Centre dedocumentation
scientifique

Beyond single receptive field: A receptive field fusion-and-stratification network for airborne laser scanning point cloud classification / Yongqiang Mao in ISPRS Journal of photogrammetry and remote sensing, vol 188 (June 2022)  

Public [article]  inISPRS Journal of photogrammetry and remote sensing > vol 188 (June 2022) . - pp 45 - 61 Titre : Beyond single receptive field: A receptive field fusion-and-stratification network for airborne laser scanning point cloud classification Type de document : Article/Communication Auteurs : Yongqiang Mao, Auteur ; Kaiqiang chen, Auteur ; Wenhui Diao, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : pp 45 - 61 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Lasergrammétrie [Termes IGN] apprentissage automatique [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] données laser [Termes IGN] données localisées 3D [Termes IGN] Perceptron multicouche [Termes IGN] représentation parcimonieuse [Termes IGN] réseau neuronal de graphes [Termes IGN] semis de points [Termes IGN] stratification de données [Termes IGN] voxel Résumé : (Auteur) The classification of airborne laser scanning (ALS) point clouds is a critical task of remote sensing and photogrammetry fields. Although recent deep learning-based methods have achieved satisfactory performance, they have ignored the unicity of the receptive field, which makes the ALS point cloud classification remain challenging for the distinguishment of the areas with complex structures and extreme scale variations. In this article, for the objective of configuring multi-receptive field features, we propose a novel receptive field fusion-and-stratification network (RFFS-Net). With a novel dilated graph convolution (DGConv) and its extension annular dilated convolution (ADConv) as basic building blocks, the receptive field fusion process is implemented with the dilated and annular graph fusion (DAGFusion) module, which obtains multi-receptive field feature representation through capturing dilated and annular graphs with various receptive regions. The stratification of the receptive fields with point sets of different resolutions as the calculation bases is performed with Multi-level Decoders nested in RFFS-Net and driven by the multi-level receptive field aggregation loss (MRFALoss) to drive the network to learn in the direction of the supervision labels with different resolutions. With receptive field fusion-and-stratification, RFFS-Net is more adaptable to the classification of regions with complex structures and extreme scale variations in large-scale ALS point clouds. Evaluated on the ISPRS Vaihingen 3D dataset, our RFFS-Net significantly outperforms the baseline (i.e. PointConv) approach by 5.3% on mF1 and 5.4% on mIoU, accomplishing an overall accuracy of 82.1%, an mF1 of 71.6%, and an mIoU of 58.2%. The experiments show that our RFFS-Net achieves a new state-of-the-art classification performance on powerline, car, and fence classes. Furthermore, experiments on the LASDU dataset and the 2019 IEEE-GRSS Data Fusion Contest dataset show that RFFS-Net achieves a new state-of-the-art classification performance. The code is available at github.com/WingkeungM/RFFS-Net. Numéro de notice : A2022-273 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2022.03.019 Date de publication en ligne : 07/04/2022 En ligne : https://doi.org/10.1016/j.isprsjprs.2022.03.019 Format de la ressource électronique : URL Article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100532 [article]

Conventional and neural network-based water vapor density model for GNSS troposphere tomography / Chen Liu in GPS solutions, vol 26 n° 1 (January 2022)  

Public [article]  inGPS solutions > vol 26 n° 1 (January 2022) . - n° 4 Titre : Conventional and neural network-based water vapor density model for GNSS troposphere tomography Type de document : Article/Communication Auteurs : Chen Liu, Auteur ; Yibin Yao, Auteur ; Chaoqian Xu, Auteur Année de publication : 2022 Article en page(s) : n° 4 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de géodésie spatiale [Termes IGN] classification par réseau neuronal [Termes IGN] erreur absolue [Termes IGN] étalonnage de modèle [Termes IGN] modèle météorologique [Termes IGN] propagation troposphérique [Termes IGN] tomographie par GPS [Termes IGN] vapeur d'eau [Termes IGN] voxel Résumé : (auteur) Global navigation satellite system (GNSS) water vapor (WV) tomography is a promising technique to reconstruct the three-dimensional (3D) WV field. However, this technique usually suffers from the ill-posed problem caused by the poor geometry of GNSS rays, resulting in underdetermined tomographic equations. Such equations often rely on iterative methods for solving, but conventional iterative approaches require accurate initial WV density. To address this demand, we proposed two models for WV density estimation. One is the conventional model (CO model) that consists of an exponential model and a linear least-squares model, which are used to describe the spatial and temporal variability of the WV density, respectively. The other is a neural network model (NN model) that uses a backpropagation neural network (BPNN) to fit the nonlinear variation of WV density in both spatial and temporal domains. WV density derived from a Hong Kong (HK) radiosonde station (RS) during 2020 was used to validate the proposed models. Validation results show that both models well describe the spatial and temporal distribution of the WV density. The NN model exhibits better prediction performance than the CO model in terms of root mean square error (RMSE) and bias. We also applied the proposed models to GNSS WV tomography to test their performance in extreme weather conditions. Test results show that the proposed model-based GNSS tomography can correct the content of WV density but cannot accurately sense its irregular distribution. Numéro de notice : A2022-005 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1007/s10291-021-01188-x Date de publication en ligne : 23/10/2021 En ligne : https://doi.org/10.1007/s10291-021-01188-x Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=98920 [article]

Modeling of precipitable water vapor from GPS observations using machine learning and tomography methods / Mir Reza Ghaffari Razin in Advances in space research, vol 69 n° 7 (April 2022)  

Public [article]  inAdvances in space research > vol 69 n° 7 (April 2022) . - pp 2671 - 2681 Titre : Modeling of precipitable water vapor from GPS observations using machine learning and tomography methods Type de document : Article/Communication Auteurs : Mir Reza Ghaffari Razin, Auteur ; Behzad Voosoghi, Auteur Année de publication : 2022 Article en page(s) : pp 2671 - 2681 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de géodésie spatiale [Termes IGN] algorithme génétique [Termes IGN] apprentissage automatique [Termes IGN] classification par séparateurs à vaste marge [Termes IGN] Inférence floue [Termes IGN] Iran [Termes IGN] précipitation [Termes IGN] radiosondage [Termes IGN] réseau neuronal artificiel [Termes IGN] retard hydrostatique [Termes IGN] retard troposphérique zénithal [Termes IGN] tomographie par GPS [Termes IGN] vapeur d'eau [Termes IGN] voxel Résumé : (auteur) This paper studies the application of two machine learning methods to model precipitable water vapor (PWV) using observations of 23 GPS stations from the local GPS network of north-west of Iran in 2011. In a first step, the zenith tropospheric delay (ZTD) and zenith hydrostatic delay (ZHD) is calculated with the Bernese GNSS software and Saastamoinen model as revised by Davis, respectively. Then, by subtracting the ZHD from the ZTD, the zenith wet delay (ZWD) is obtained at each GPS station, for all times. In a second step, ZWD is modeled by two different machine learning methods, based on the latitude, longitude, DOY, time, relative humidity, temperature and pressure. After training a Support Vector Machine (SVM) and an Artificial Neural Network (ANN), ZWD temporal and spatial variations are estimated. Using the formula by Bevis, the ZWD can be converted to PWV at any time and space, for each machine learning method. The accuracy of the two new models is evaluated using control stations, exterior and radiosonde station, whose observations were not used in the training step. Also, all the results of the SVM and ANN are compared with a voxel-based tomography (VBT) model. In the control and exterior stations, ZWD estimated by the SVM (ZWDSVM) and ANN (ZWDANN) is compared with the ZWD obtained from the GPS (ZWDGPS). Also, in the control and exterior stations, precise point positioning (PPP) is used to evaluate the accuracy of the new models. In the radiosonde station, the PWV of the new models (PWVSVM, PWVANN) is compared with the radiosonde PWV (PWVradiosonde) and voxel-based PWV (PWVVBT). The averaged relative error of the SVM, ANN and VBT models in the control stations is 10.50%, 12.71% and 12.91%, respectively. For SVM, ANN and VBT models, the averaged RMSE at the control stations is 1.87 (mm), 2.22 (mm) and 2.29 (mm), respectively. Analysis of the results of PWV estimated by the SVM, ANN and VBT, as well as the surface precipitation obtained from meteorological stations, indicate the high accuracy of the SVM in comparison with the ANN and VBT model. In the results shown in this paper, the SVM has the best ability to accurately estimate ZWD and PWV, using local GPS network observations. Numéro de notice : A2022-219 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Article DOI : 10.1016/j.asr.2022.01.003 Date de publication en ligne : 13/01/2022 En ligne : https://doi.org/10.1016/j.asr.2022.01.003 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100106 [article]

Spatiotemporal analysis of precipitable water vapor using ANFIS and comparison against voxel-based tomography and radiosonde / Mir Reza Ghaffari Razin in GPS solutions, vol 26 n° 1 (January 2022)  

Public [article]  inGPS solutions > vol 26 n° 1 (January 2022) . - n° 1 Titre : Spatiotemporal analysis of precipitable water vapor using ANFIS and comparison against voxel-based tomography and radiosonde Type de document : Article/Communication Auteurs : Mir Reza Ghaffari Razin, Auteur ; Samed Inyurt, Auteur Année de publication : 2022 Article en page(s) : n° 1 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de géodésie spatiale [Termes IGN] Inférence floue [Termes IGN] modélisation spatio-temporelle [Termes IGN] positionnement ponctuel précis [Termes IGN] précipitation [Termes IGN] radiosondage [Termes IGN] retard troposphérique zénithal [Termes IGN] station GPS [Termes IGN] vapeur d'eau [Termes IGN] voxel Résumé : (auteur) Water vapor (WV) is one of the most important parameters in meteorological studies. Using an adaptive neuro-fuzzy inference system (ANFIS), a new method has been proposed for spatiotemporal modeling of precipitable WV (PWV). In a first step, the tropospheric zenith wet delay (ZWD) is calculated using the observations of 23 GPS stations in the northwest of Iran. Out of these 23 stations, 21 stations for training and 2 stations for testing and validating were selected. The observations are for 15 days, ranging from day of year (DOY) 300 to 314 in 2011. The reason for choosing this area and time interval is the availability of a complete set of data. Then, the values of ZWD are converted to PWV. The PWV values obtained from this step are considered as the output of the ANFIS. Also, the latitude and longitude values of the GPS stations, the DOY, observational time (min), temperature (T), pressure (P), and relative humidity (RH) are considered input to ANFIS. The ANFIS network is trained using the back-propagation algorithm. After the training step, the PWV values are evaluated at 2 test stations, KLBR and GGSH, and at Tabriz radiosonde station (38.08° N, 46.28°E). For a more accurate evaluation, all the results of the new method are compared with the voxel-based tomography model. The evaluation of the results is performed using the relative error, standard deviation, correlation coefficient, and root-mean-square error (RMSE). Also, precise point positioning (PPP) is used to better evaluate the proposed model at test stations. The value of the correlation coefficient at the radiosonde station for the ANFIS and voxel is 0.90 and 0.87, respectively. Also, the minimum RMSE calculated for the ANFIS and voxel are 1.02 and 1.06 mm, respectively. In the PPP analysis, an improvement of about 4 mm is observed in the coordinates of the test stations using ANFIS. The results confirm the capability and high accuracy of the proposed model in determining the temporal and spatial variations of PWV. Numéro de notice : A2022-003 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1007/s10291-021-01184-1 Date de publication en ligne : 19/10/2021 En ligne : https://doi.org/10.1007/s10291-021-01184-1 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=98828 [article]

Ionospheric tomographic common clock model of undifferenced uncombined GNSS measurements / German Olivares-Pulido in Journal of geodesy, vol 95 n° 11 (November 2021)  

Public [article]  inJournal of geodesy > vol 95 n° 11 (November 2021) . - n° 122 Titre : Ionospheric tomographic common clock model of undifferenced uncombined GNSS measurements Type de document : Article/Communication Auteurs : German Olivares-Pulido, Auteur ; Manuel Hernández-Pajares, Auteur ; Haixia Lyu, Auteur ; et al., Auteur Année de publication : 2021 Article en page(s) : n° 122 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] correction ionosphérique [Termes IGN] horloge du satellite [Termes IGN] mesurage par GNSS [Termes IGN] modèle ionosphérique [Termes IGN] phase [Termes IGN] positionnement ponctuel précis [Termes IGN] teneur totale en électrons [Termes IGN] tomographie par GPS [Termes IGN] voxel [Termes IGN] Wide Area Augmentation System [Vedettes matières IGN] Traitement de données GNSS Résumé : (auteur) In this manuscript, we introduce the Ionospheric Tomographic Common Clock (ITCC) model of undifferenced uncombined GNSS measurements. It is intended for improving the Wide Area precise positioning in a consistent and simple way in the multi-GNSS context, and without the need of external precise real-time products. This is the case, in particular, of the satellite clocks, which are estimated at the Wide Area GNSS network Central Processing Facility (CPF) referred to the reference receiver one; and the precise realtime ionospheric corrections, simultaneously computed under a voxel-based tomographic model with satellite clocks and other geodetic unknowns, from the uncombined and undifferenced pseudoranges and carrier phase measurements at the CPF from the Wide Area GNSS network area. The model, without fixing the carrier phase ambiguities for the time being (just constraining them by the simultaneous solution of both ionospheric and geometric components of the uncombined GNSS model), has been successfully applied and assessed against previous precise positioning techniques. This has been done by emulating real-time conditions for Wide Area GPS users during 2018 in Poland. Numéro de notice : A2021-776 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1007/s00190-021-01568-8 Date de publication en ligne : 13/10/2021 En ligne : https://doi.org/10.1007/s00190-021-01568-8 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=98839 [article]

Sensitivity of voxel-based estimations of leaf area density with terrestrial LiDAR to vegetation structure and sampling limitations: A simulation experiment / Maxime Soma in Remote sensing of environment, vol 257 (May 2021)  

Permalink

Aleatoric uncertainty estimation for dense stereo matching via CNN-based cost volume analysis / Max Mehltretter in ISPRS Journal of photogrammetry and remote sensing, vol 171 (January 2021)  

Permalink

Benefiting from local rigidity in 3D point cloud processing / Zan Gojcic (2021)  

Permalink

Learning-based representations and methods for 3D shape analysis, manipulation and reconstruction / Marie-Julie Rakotosaona (2021)  

Permalink

Hierarchical instance recognition of individual roadside trees in environmentally complex urban areas from UAV laser scanning point clouds / Yongjun Wang in ISPRS International journal of geo-information, vol 9 n° 10 (October 2020)  

Permalink

An improved constrained simultaneous iterative reconstruction technique for ionospheric tomography / Yi Bin Yao in GPS solutions, Vol 24 n° 3 (July 2020)  

Permalink

Estimation of tropospheric wet refractivity using tomography method and artificial neural networks in Iranian case study / Mir Reza Ghaffari Razin in GPS solutions, Vol 24 n° 3 (July 2020)  

Permalink

Tree annotations in LiDAR data using point densities and convolutional neural networks / Ananya Gupta in IEEE Transactions on geoscience and remote sensing, vol 58 n° 2 (February 2020)  

Permalink

Deep learning for semantic feature extraction in aerial imagery / Ananya Gupta (2020)  

Permalink

Learning 3D generation and matching / Thibault Groueix (2020)  

Permalink