Centre dedocumentation
scientifique

Deep learning in denoising of micro-computed tomography images of rock samples / Mikhail Sidorenko in Computers & geosciences, vol 151 (June 2021)  

Public [article]  inComputers & geosciences > vol 151 (June 2021) . - n° 104716 Titre : Deep learning in denoising of micro-computed tomography images of rock samples Type de document : Article/Communication Auteurs : Mikhail Sidorenko, Auteur ; Denis Orlov, Auteur ; Mohammad Ebadi, Auteur ; et al., Auteur Année de publication : 2021 Article en page(s) : n° 104716 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image [Termes IGN] accentuation d'image [Termes IGN] apprentissage profond [Termes IGN] classification dirigée [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] filtrage du bruit [Termes IGN] filtre passe-bande [Termes IGN] roche [Termes IGN] tomographie Résumé : (auteur) Nowadays, the advantages of Digital Rock Physics (DRP) are well known and widely applied in comprehensive core analysis. It is also known that the quality of the 3D pore scale model drastically influences the results of rock properties simulation, which makes the preprocessing stage of DRP very important. In this work, we consider the application of Deep Convolutional Neural Networks (CNNs) for the preprocessing of CT images, specifically for denoising, in two setups - conventional fully-supervised learning and the self-supervised learning, when the only available data is the noisy images. To train CNNs in a supervised setup, we use images processed by a combination of bilateral and bandpass filters. We trained CNNs of the same architecture with different loss functions to find out how the choice of a loss function influences the model's performance. Some of the obtained CNNs yielded the highest quality in terms of full-reference and no-reference metrics and significant histogram effect (bimodal intensity distribution). Images denoised with these models were qualitatively and quantitatively better than the reference “ground truth” images used for training. We use the Deep Image Prior algorithm to train denoising models in a self-supervised setup. The obtained models are much better than ones obtained in fully-supervised setup, but are too slow, as they are optimization-based rather than feed-forward. Such an algorithm can be used in the dataset generation for feed-forward meta-models. These results could help to develop an AI-based instrument to build high-quality 3D segmented models of rocks for DRP applications. Numéro de notice : A2021-389 Affiliation des auteurs : non IGN Thématique : IMAGERIE/INFORMATIQUE Nature : Article DOI : 10.1016/j.cageo.2021.104716 date de publication en ligne : 02/03/2021 En ligne : https://doi.org/10.1016/j.cageo.2021.104716 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=97672 [article]

An improved computerized ionospheric tomography model fusing 3-D multisource ionospheric data enabled quantifying the evolution of magnetic storm / Jian Kong in IEEE Transactions on geoscience and remote sensing, vol 59 n° 5 (May 2021)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 59 n° 5 (May 2021) . - pp 3725 - 3736 Titre : An improved computerized ionospheric tomography model fusing 3-D multisource ionospheric data enabled quantifying the evolution of magnetic storm Type de document : Article/Communication Auteurs : Jian Kong, Auteur ; Lulu Shan, Auteur ; Chen Zhou, Auteur ; et al., Auteur Année de publication : 2021 Article en page(s) : pp 3725 - 3736 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de géodésie spatiale [Termes IGN] données GNSS [Termes IGN] erreur absolue [Termes IGN] filtre de Kalman [Termes IGN] fusion de données multisource [Termes IGN] modèle ionosphérique [Termes IGN] modèle stochastique [Termes IGN] perturbation ionosphérique [Termes IGN] tempête magnétique [Termes IGN] teneur totale en électrons [Termes IGN] tomographie Résumé : (auteur) Global Navigation Satellite System (GNSS) ionospheric tomography is a typical ill-posed problem. Joint inversion with external observation data is one of the effective ways to mitigate the problem. In this article, by fusing 3-D multisource ionospheric data, and improving the stochastic model, an improved GNSS tomographic algorithm MFCIT [computerized ionospheric tomography (CIT) using mapping function] is presented. The accuracy of the algorithm is validated by selected data under different geomagnetic and solar conditions acquired in Europe. The results show that the estimated, statistically significant uncertainty for each of the layers is about 0.50–3.0TECU, with the largest absolute error within 6.0TECU. The advantage of the MFCIT is that it is based on the Kalman filter, which enables efficient near real-time 3-D monitoring of ionosphere. The temporal resolution can reach ~1 min level. Here, we apply the ionospheric tomography inversion to the magnetic storm on January 7, 2015, in the European region, and quantified the evolution of the storm. The results show that the difference of the core region between the MFCIT and CODE GIM is less than 1TECU. More importantly, during the initial phase of the storm, when the ionospheric disturbance is not evident in the single layer CODE GIM model, the MFCIT shows obvious positive disturbances in the upper ionosphere, although there is no disturbance in the F2 layer. The MFCIT further tracks the evolution of the magnetic storm that the ionospheric disturbance expands from the upper to the lower ionosphere layers, and at UT12:00, the disturbance continues to spread to the F2 layer. Numéro de notice : A2021-396 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2020.3022949 date de publication en ligne : 24/09/2020 En ligne : https://doi.org/10.1109/TGRS.2020.3022949 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=97686 [article]

SAR speckle removal using hybrid frequency modulations / Shuaiqi Liu in IEEE Transactions on geoscience and remote sensing, vol 59 n° 5 (May 2021)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 59 n° 5 (May 2021) . - pp 3956 - 3966 Titre : SAR speckle removal using hybrid frequency modulations Type de document : Article/Communication Auteurs : Shuaiqi Liu, Auteur ; Lele Gao, Auteur ; Yu Lei, Auteur ; et al., Auteur Année de publication : 2021 Article en page(s) : pp 3956 - 3966 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image radar et applications [Termes IGN] artefact [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] filtrage du bruit [Termes IGN] filtre de déchatoiement [Termes IGN] image radar moirée [Termes IGN] modulation de fréquence Résumé : (auteur) Synthetic aperture radar (SAR) images often interfere with speckle artifacts that have a great impact on subsequent processing and analysis operations. To remove speckle artifacts, this article introduces a hybrid denoising approach by using a convolutional neural network (CNN) and consistent cycle spinning (CCS) in the nonsubsample shearlet transform (NSST) domain. First, we apply NSST to a noisy SAR image to gain low- and high-frequency coefficients. Second, we adopt a learned deep CNN model to eliminate the speckle noise in the low-frequency coefficients, which retains more contour information. Third, we employ CCS to enhance the high-frequency coefficients, which preserves more details of the original SAR image. Finally, we obtain the denoised image by using inverse NSST applied to the denoised coefficients. Compared with state-of-the-art algorithms, the results of the experiment indicate that our method not only achieves better speckle removal performance but also maintains more detailed information retention. Numéro de notice : A2021-397 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2020.3014130 date de publication en ligne : 18/08/2020 En ligne : https://doi.org/10.1109/TGRS.2020.3014130 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=97688 [article]

SNR-based water height retrieval in rivers: Application to high amplitude asymmetric tides in the Garonne river / Pierre Zeiger in Remote sensing, vol 13 n° 9 (May 2021)  

Public [article]  inRemote sensing > vol 13 n° 9 (May 2021) . - n° 1856 Titre : SNR-based water height retrieval in rivers: Application to high amplitude asymmetric tides in the Garonne river Type de document : Article/Communication Auteurs : Pierre Zeiger, Auteur ; Frédéric Frappart, Auteur ; José Darrozes, Auteur ; et al., Auteur Année de publication : 2021 Article en page(s) : n° 1856 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de géodésie spatiale [Termes IGN] crue [Termes IGN] filtrage du bruit [Termes IGN] Garonne (bassin) [Termes IGN] interférence [Termes IGN] marée océanique [Termes IGN] modèle d'inversion [Termes IGN] rapport signal sur bruit [Termes IGN] réflectométrie par GNSS [Termes IGN] série temporelle [Termes IGN] surveillance hydrologique Résumé : (auteur) Signal-to-noise ratio (SNR) time series acquired by a geodetic antenna were analyzed to retrieve water heights during asymmetric tides on a narrow river using the Interference Pattern Technique (IPT) from Global Navigation Satellite System Reflectometry (GNSS-R). The dynamic SNR method was selected because the elevation rate of the reflecting surface during rising tides is high in the Garonne River with macro tidal conditions. A new process was developed to filter out the noise introduced by the environmental conditions on the reflected signal due to the narrowness of the river compared to the size of the Fresnel areas, the presence of vegetation on the river banks, and the presence of boats causing multiple reflections. This process involved the removal of multipeaks in the Lomb-Scargle Periodogram (LSP) output and an iterative least square estimation (LSE) of the output heights. Evaluation of the results was performed against pressure-derived water heights. The best results were obtained using all GNSS bands (L1, L2, and L5) simultaneously: R = 0.99, ubRMSD = 0.31 m. We showed that the quality of the retrieved heights was consistent, whatever the vertical velocity of the reflecting surface, and was highly dependent on the number of satellites visible. The sampling period of our solution was 1 min with a 5-min moving window, and no tide models or fit were used in the inversion process. This highlights the potential of the dynamic SNR method to detect and monitor extreme events with GNSS-R, including those affecting inland waters such as flash floods. Numéro de notice : A2021-406 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Article DOI : 10.3390/rs13091856 date de publication en ligne : 10/05/2021 En ligne : https://doi.org/10.3390/rs13091856 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=97722 [article]

A stacked dense denoising–segmentation network for undersampled tomograms and knowledge transfer using synthetic tomograms / Dimitrios Bellos in Machine Vision and Applications, vol 32 n° 3 (May 2021)  

Public [article]  inMachine Vision and Applications > vol 32 n° 3 (May 2021) . - n° 75 Titre : A stacked dense denoising–segmentation network for undersampled tomograms and knowledge transfer using synthetic tomograms Type de document : Article/Communication Auteurs : Dimitrios Bellos, Auteur ; Mark Basham, Auteur ; Tony Pridmore, Auteur ; et al., Auteur Année de publication : 2021 Article en page(s) : n° 75 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image [Termes IGN] acquisition de connaissances [Termes IGN] apprentissage profond [Termes IGN] échantillonnage [Termes IGN] filtrage du bruit [Termes IGN] rapport signal sur bruit [Termes IGN] rayon X [Termes IGN] reconstruction d'image [Termes IGN] segmentation sémantique [Termes IGN] série temporelle [Termes IGN] tomographie Résumé : (auteur) Over recent years, many approaches have been proposed for the denoising or semantic segmentation of X-ray computed tomography (CT) scans. In most cases, high-quality CT reconstructions are used; however, such reconstructions are not always available. When the X-ray exposure time has to be limited, undersampled tomograms (in terms of their component projections) are attained. This low number of projections offers low-quality reconstructions that are difficult to segment. Here, we consider CT time-series (i.e. 4D data), where the limited time for capturing fast-occurring temporal events results in the time-series tomograms being necessarily undersampled. Fortunately, in these collections, it is common practice to obtain representative highly sampled tomograms before or after the time-critical portion of the experiment. In this paper, we propose an end-to-end network that can learn to denoise and segment the time-series’ undersampled CTs, by training with the earlier highly sampled representative CTs. Our single network can offer two desired outputs while only training once, with the denoised output improving the accuracy of the final segmentation. Our method is able to outperform state-of-the-art methods in the task of semantic segmentation and offer comparable results in regard to denoising. Additionally, we propose a knowledge transfer scheme using synthetic tomograms. This not only allows accurate segmentation and denoising using less real-world data, but also increases segmentation accuracy. Finally, we make our datasets, as well as the code, publicly available. Numéro de notice : A2021-456 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article DOI : 10.1007/s00138-021-01196-4 date de publication en ligne : 27/04/2021 En ligne : https://doi.org/10.1007/s00138-021-01196-4 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=97902 [article]

Hyperspectral image denoising via clustering-based latent variable in variational Bayesian framework / Peyman Azimpour in IEEE Transactions on geoscience and remote sensing, vol 59 n° 4 (April 2021)  

Permalink

The Realization and evaluation of PPP ambiguity resolution with INS aiding in marine survey / Zhenqiang Du in Marine geodesy, vol 44 n° 2 (March 2021)  

Permalink

Denoising Sentinel-1 extra-wide mode cross-polarization images over sea ice / Yan Sun in IEEE Transactions on geoscience and remote sensing, Vol 59 n° 3 (March 2021)  

Permalink

Robust unsupervised small area change detection from SAR imagery using deep learning / Xinzheng Zhang in ISPRS Journal of photogrammetry and remote sensing, vol 173 (March 2021)  

Permalink

Space-time disease mapping by combining Bayesian maximum entropy and Kalman filter: the BME-Kalman approach / Bisong Hu in International journal of geographical information science IJGIS, vol 35 n° 3 (March 2021)  

Permalink

A feature-preserving point cloud denoising algorithm for LiDAR-derived DEM construction / Chuanfa Chen in Survey review, Vol 53 n° 377 (February 2021)  

Permalink

A highly adaptable method for GNSS cycle slip detection and repair based on Kalman filter / Xianwen Yu in Survey review, Vol 53 n° 377 (February 2021)  

Permalink

Uncertainties and errors in algorithms for elevation gradients / Dong Shi in International journal of geographical information science IJGIS, vol 35 n° 2 (February 2021)  

Permalink

Application of baseline constraint Kalman filter to BeiDou precise point positioning / Xiaoguo Guan in Survey review, vol 53 n°376 (January 2021)  

Permalink

Du drone LiDAR à un nuage de points précis et exact : une chaîne de traitement LiDAR adaptée et quasi automatique / Maxime Lafleur in XYZ, n° 165 (décembre 2020)

Permalink