspectral images. Second, a superpixel-based spatial-spectral Schroedinger eigenmaps (S4E) method is adopted to effectively reduce the dimensions of each extracted Gabor cube. Finally, the classification is carried out by a support vector machine (SVM)-based multitask learning framework. The proposed approach is thus termed Gabor S4E and SVM-based multitask learning (GS4E-MTLSVM). A series of experiments is conducted on three real hyperspectral image data sets to demonstrate the effectiveness of the proposed GS4E-MTLSVM approach. The experimental results show that the performance of the proposed GS4E-MTLSVM is better than those of several state-of-the-art methods, while the computational complexity has been greatly reduced, compared with the pixel-based spatial-spectral Schroedinger eigenmaps method. Numéro de notice : A2017-466 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2017.2647815 En ligne : http://dx.doi.org/10.1109/TGRS.2017.2647815 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=86389 [article]

Self-taught feature learning for hyperspectral image classification / Ronald Kemker in IEEE Transactions on geoscience and remote sensing, vol 55 n° 5 (May 2017)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 55 n° 5 (May 2017) . - pp 2693 - 2705 Titre : Self-taught feature learning for hyperspectral image classification Type de document : Article/Communication Auteurs : Ronald Kemker, Auteur ; Christopher Kanan, Auteur Année de publication : 2017 Article en page(s) : pp 2693 - 2705 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] analyse en composantes indépendantes [Termes IGN] apprentissage automatique [Termes IGN] classification dirigée [Termes IGN] extraction de traits caractéristiques [Termes IGN] filtrage numérique d'image [Termes IGN] image hyperspectrale Résumé : (Auteur) In this paper, we study self-taught learning for hyperspectral image (HSI) classification. Supervised deep learning methods are currently state of the art for many machine learning problems, but these methods require large quantities of labeled data to be effective. Unfortunately, existing labeled HSI benchmarks are too small to directly train a deep supervised network. Alternatively, we used self-taught learning, which is an unsupervised method to learn feature extracting frameworks from unlabeled hyperspectral imagery. These models learn how to extract generalizable features by training on sufficiently large quantities of unlabeled data that are distinct from the target data set. Once trained, these models can extract features from smaller labeled target data sets. We studied two self-taught learning frameworks for HSI classification. The first is a shallow approach that uses independent component analysis and the second is a three-layer stacked convolutional autoencoder. Our models are applied to the Indian Pines, Salinas Valley, and Pavia University data sets, which were captured by two separate sensors at different altitudes. Despite large variation in scene type, our algorithms achieve state-of-the-art results across all the three data sets. Numéro de notice : A2017-467 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2017.2651639 En ligne : https://doi.org/10.1109/TGRS.2017.2651639 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=86390 [article]

Evaluation of the range accuracy and the radiometric calibration of multiple terrestrial laser scanning instruments for data interoperability / Kim Calders in IEEE Transactions on geoscience and remote sensing, vol 55 n° 5 (May 2017)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 55 n° 5 (May 2017) . - pp 2716 - 2724 Titre : Evaluation of the range accuracy and the radiometric calibration of multiple terrestrial laser scanning instruments for data interoperability Type de document : Article/Communication Auteurs : Kim Calders, Auteur ; Mathias I. Disney, Auteur ; John Armston, Auteur ; et al., Auteur Année de publication : 2017 Article en page(s) : pp 2716 - 2724 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Lasergrammétrie [Termes IGN] acquisition de données [Termes IGN] analyse comparative [Termes IGN] étalonnage radiométrique [Termes IGN] évaluation des données [Termes IGN] interopérabilité sémantique [Termes IGN] précision métrique [Termes IGN] réflectance [Termes IGN] télémètre laser terrestre Résumé : (Auteur) Terrestrial laser scanning (TLS) data provide 3-D measurements of vegetation structure and have the potential to support the calibration and validation of satellite and airborne sensors. The increasing range of different commercial and scientific TLS instruments holds challenges for data and instrument interoperability. Using data from various TLS sources will be critical to upscale study areas or compare data. In this paper, we provide a general framework to compare the interoperability of TLS instruments. We compare three TLS instruments that are the same make and model, the RIEGL VZ-400. We compare the range accuracy and evaluate the manufacturer's radiometric calibration for the uncalibrated return intensities. Our results show that the range accuracy between instruments is comparable and within the manufacturer's specifications. This means that the spatial XYZ data of different instruments can be combined into a single data set. Our findings demonstrate that radiometric calibration is instrument specific and needs to be carried out for each instrument individually before including reflectance information in TLS analysis. We show that the residuals between the calibrated reflectance panels and the apparent reflectance measured by the instrument are greatest for highest reflectance panels (residuals ranging from 0.058 to 0.312). Numéro de notice : A2017-468 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2017.2652721 En ligne : http://dx.doi.org/10.1109/TGRS.2017.2652721 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=86391 [article]

Hyperspectral and lidar intensity data fusion : A framework for the rigorous correction of illumination, anisotropic effects, and cross calibration / Maximilian Brell in IEEE Transactions on geoscience and remote sensing, vol 55 n° 5 (May 2017)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 55 n° 5 (May 2017) . - pp 2799 - 2810 Titre : Hyperspectral and lidar intensity data fusion : A framework for the rigorous correction of illumination, anisotropic effects, and cross calibration Type de document : Article/Communication Auteurs : Maximilian Brell, Auteur ; Karl Segl, Auteur ; Luis Guanter, Auteur ; Bodo Bookhagen, Auteur Année de publication : 2017 Article en page(s) : pp 2799 - 2810 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications photogrammétriques [Termes IGN] anisotropie [Termes IGN] correction radiométrique [Termes IGN] données lidar [Termes IGN] données localisées 3D [Termes IGN] étalonnage croisé [Termes IGN] fusion d'images [Termes IGN] image hyperspectrale [Termes IGN] intensité lumineuse [Termes IGN] réflectance spectrale Résumé : (Auteur) The fusion of hyperspectral imaging (HSI) sensor and airborne lidar scanner (ALS) data provides promising potential for applications in environmental sciences. Standard fusion approaches use reflectance information from the HSI and distance measurements from the ALS to increase data dimensionality and geometric accuracy. However, the potential for data fusion based on the respective intensity information of the complementary active and passive sensor systems is high and not yet fully exploited. Here, an approach for the rigorous illumination correction of HSI data, based on the radiometric cross-calibrated return intensity information of ALS data, is presented. The cross calibration utilizes a ray tracing-based fusion of both sensor measurements by intersecting their particular beam shapes. The developed method is capable of compensating for the drawbacks of passive HSI systems, such as cast and cloud shadowing effects, illumination changes over time, across track illumination, and partly anisotropy effects. During processing, spatial and temporal differences in illumination patterns are detected and corrected over the entire HSI wavelength domain. The improvement in the classification accuracy of urban and vegetation surfaces demonstrates the benefit and potential of the proposed HSI illumination correction. The presented approach is the first step toward the rigorous in-flight fusion of passive and active system characteristics, enabling new capabilities for a variety of applications. Numéro de notice : A2017-469 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2017.2654516 En ligne : https://doi.org/10.1109/TGRS.2017.2654516 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=86392 [article]

Baltic sea ice concentration estimation using SENTINEL-1 SAR and AMSR2 microwave radiometer data / Juha Karvonen in IEEE Transactions on geoscience and remote sensing, vol 55 n° 5 (May 2017)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 55 n° 5 (May 2017) . - pp 2871 - 2883 Titre : Baltic sea ice concentration estimation using SENTINEL-1 SAR and AMSR2 microwave radiometer data Type de document : Article/Communication Auteurs : Juha Karvonen, Auteur Année de publication : 2017 Article en page(s) : pp 2871 - 2883 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image radar et applications [Termes IGN] analyse comparative [Termes IGN] Baltique, mer [Termes IGN] épaisseur de la glace [Termes IGN] glace de mer [Termes IGN] image Aqua-AMSR [Termes IGN] image Sentinel-SAR [Termes IGN] navigation maritime [Termes IGN] Sentinel-1 [Termes IGN] télédétection en hyperfréquence Résumé : (Auteur) Sea ice concentration (SIC) is an important sea ice parameter for sea ice navigation, environmental research, and weather and ice forecasting. We have developed and tested a method for estimation of the Baltic Sea SIC using SENTINEL-1 synthetic aperture radar (SAR) and Advanced Microwave Scanning Radiometer 2 passive microwave radiometer (MWR) data. Here, we present the method and results for January 2016. Ice concentration grids of Finnish Meteorological Institute daily ice charts have been used as reference data in this paper. We present a comparison of four SIC estimation methods with our reference data. In addition to the combined SAR/MWR SIC estimation method, we also compare SIC estimates produced using SAR alone and two MWR-based methods. The main target of this paper was to develop and test a high-resolution SIC estimation method suitable for operational use. Numéro de notice : A2017-470 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2017.2655567 En ligne : http://dx.doi.org/10.1109/TGRS.2017.2655567 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=86393 [article]

An internal crown geometric model for conifer species classification with high-density LiDAR data / Aravind Harikumar in IEEE Transactions on geoscience and remote sensing, vol 55 n° 5 (May 2017)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 55 n° 5 (May 2017) . - pp 2924 - 2940 Titre : An internal crown geometric model for conifer species classification with high-density LiDAR data Type de document : Article/Communication Auteurs : Aravind Harikumar, Auteur ; Francesca Bovolo, Auteur ; Lorenzo Bruzzone, Auteur Année de publication : 2017 Article en page(s) : pp 2924 - 2940 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Lasergrammétrie [Termes IGN] analyse en composantes principales [Termes IGN] classification dirigée [Termes IGN] données lidar [Termes IGN] données localisées 3D [Termes IGN] forêt [Termes IGN] houppier [Termes IGN] modèle de croissance végétale [Termes IGN] modèle géométrique [Termes IGN] Pinophyta [Termes IGN] structure d'un peuplement forestier [Termes IGN] Trente Résumé : (Auteur) The knowledge of the tree species is a crucial information that governs the success of precision forest management practice. High-density small footprint multireturn airborne light detection and ranging (LiDAR) scanning can collect a huge amount of point samples containing structural details of the forest vertical profile, which can reveal important structural information of the forest components. LiDAR data have been successfully used to distinguish between coniferous and deciduous/broadleaved tree species. However, species classification within a class (e.g., the conifer class) using LiDAR data is a challenging problem when considering the tree external crown characteristics only. This paper presents a novel method for conifer species classification based on the use of geometric features describing both the internal and external structures of the crown. The internal crown geometric features (IGFs) are defined based on a novel internal branch structure model, which uses 3-D region growing and principal component analysis to delineate the branch structure of a conifer tree accurately. IGFs are used together with external crown geometric features to perform conifer species classification. Three different support vector machines have been considered for classification performance evaluation. The experimental analysis conducted on high-density LiDAR data acquired over a portion of the Trentino region in Italy proves the effectiveness of the proposed method. Numéro de notice : A2017-471 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2017.2656152 En ligne : http://dx.doi.org/10.1109/TGRS.2017.2656152 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=86394 [article]

Sentinel-1 interferometric SAR mapping of precipitable water vapor over a country-spanning area / Pedro Mateus in IEEE Transactions on geoscience and remote sensing, vol 55 n° 5 (May 2017)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 55 n° 5 (May 2017) . - pp 2993 - 2999 Titre : Sentinel-1 interferometric SAR mapping of precipitable water vapor over a country-spanning area Type de document : Article/Communication Auteurs : Pedro Mateus, Auteur ; João Catalão, Auteur Année de publication : 2017 Article en page(s) : pp 2993 - 2999 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image radar et applications [Termes IGN] image Sentinel-SAR [Termes IGN] image TOPSAR [Termes IGN] interféromètrie par radar à antenne synthétique [Termes IGN] Sentinel-1 [Termes IGN] vapeur d'eau Résumé : (Auteur) This paper presents a methodology to generate maps of atmosphere's precipitable water vapor (PWV) over large areas with a length of hundreds of kilometers and a width of about 250 km, based on the use of interferometric Sentinel-1A/B C-band synthetic aperture radar (SAR) data with a high spatial resolution of 5 × 20 m2 and the revisiting time of six days. An algorithm to calibrate and merge PWV maps from different swaths of Sentinel-1 acquired along the same track, using global navigation satellite system (GNSS) measurements, is described. The proposed methodology is tested on Sentinel-1A SAR images acquired over the Iberian Peninsula, along both descending and ascending tracks. The assessment with an independent set of GNSS measurements shows a mean difference of a fraction of millimeter and a dispersion lower than 2 mm. Both the use of Sentinel-1A/B SAR images and the proposed methodology open new perspectives on the application of SAR meteorology for the high-resolution mapping of PWV over large region-spanning areas and the assimilation of interferometric SAR data into numerical weather models. Numéro de notice : A2017-472 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2017.2658342 En ligne : https://doi.org/10.1109/TGRS.2017.2658342 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=86395 [article]