Centre dedocumentation
scientifique

Effect of label noise in semantic segmentation of high resolution aerial images and height data / Arabinda Maiti in ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol V-2-2022 (2022 edition)  

Public [article]  inISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences > vol V-2-2022 (2022 edition) . - pp 275 - 282 Titre : Effect of label noise in semantic segmentation of high resolution aerial images and height data Type de document : Article/Communication Auteurs : Arabinda Maiti, Auteur ; Sander J. Oude Elberink, Auteur ; M. George Vosselman, Auteur Année de publication : 2022 Article en page(s) : pp 275 - 282 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] apprentissage profond [Termes IGN] bruit (théorie du signal) [Termes IGN] données altimétriques [Termes IGN] données d'entrainement (apprentissage automatique) [Termes IGN] image à très haute résolution [Termes IGN] image aérienne [Termes IGN] orthoimage [Termes IGN] segmentation sémantique Résumé : (auteur) The performance of deep learning models in semantic segmentation is dependent on the availability of a large amount of labeled data. However, the influence of label noise, in the form of incorrect annotations, on the performance is significant and mostly ignored. This is a big concern in remote sensing applications, wherein acquired datasets are spatially limited, labeling is done by domain experts with possible sources of high inter-and intra-observer variability leading to erroneous predictions. In this paper, we first simulate the label noise while conducting experiments on two different datasets with very high-resolution aerial images, height data, and inaccurate labels, responsible for the training of deep learning models. We then focus on the effect of these noises on the model performance. Different classes respond differently to the label noise. The typical size of an object belonging to a class is a crucial factor regarding the class-specific performance of the model trained with erroneous labels. Errors caused by relative shifts of labels are the most influential label errors. The model is generally more tolerant of the random label noise than other label errors. It has been observed that the accuracy gets reduced by at least 3% while 5% of label pixels are erroneous. In this regard, our study provides a new perspective of evaluating and quantifying the propagation of label noise in the model performance that is indeed important for adopting reliable semantic segmentation practices. Numéro de notice : A2022-434 Affiliation des auteurs : non IGN Thématique : IMAGERIE/INFORMATIQUE Nature : Article DOI : 10.5194/isprs-annals-V-2-2022-275-2022 Date de publication en ligne : 17/05/2022 En ligne : https://doi.org/10.5194/isprs-annals-V-2-2022-275-2022 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100741 [article]

Modeling gravimetric signatures of third-degree ocean tides and their detection in superconducting gravimeter records / Roman Sulzbach in Journal of geodesy, vol 96 n° 5 (May 2022)  

Public [article]  inJournal of geodesy > vol 96 n° 5 (May 2022) . - n° 35 Titre : Modeling gravimetric signatures of third-degree ocean tides and their detection in superconducting gravimeter records Type de document : Article/Communication Auteurs : Roman Sulzbach, Auteur ; Hartmut Wziontek, Auteur ; Michael Hart-Davis, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° 35 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géodésie physique [Termes IGN] données marégraphiques [Termes IGN] gravimètre supraconducteur [Termes IGN] levé gravimétrique [Termes IGN] marée océanique [Termes IGN] montée du niveau de la mer [Termes IGN] simulation hydrodynamique [Termes IGN] surcharge océanique Résumé : (auteur) We employ the barotropic, data-unconstrained ocean tide model TiME to derive an atlas for degree-3 tidal constituents including monthly to terdiurnal tidal species. The model is optimized with respect to the tide gauge data set TICON-td that is extended to include the respective tidal constituents of diurnal and higher frequencies. The tide gauge validation shows a root-mean-square (RMS) deviation of 0.9–1.3 mm for the individual species. We further model the load tide-induced gravimetric signals by two means (1) a global load Love number approach and (2) evaluating Greens-integrals at 16 selected locations of superconducting gravimeters. The RMS deviation between the amplitudes derived using both methods is below 0.5 nGal (1 nGal =0.01nms2) when excluding near-coastal gravimeters. Utilizing ETERNA-x, a recently upgraded and reworked tidal analysis software, we additionally derive degree-3 gravimetric tidal constituents for these stations, based on a hypothesis-free wave grouping approach. We demonstrate that this analysis is feasible, yielding amplitude predictions of only a few 10 nGal, and that it agrees with the modeled constituents on a level of 63–80% of the mean signal amplitude. Larger deviations are only found for lowest amplitude signals, near-coastal stations, or shorter and noisier data sets. Numéro de notice : A2022-299 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1007/s00190-022-01609-w Date de publication en ligne : 30/04/2022 En ligne : https://doi.org/10.1007/s00190-022-01609-w Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100511 [article]

Coastal observation of sea surface tide and wave height using opportunity signal from Beidou GEO satellites: analysis and evaluation / Feng Wang in Journal of geodesy, vol 96 n° 4 (April 2022)  

Public [article]  inJournal of geodesy > vol 96 n° 4 (April 2022) . - n° 17 Titre : Coastal observation of sea surface tide and wave height using opportunity signal from Beidou GEO satellites: analysis and evaluation Type de document : Article/Communication Auteurs : Feng Wang, Auteur ; Dongkai Yang, Auteur ; Guodong Zhang, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° 17 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géodésie spatiale [Termes IGN] Chine [Termes IGN] données altimétriques [Termes IGN] données marégraphiques [Termes IGN] hauteurs de mer [Termes IGN] méthode de Monte-Carlo [Termes IGN] modèle géométrique [Termes IGN] rapport signal sur bruit [Termes IGN] récepteur GNSS [Termes IGN] réflectométrie par GNSS [Termes IGN] signal BeiDou [Termes IGN] surface de la mer [Termes IGN] vague Résumé : (auteur) In this paper, the methods retrieving tide and SWH using reflected BeiDou GEO satellite signals are proposed, and a data-driven method is proposed to calibrate sea state bias of the retrieved tide. In addition, an estimator combining multi-satellite observation based on linear unbiased minimum variance (LUMV) is developed to improve the retrieved precision. The B1I signal experiments in Qingdao and Shenzhen show after calibrating sea state influence using the proposed method, the root-mean-square error (RMSE) could fall to 0.40 m from 0.45 m, and compared to the single-satellite observation, the multi-satellite observation based on the LUMV estimator could significantly reduce the RMSE of the retrieved tide to 0.16 m. Shenzhen experiment is also used to evaluate the performance of retrieving SWH and the determination coefficient of 0.60 is obtained. This paper also conducts Monte Carlo simulation and experiment to evaluate the altimetry and measuring SWH precision using reflected B3I signal. The simulated results when SNR is over 5 dB, incoherent averaging number is 10000, and the receiver bandwidth is over 45 MHz, the estimated precision of the delay can reach up ∼0.15 m, and the precision of the normalized area ranges from 0.2 to 0.3 m. The B3I experiment show that compared to B1I signal, when the reflected signal from individual satellite is used, the better precision with the RMSE of 0.25 can be obtained, and when combining the measurements from the three satellites using LUMV estimator, the RMSE reduces to 0.16 m. Further, the precision of 0.12 m can be obtained by calibrating the sea state influence. Numéro de notice : A2022-213 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1007/s00190-022-01605-0 Date de publication en ligne : 06/03/2022 En ligne : https://doi.org/10.1007/s00190-022-01605-0 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100050 [article]

Effect of PCV and attitude on the precise orbit determination of Jason-3 satellite / Kai Li in Journal of applied geodesy, vol 16 n° 2 (April 2022)  

Public [article]  inJournal of applied geodesy > vol 16 n° 2 (April 2022) . - pp 143 - 150 Titre : Effect of PCV and attitude on the precise orbit determination of Jason-3 satellite Type de document : Article/Communication Auteurs : Kai Li, Auteur ; Xuhua Zhou, Auteur ; Nannan Guo, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : pp 143 - 150 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Techniques orbitales [Termes IGN] capteur d'orientation [Termes IGN] centre de phase [Termes IGN] données Jason [Termes IGN] orbite basse [Termes IGN] orbitographie [Termes IGN] perturbation orbitale Résumé : (auteur) Satellite attitude modes and antenna phase center variations have a great influence on the Precise Orbit Determination (POD) of Low Earth Orbit satellites (LEOs). Inaccurate information about spacecraft attitude, phase center offsets and variations in the POD leads to orbital error. The Jason-3 satellite experienced complex attitude modes which are fixed, sinusoidal, ramp-up/down and yaw-flip. Therefore, it is necessary to properly construct the attitude model in the process of POD especially when there is no external attitude data. For the antenna phase center correction, the PCO which is the deviation between Antenna Reference Point (ARP) and Mean Antenna Phase Center (MAPC) usually can be calibrated on the ground accurately, but the PCV which is the deviation between Instantaneous Antenna Phase Center (IAPC) and Mean Antenna Phase Center (MAPC) will change greatly with the change of space environment. Residual approach can be used to estimate the receiver PCV map. In this paper, we collected the on-board GPS data of Jason-3 satellite from January 2019 and analyzed the impacts of PCV and spacecraft attitude on the orbit accuracy by performing the reduced-dynamic POD. Compared with the reference orbit released by the Centre National d’Études Spatiales (CNES), using the PCV map can reduce the Root Mean Square (RMS) of orbit differences in the Radial (R), Along-track (T), Cross-track (N) and 3D direction about 0.3, 1.0, 0.9, and 1.4 mm. Based on the estimated PCV map, the orbit accuracy in R, T, N and 3D direction is 1.24, 2.81, 1.17, and 3.29 cm respectively by using the measured attitude data. When using the attitude model, the orbit accuracy in R, T, N and 3D directions is 1.60, 3.54, 1.33, and 4.13 cm, respectively. The results showed that the combination of measured attitude data and modeled PCV map can obtain the better orbit solution. It is essential to build a corresponding model in high-precision orbit determination, when there is no attitude data and PCV map. Numéro de notice : A2022-251 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Article DOI : 10.1515/jag-2021-0052 Date de publication en ligne : 26/01/2022 En ligne : https://doi.org/10.1515/jag-2021-0052 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100204 [article]

Recent changes in the climate-growth response of European larch (Larix decidua Mill.) in the Polish Sudetes / Malgorzata Danek in Trees, vol 36 n° 2 (April 2022)  

Public [article]  inTrees > vol 36 n° 2 (April 2022) . - pp 803 - 817 Titre : Recent changes in the climate-growth response of European larch (Larix decidua Mill.) in the Polish Sudetes Type de document : Article/Communication Auteurs : Malgorzata Danek, Auteur ; Tomasz Danek, Auteur Année de publication : 2022 Article en page(s) : pp 803 - 817 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] altitude [Termes IGN] classification par forêts d'arbres décisionnels [Termes IGN] données dendrométriques [Termes IGN] Larix decidua [Termes IGN] modèle de croissance végétale [Termes IGN] Pologne [Termes IGN] stress hydrique [Vedettes matières IGN] Végétation et changement climatique Résumé : (auteur) Larches in the Sudetes are very sensitive to the currently changing climatic factors, and an extreme negative response to drought is observed. In this study, temporal changes in the climate-growth relationship of European larch were analyzed using moving-window correlation. Change-point detection analysis was performed to determine whether there is a temporal connection between tree-ring growth responses and changes in climatic factors trends. The Random Forest predictor importance determination method was used to establish the set of climatic factors that influence larch tree-ring growth the most and to show how this set changes over time. Additionally, cluster analysis was applied to find spatial growth patterns and to generalize the growth response of larch. The results indicate that the main clustering factor is altitude. Nevertheless, an increasing unification of the larch’s response to dominant climatic factors is observable throughout the whole study area. This unification is expressed in the increasingly positive and recently dominant effect of May temperature. A progressively negative influence of the temperature in the summer and late autumn of the year preceding growth was observed, as was an increasing influence of water availability in the summer months. The study indicates that there is a connection between the observed changes and the recent rapid rise in temperature, which has consequently had a negative influence on water availability. The growth of this tree species in the Sudetes is expected to be very limited in the future due to its sensitivity to drought, the predicted increase in temperatures and thermal extremes, and the decrease of the share of summer precipitation in the annual total. Numéro de notice : A2022-316 Affiliation des auteurs : non IGN Thématique : FORET Nature : Article DOI : 10.1007/s00468-021-02251-3 Date de publication en ligne : 09/12/2021 En ligne : http://dx.doi.org/10.1007/s00468-021-02251-3 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100420 [article]

Challenges related to the determination of altitudes of mountain peaks presented on cartographic sources / Katarzyna Chwedczuk in Geodetski vestnik, vol 66 n° 1 (March 2022)  

Permalink

Estimation of the height datum geopotential value of Hong Kong using the combined Global Geopotential Models and GNSS/levelling data / Panpan Zhang in Survey review, vol 54 n° 383 (March 2022)  

Permalink

Validating a new GNSS-based sea level instrument (CalNaGeo) at Senetosa Cape / Pascal Bonnefond in Marine geodesy, vol 45 n° 2 (March 2022)  

Permalink

A comparison of linear-mode and single-photon airborne LiDAR in species-specific forest inventories / Janne Raty in IEEE Transactions on geoscience and remote sensing, vol 60 n° 1 (January 2022)  

Permalink

Empirical comparison between stochastic and deterministic modifiers over the French Auvergne geoid computation test-bed / Ropesh Goyal in Survey review, vol 54 n° 382 (January 2022)  

Permalink

Le radar révèle des montagnes cachées / Laurent Polidori in Géomètre, n° 2198 (janvier 2022)

Permalink

Use of multi-temporal and multi-sensor data for continental water body extraction in the context of the SWOT mission / Nicolas Gasnier (2022)  

Permalink

Assessing the agreement of ICESat-2 terrain and canopy height with airborne lidar over US ecozones / Lonesome Malambo in Remote sensing of environment, vol 266 (December 2021)  

Permalink

Comparative analysis for methods of building digital elevation models from topographic maps using geoinformation technologies / Vadim Belenok in Geodesy and cartography, vol 47 n° 4 (December 2021)  

Permalink

La modélisation des eaux / Michel Kasser in Géomètre, n° 2197 (décembre 2021)

Permalink