Search

International audience; Raw data issued from meteorological radars are often corrupted by unwanted signals generically called clutter. Hills, tall buildings, atmospheric turbulence, birds, and insects yield patterns that complicate the interpretation of radar images and might add bias in the quantitative precipitation estimates (QPE). Clutter differs from precipitating echoes by both their polarimetric signatures and their particular shapes. This work deals with the removal of clutter. The core idea is to use a fully convolutional network (FCN) to take clutter shapes into account. For a straightforward approach by supervised learning, one would need radar images and their cleaned counterpart, which are not available. We developed a weakly supervised learning method that allows circumventing this issue. This method only requires auxiliary data from rain gauges. The additivity of the reflectivity allows presenting the learning problem in the form of a supervised restoration task with noisy targets. This problem is solved by successive training of a standard FCN (U-net). As ground truth is missing, standard metrics cannot be employed to make a global evaluation. Nevertheless, our method is quantitatively assessed on two clutter classes: ground clutter and interferences. A case study completes the evaluation. A qualitative comparison with the Météo-France algorithm is also performed on a couple of difficult cases.

The contributions of this paper are twofold. First, we show the potential interest of Complex-Valued Neural Network (CVNN) on classification tasks for complex-valued datasets. To highlight this assertion, we investigate an example of complex-valued data in which the real and imaginary parts are statistically dependent through the property of non-circularity. In this context, the performance of fully connected feed-forward CVNNs is compared against a real-valued equivalent model. The results show that CVNN performs better for a wide variety of architectures and data structures. CVNN accuracy presents a statistically higher mean and median and lower variance than Real-Valued Neural Network (RVNN). Furthermore, if no regularization technique is used, CVNN exhibits lower overfitting. The second contribution is the release of a Python library (Barrachina 2019) using Tensorflow as back-end that enables the implementation and training of CVNNs in the hopes of motivating further research on this area., Comment: 6 pages, 5 figures, conference, preprint

International audience; We investigate in this paper the wireless interrogation of chipless and passive mechanical sensors using a range-Doppler imaging algorithm. As an example, a Crookes radiometer illuminated by a light of controlled illuminance is interrogated by using a 24GHz FM-CW radar. A remote detection algorithm is proposed to estimate the reading range and Doppler harmonics embedded in the electromagnetic signal backscattered by the radiometer. From these harmonics the accurate estimation of illuminance is achieved for long reading ranges of at least 3.5m in a cluttered environment.

International audience; The purpose of this work is to perform co-engineering of an object classification system including both a radar sensor and a software of image processing. We aim at the smallest possible false recognition rate, considering three classes of imaged objects. For this we retain five relevant parameters which impact the recognition performances. We adopt the mixed grey wolf optimizer to provide the best set of parameters.

This article describes the wireless and simultaneous interrogation and identification of multiple passive (zero-power) pressure sensors in an industrial environment with a reading range of at least 4 m. The 3-D beamscanning of the scene is performed from a 24-GHz frequency-modulated continuous-wave (FM-CW) radar and for diverse electric field polarizations. The identification is performed using a k-nearest neighbor classification. The benefit of using the cross-polarized electric fields combined with a radar imagery technique is enlightened from the analysis of the background clutter and the simultaneous remote interrogation of several passive pressure sensors at a distance up to 17.7 m. The measurement uncertainty on pressure obtained from the proposed long-range wireless technique is finally reported and discussed.

International audience; In this article, we demonstrate the use of general software-defined radio (SDR) hardware for ground-based interferometric radar (GBIR) system development for static target imaging and displacement estimation purposes. First, a system synchronization approach is proposed within the free and open-source framework GNU Radio, followed by a frequency-domain bandwidth synthesis method used to improve the range resolution with three different waveforms. Second, data preprocessing and target imaging methods are proposed to reduce the negative influences of practical nonideal factors and get high-quality target images. Finally, various experiments are conducted to show the performance of the developed SDR-GBIR system and the proposed methods. It is shown that high-resolution target image and high-accuracy displacement measurement can be obtained by our SDR-GBIR systems.

International audience; This paper is devoted to the study of an automatic target recognition method to classify Inverse Synthetic Aperture Radar (ISAR) images of moving targets. 2-D ISAR imagery generation allows to obtain a pertinent representation of the moving target reflectivity distribution employing radar imaging system. The proposed target recognition technique is based on a combined the polar representation with shape matrix description applied on ISAR image. These descriptors (features) are used to achieve the recognition task method using Neural Network and Support Vector Machine. Several simulations are provided to validate the performances of the proposed method for the automatic aircraft target recognition.

International audience; The estimation of the foliage growth of vineplants based on 3D radar proximal sensing is proposed in this paper. Measurements are performed with a ground-based 24GHz FM-CW radar and for different field polarizations. The statistical estimator derived from the 3D radar images is defined and correlated with the foliage growth of vineplants. This work and results may find applications in precision viticulture for estimating the leaf -to fruit ratio.

International audience; This paper introduces and evaluates numerically a multigrid solver for non-linear tomographic radar imaging. Our goal is to enable the fast and robust inversion of sparse time-domain data with a mathematical full-wave approach utilizing a higher-order Born approximation (BA). Full-wave inversion is computationally expensive, hence techniques to speed up the numerical procedures are needed. To model the wave propagation effectively, we use the finite element time-domain (FETD) method, which is equipped with a multigrid scheme to enable the rapid evaluation of the higher-order BA. As a potential application, we consider the tomography of small solar system bodies (SSSBs) and asteroid interiors particular, the latter of which can contain internal details observable by radar, e.g., layers, voids and cracks. In the numerical experiments, we investigated monostatic, bistatic and multistatic measurement configurations. The results obtained suggest that, with a relevant noise level, the tomographic reconstruction quality can be improved by applying the higher-order BA in comparison to the first-order case, which we interpret as a linearization of the inverse problem. Our open multigrid-FETD solver for Matlab (The Mathworks Inc.) is available online. It applies Matlabs features for graphics computing unit acceleration to enhance computational performance.

1. The recent availability of high spatial and temporal resolution optical and radar satellite imagery has dramatically increased opportunities for mapping land cover at fine scales. Fusion of optical and radar images has been found useful in tropical areas affected by cloud cover because of their complementarity. However, the multitemporal dimension these data now offer is often neglected because these areas are primarily characterised by relatively low levels of seasonality and because the consideration of multitemporal data requires more processing time. Hence, land cover mapping in these regions is often based on imagery acquired for a single date or on an average of multiple dates. 2. The aim of this work is to assess the added value brought by the temporal dimension of optical and radar time series when mapping land cover in tropical environments. Specifically, we compared the accuracies of classifications based on (i) optical time series, (ii) their temporal average, (iii) radar time series, (iv) their temporal average, (v) a combination of optical and radar time series and (vi) a combination of their temporal averages for mapping land cover in Jambi province, Indonesia, using Sentinel‐1 and Sentinel‐2 imagery. 3. Using the full information contained in the time series resulted in significantly higher classification accuracies than using temporal averages (+14.7% for Sentinel‐1, +2.5% for Sentinel‐2 and +2% combining Sentinel‐1 and Sentinel‐2). Overall, combining Sentinel‐2 and Sentinel‐1 time series provided the highest accuracies (Kappa = 88.5%). 4. Our study demonstrates that preserving the temporal information provided by satellite image time series can significantly improve land cover classifications in tropical biodiversity hotspots, improving our capacity to monitor ecosystems of high conservation relevance such as peatlands. The proposed method is reproducible, automated, and based on open‐source tools satellite imagery.

International audience; Deforestation has become a major problem consisting of a continuous regression of forested areas in the world, and for this purpose, an efficient detection of these changes has become more than necessary. In this work, a new method for deforestation change detection is proposed. This approach is based on a supervised fusion of local texture features extracted from SAR images. ALOS PALSAR (Advanced Land Observation Satellite Phased Array type L-band Synthetic Aperture Radar) multi-temporal data have been used in this work. Normalised radar cross-section (NRCS) and polarimetric features extracted from HH and HV polarised data allowed recognising different categories of land covers termed as NRCS classification. Grey-level co-occurrence matrix (GLCM) texture features were extracted by using a different moving window sizes applied on local regions previously obtained by binarisation of the NRCS results. A total of 300 samples of regions and five GLCM characteristics have been used here. The detection of deforestation appears clearly in the resulted images with a very satisfactory precision of the reached regions, and the obtained results of the proposed supervised approach have indeed led to very good detection results of the deforestation change.

International audience; The huge amount of data currently produced by modern Earth Observation (EO) missions has allowed for the design of advanced machine learning techniques able to support complex Land Use/Land Cover (LULC) mapping tasks. The Copernicus programme developed by the European Space Agency provides, with missions such as Sentinel-1 (S1) and Sentinel-2 (S2), radar and optical (multi-spectral) imagery, respectively, at 10 m spatial resolution with revisit time around 5 days. Such high temporal resolution allows to collect Satellite Image Time Series (SITS) that support a plethora of Earth surface monitoring tasks. How to effectively combine the complementary information provided by such sensors remains an open problem in the remote sensing field. In this work, we propose a deep learning architecture to combine information coming from S1 and S2 time series, namely TWINNS (TWIn Neural Networks for Sentinel data), able to discover spatial and temporal dependencies in both types of SITS. The proposed architecture is devised to boost the land cover classification task by leveraging two levels of complementarity, i.e., the interplay between radar and optical SITS as well as the synergy between spatial and temporal dependencies. Experiments carried out on two study sites characterized by different land cover characteristics (i.e., the Koumbia site in Burkina Faso and Reunion Island, a overseas department of France in the Indian Ocean), demonstrate the significance of our proposal.

Passive sensors such as multi-spectral (e.g., Single Lens Reflex, SLR) cameras are increasingly being used for geohazards monitoring (landslides, cliffs affected by rock falls, ice glaciers, and volcano flanks) because of their low cost compared to expensive terrestrial laser scanner (TLS) or radar imaging (GB-InSAR) systems. Indeed, due to the large consumer market, sensor resolution and quality (e.g., gain, dynamic range, and geometry) are increasing rapidly. For gravitational processes, such as landslides, recent research has focused on the development and implementation of image correlation techniques to estimate the spatial shift between at least a pair of images by maximizing a cross-correlation function. A generic and fully automated pipeline is proposed for the processing of long image time series acquired for several site configurations. The system associates modules for 1) the selection of the image sequences, 2) the registration of the image stacks and the correction of the camera movements, and 3) the calculation of the terrain motion using change detection approaches. The system is based on the open-source photogrammetric library MicMac and tailored for the processing of monoscopic images. A sensitivity analysis is conducted to design and test the image processing for two use cases respectively the Chambon landslide (Isère, France) characterized by slow motion (< 10 cm.day−1), and the Pas de l’Ours landslide (Hautes-Alpes, France) characterized by moderate motion (> 50 cm.day−1). Four categories of parameters are tested: the image modality, the image matching parameters, the size of the stable area used in the co-registration stage, and the strategy used to combine the images in the time series. The application of the pipeline on the two use cases provides information about the kinematics and the spatial behavior of the landslides.

International audience; This paper reports a new technique for detecting and remotely analysing zero-power (or fully passive) and wireless sensors. It consists of segmenting with isolines three-dimensional radar images provided by Frequency-Modulated ContinuousWave radars. An isoline refers here to a line along which the radar echo amplitude is the same. Recent applications of this technique on temperature and humidity zero-power sensors are reported.

International audience; Rainfall forecasting is a major issue for anticipating severe meteorological events and for agriculture management. Weather radar imaging has been identified in the literature as the best way to measure rainfall on a large domain, with a fine spatial and temporal resolution. This paper describes two methods allowing to improve rain nowcast from radar images at two different scales. These methods are further compared to an operational chain relying on only one type of radar observation. The comparison is led with regional and local criteria. For both, significant improvements are quantified compared to the original method.

International audience; Characterizing the spatiotemporal evolution of creep is essential to constrain fault slip budget and understand creep mechanism. Studies based on interferometric synthetic aperture radar and Global Positioning System (GPS) satellite observations until 2012 have shown that the central segment of the 17 August 1999 Mw 7.4 Izmit earthquake on the North Anatolian Fault began slipping aseismically following the event. In the present study, we combine new interferometric synthetic aperture radar time series, based on TerraSAR‐X and Sentinel 1A/B radar images acquired over the period 2011–2017, with near‐field GPS measurement campaigns performed every 6 months from 2014 to 2016. The mean velocity fields reveal that creep on the central segment of the 1999 Izmit fault rupture continues to decay, more than 19 years after the earthquake, in overall agreement with models of postseismic afterslip decaying logarithmically with time for a long period of time. Along the fault section that experienced supershear velocity rupture during the Izmit earthquake creep continues with a rate up to ~ 8 mm/year. A significant transient accelerating creep is detected in December 2016 on the Sentinel‐1 time series, near the maximum creep rate location, associated with a total surface slip of 10 mm released in 1 month only. Additional analyses of the vertical velocity fields show a persistent subsidence on the hanging wall block of the Golcuk normal fault that also ruptured during the Izmit earthquake. Our results demonstrate that afterslip processes along the North Anatolian Fault east‐southeast of Istanbul are more complex than previously proposed as they vary spatiotemporally along the fault.

International audience; Extracting information from a polarimetric radar representation usually consists in decomposing it with target decomposition algorithms. This first step can be seen as a geometric analysis of the polarimetric information: the identification of physical radar scattering mechanisms. The problem is that average physical parameters are estimated. As a consequence, these parameters might not describe correctly any of the land cover types that can be mixed together into the radar resolution cell. Therefore, using the polarimetric parameters for land cover classification is challenging. The novelty of the method is to propose a thematic analysis of the polarimetric information preceding the geometric one. The objective is to assess if splitting off polarimetric information on a land cover type basis before applying usual target decomposition algorithms can produce more consistent radar scattering mechanisms when land cover classes are mixed inside the radar resolution cell. A cooperative fusion framework in which very high-resolution optical images are used to unmix physical radar scattering mechanisms is proposed. For bare soil and forests, we point out that a linear unmixing model applied to the covariance matrix is able to split off polarimetric information on a land cover type basis. The assessment of the unmixed radar matrices is carried out with polarimetric radar images from the Radarsat-2 satellite. It was found that despite speckle, the reconstructed radar information after the unmixing process is statistically relevant with the observations. The question whether the unmixed radar images contain relevant thematic information is more challenging, but results tend to validate this property. This method could be used to have a better estimation of vegetation biomass in the context of open forested areas.

International audience; The research and improvement of methods to be used for crop monitoring are currently major challenges, especially for radar images due to their speckle noise nature. The European Space Agency's (ESA) Sentinel-1 constellation provides synthetic aperture radar (SAR) images coverage with a 6 days revisit period at a high spatial resolution of pixel spacing 20 m. Sentinel-1 data are considerable useful, as they provide valuable information of the vegetation cover. The objective of this paper is to provide a better understanding of the capabilities of Sentinel-1 radar images for rice height and dry biomass retrievals. To do this, we train Sentinel-1 data against ground measurements with classical machine learning techniques (Multiple Linear Regression (MLR), Support Vector Regression (SVR) and Random Forest (RF)) to estimate rice height and dry biomass. The study is carried out on a multi-temporal Sentinel-1 dataset acquired from May 2017 to September 2017 over the Camargue region, southern France. The ground in-situ measurements were made in the same period to collect rice height and dry biomass over 11 rice fields. The images were processed in order to produce an intensity radar data stack in C-band including dual-polarization VV (Vertical receive and Vertical transmit) and VH (Vertical receive and Horizontal transmit) data. We found that non-parametric methods (SVR and RF) had a better performance over the parametric MLR method for rice biophysical parameter retrievals. The accuracy of rice height estimation showed that rice height retrieval was strongly correlated to the in-situ rice height from dual-polarization, in which Random Forest yielded the best performance with correlation coefficient R-2 = 0:92 and the root mean square error (RMSE) 16% (7.9 cm). In addition, we demonstrated that the correlation of Sentinel-1 signal to the biomass was also very high in VH polarization with R-2 = 0.9 and RMSE = 18% (162 g.m(-2) (with Random Forest method). Such results indicate that the highly qualified Sentinel-1 radar data could be well exploited for rice biomass and height retrieval and they could be used for operational tasks.

International audience; The aim of this paper is to provide a better understanding of potentialities of the new Sentinel-1 radar images for mapping the different crops in the Camargue region in the South France. The originality relies on deep learning techniques. The analysis is carried out on multitemporal Sentinel-1 data over an area in Camargue, France. 50 Sentinel-1 images processed in order to produce an intensity radar data stack from May 2017 to September 2017. We revealed that even with classical machine learning approaches (K nearest neighbors, random forest, and support vector machine), good performance classification could be achieved with F-measure/Accuracy greater than 86 % and Kappa coefficient better than 0.82. We found that the results of the two deep recurrent neural network (RNN)-based classifiers clearly outperformed the classical approaches. Finally, our analyses of Camargue area results show that the same performance was obtained with two different RNN-based classifiers on the Rice class, which is the most dominant crop of this region, with a F-measure metric of 96 %. These results thus highlight that in the near future, these RNN-based techniques will play an important role in the analysis of remote sensing time series.

International audience; In this paper, we present a C-Band 2 ×2 MIMO Multi-Carrier Radar system that has been realised at the IETR laboratory and a new modulation technique that permits to avoid the use of switches. To validate this system, measurements have been done with two trihedral corner reflectors lying on a grass field, using our system and a VNA. We have processed the acquired data using the Back Projection Algorithm (BPA). VNA and MIMO-MCR give similar results with a vertical resolution of 0.37 m@1.6 m ground range and a vertical ambiguity of 0.74 m@1.6 m ground range. Therefore, the system is fully operational. © 2018 IEEE.

International audience; In this paper we propose to investigate a mm- Wave radar solution that relies on CW measurements on a reduced scan area and to form radar images with backpropagation. This greatly simplifies the radar architecture and improves the SNR. The response of the antenna, critical in monostatic systems, is removed by normalization with respect to the mean value of the measurements.

A $4 \times 1$ passive compressive device (PCD), which works at millimeter wave, is proposed in this study. In a passive manner, the device allows to compress the signals from four separated channels into a single channel. The PCD is a three-dimensional cavity that ensures each channel to be decorrelated with each others. Associated with a Luneburg lens, this allows to generate four distinct beams that are focused in different sectors. This is a multibeam real aperture radar configuration that allows to detect targets in real time. The configuration is validated by measurements.

International audience; Mapping winter vegetation quality is a challenging problem in remote sensing. This is due to cloud coverage in winter periods, leading to a more intensive use of radar rather than optical images. The aim of this letter is to provide a better understanding of the capabilities of Sentinel-1 radar images for winter vegetation quality mapping through the use of deep learning techniques. Analysis is carried out on a multitemporal Sentinel-1 data over an area around Charentes-Maritimes, France. This data set was processed in order to produce an intensity radar data stack from October 2016 to February 2017. Two deep recurrent neural network (RNN)-based classifiers were employed. Our work revealed that the results of the proposed RNN models clearly outperformed classical machine learning approaches (support vector machine and random forest).

The research and improvement of methods to be used for crop monitoring are currently major challenges, especially for radar images due to their speckle noise nature. The European Space Agency’s (ESA) Sentinel-1 constellation provides synthetic aperture radar (SAR) images coverage with a 6-day revisit period at a high spatial resolution of pixel spacing of 20 m. Sentinel-1 data are considerably useful, as they provide valuable information of the vegetation cover. The objective of this work is to study the capabilities of multitemporal radar images for rice height and dry biomass retrievals using Sentinel-1 data. To do this, we train Sentinel-1 data against ground measurements with classical machine learning techniques (Multiple Linear Regression (MLR), Support Vector Regression (SVR) and Random Forest (RF)) to estimate rice height and dry biomass. The study is carried out on a multitemporal Sentinel-1 dataset acquired from May 2017 to September 2017 over the Camargue region, southern France. The ground in-situ measurements were made in the same period to collect rice height and dry biomass over 11 rice fields. The images were processed in order to produce a radar stack in C-band including dual-polarization VV (Vertical receive and Vertical transmit) and VH (Vertical receive and Horizontal transmit) data. We found that non-parametric methods (SVR and RF) had a better performance over the parametric MLR method for rice biophysical parameter retrievals. The accuracy of rice height estimation showed that rice height retrieval was strongly correlated to the in-situ rice height from dual-polarization, in which Random Forest yielded the best performance with correlation coefficient R 2 = 0.92 and the root mean square error (RMSE) 16% (7.9 cm). In addition, we demonstrated that the correlation of Sentinel-1 signal to the biomass was also very high in VH polarization with R 2 = 0.9 and RMSE = 18% (162 g·m − 2 ) (with Random Forest method). Such results indicate that the highly qualified Sentinel-1 radar data could be well exploited for rice biomass and height retrieval and they could be used for operational tasks.

International audience; The development and improvement of methods to map agricultural land cover are currently major challenges, especially for radar images. This is due to the speckle noise nature of radar, leading to a less intensive use of radar rather than optical images. The European Space Agency Sentinel-1 constellation, which recently became operational, is a satellite system providing global coverage of Synthetic Aperture Radar (SAR) with a 6-days revisit period at a high spatial resolution of about 20 m. These data are valuable, as they provide spatial information on agricultural crops. The aim of this paper is to provide a better understanding of the capabilities of Sentinel-1 radar images for agricultural land cover mapping through the use of deep learning techniques. The analysis is carried out on multitemporal Sentinel-1 data over an area in Camargue, France. The data set was processed in order to produce an intensity radar data stack from May 2017 to September 2017. We improved this radar time series dataset by exploiting temporal filtering to reduce noise, while retaining as much as possible the fine structures present in the images. We revealed that even with classical machine learning approaches (K nearest neighbors, random forest, and support vector machines), good performance classification could be achieved with F-measure/Accuracy greater than 86% and Kappa coefficient better than 0.82. We found that the results of the two deep recurrent neural network (RNN)-based classifiers clearly outperformed the classical approaches. Finally, our analyses of the Camargue area results show that the same performance was obtained with two different RNN-based classifiers on the Rice class, which is the most dominant crop of this region, with a F-measure metric of 96%. These results thus highlight that in the near future these RNN-based techniques will play an important role in the analysis of remote sensing time series.This article belongs to the Special Issue High Resolution Image Time Series for Novel Agricultural Applications

International audience; This paper presents a method for the remote interrogation of batteryless humidity sensors based on statistical estimations and three-dimensional beam scanning techniques. As a proof-of-concept demonstration, a Van-Atta reflectarray sensor is interrogated with a 24 GHz Frequency-Modulated Continuous-Wave transceiver at various reading ranges. The extension of the proposed approach to passive or moving sensors is reported and measurement results are discussed. The relative humidity is derived from the statistical analysis of multidimensional radar echoes, while method is proposed to detect the sensor among clutter without knowing its exact position. Moreover it is demonstrated that the cross-polarization effect of the Van-Atta reflectarray combined with the 3D beam scanning technique allows the indoor measurement of the relative humidity up to 58 meters.

International audience; Synthetic aperture radar tomography, or volume imaging with a side-looking radar uses a two-dimensional aperture made of several antenna paths along more or less parallel trajectories. The carrier motion providing the along track extension of the aperture, and the across-track extension is provided by multiple cross-track antenna or by repeated acquisition along parallel lines. The relative position of the individual antenna centres within the aperture is highly demanding (its accuracy requirement is in the order of magnitude of on tenth of wavelength). The absolute positioning is less demanding: it has an impact on the image registration (which does not require accuracy beyond the image resolution, typically several wavelengths) and to a lesser extend to the topographic effects in motion compensation (but DTM resolution is generally even much lower than that of the image). The relative successive positions during the aperture can be refined by autofocus (using the resulting image sharpness to increase the trajectory accuracy) for the along-track separation. For the across-track separation, especially in repeat-pass airborne acquisitions, the same direct approach is not feasible because volume extension of the landscape is low (typically the height of a tree compared to the image stripe length) and the aperture sampling across-track is sparse (each new antenna centre require a new acquisition line that takes tens of minutes of flight). Co-registration by correlating the images obtained from individual acquisition lines is not enough since the autofocus only refines the relative positions within the along-track aperture, leaving error build up at low frequency. Here we propose and evaluate two solutions: The first one is to separately autofocus the acquisition lines and then recover the low frequency positioning errors between the acquisition lines by measuring the distortion field between the images obtained from each acquisition line and deriving the low frequency relative errors. The second is to autofocus the first (or master) acquisition line and then "refocus" the other acquisition lines using the master signal as reference in a way derived from the bistatic autofocus. To compare the methods, we use the 18 calibrations passes of a three week long airborne acquisition campaign with two antenna centre, providing full polarimetric measurement on one side and dual polar measurement on the other side (thus providing 36 separate trajectories for the Vv and Hv channels and 18 for the Hh and Vh channels).

International audience; The goal of most modern automotive safety driver assistance functions is to avoid possible collisions. Pedestrian protection, predictive emergency braking or turn and crossing assist functions are usually based on two steps. First, the radar provides detailed information on the environment, and then a detection procedure is driven. Because of the complicated environment near the vehicle, this second step is a difficult task to achieve in order to give reliable information to the driver. In this paper, we propose to fuse the environment estimation and the detection step into a simple and direct collision target detector. Indeed, this procedure allows detecting possible collision targets, based on their typical Doppler signature, while rejecting all fixed and non-dangerous targets (clutter). Moreover, this detector only exploits a single antenna, and the necessary target speed vector information is obtained by a second order phase expansion using a long integration time, making the use of an antenna array unnecessary.

International audience; For better benefits and yields, a good estimation of grapes quantity in vineyard is necessary. In this paper, three-dimensional (3D) imagery technique using conventional 24GHz Frequency-Modulated Continuous-Wave (FMCW) Radar is applied for detecting and remotely estimating the intra-parcel quantity of grapes. An estimation is possible even in presence of natural or artificial clutters such as leaves, wood or irrigation hoses. The microwave sensing is performed from the radar beam scanning of a vineyard and an estimator is defined to derive the quantity of grapes in grapevines from radar echoes distribution in the interrogated 3D scene. An algorithm based on contour detection is applied to the 3D radar image and a new parameter, called the spread factor, is defined for classifying the echo levels of grapes. The quantity of grapes is finally deduced from an appropriate estimator. This remote sensing approach brings a new and flexible solution for precision viticulture by estimating the grape quantity even for grapes hidden by leaves.

International audience; Cet article examine le problème de la détection et de la localisation d'une cible NLOS dans un milieu urbain en exploitant les multitrajets. Nous proposons un algorithme de détection-localisation basé sur un filtre adapté sous-espace qui fonctionne directement dans l'espace d'état de la cible. Une expérimentation a été réalisée pour montrer qu'un petit radar portatif peut fournir des images des multitrajets. Ces dernières peuvent être clairement interprétées grâce à un modèle simple de tracé de rayon. L'application de l'algorithme de détectionlocalisation sur ces mesures expérimentales montre que la cible peut être détectée et localisée. De fortes ambiguïtés inhérentes au problème de localisation-détection multitrajets sont toutefois visibles et devront être traitées dans les travaux futurs.

International audience; The classification of radar emitters in a naval electronic-warfare (EW) context is a challenging problem. Furthermore, over the last few years, new technologies and the reduction of the fabrication costs have led to an increase in the number of radars. Thus, EW systems need to improve their capabilities of radar identification. The common operational methods, which consist of clustering pulses with similar parameters, are no longer able to solve the classification task efficiently. The main challenges of this data clustering of unknown radar emitters from passive measurements are the high dimension of the received radar pulse samples, the small sample group size, and closely located radar pulse clusters. In this paper, we propose modeling the instantaneous frequency law of the recorded pulses using Bézier curves. The spatial distribution of the control points of the Bézier curves is used as a new feature for the classification process based on the common density-based spatial clustering of applications with noise algorithms. Real data recorded near Brest, France, are used to highlight the interest in using the proposed approach. We also show the potential of the approach for specific emitter identification.

International audience; This paper provides an overview of the SWIM (Surface Waves Investigation and Monitoring) instrument which will be one of the two payload instruments carried by CFOSAT (China France Oceanography SATellite) with a planned launch date in mid 2018. SWIM is a real aperture wave scatterometer operated at near-nadir incidence angles and dedicated to the measurement of directional spectra of ocean waves. The SWIM flight model is currently being assembled and tested, its performance is being assessed and its prototype data processing algorithm is being developed. The aim of this paper is to provide a complete overview on the motivations and scientific requirements of this mission, together with a description of the design and characteristics of the SWIM instrument, and the analysis of its expected performances based on a pre-launch study. An end-to-end simulator has been developed to evaluate the quality of the data products, thus allowing the overall performance of the instrument to be assessed. Simulations run with two subsets of full orbit subsets show that the performances of the instrument and the inversion algorithms will meet the scientific requirements for the mission.

The Shigaraki unmanned aerial vehicle (UAV)-Radar Experiment (ShUREX) is an international (USA-Japan-France) observational campaign, whose overarching goal is to demonstrate the utility of small, lightweight, inexpensive, autonomous UAVs in probing and monitoring the lower troposphere and to promote synergistic use of UAVs and very high frequency (VHF) radars. The 2-week campaign lasting from June 1 to June 14, 2015, was carried out at the Middle and Upper Atmosphere (MU) Observatory in Shigaraki, Japan. During the campaign, the DataHawk UAV, developed at the University of Colorado, Boulder, and equipped with high-frequency response cold wire and pitot tube sensors (as well as an iMET radiosonde), was flown near and over the VHF-band MU radar. Measurements in the atmospheric column in the immediate vicinity of the radar were obtained. Simultaneous and continuous operation of the radar in range imaging mode enabled fine-scale structures in the atmosphere to be visualized by the radar. It also permitted the UAV to be commanded to sample interesting structures, guided in near real time by the radar images. This overview provides a description of the ShUREX campaign and some interesting but preliminary results of the very first simultaneous and intensive probing of turbulent structures by UAVs and the MU radar. The campaign demonstrated the validity and utility of the radar range imaging technique in obtaining very high vertical resolution (~20 m) images of echo power in the atmospheric column, which display evolving fine-scale atmospheric structures in unprecedented detail. The campaign also permitted for the very first time the evaluation of the consistency of turbulent kinetic energy dissipation rates in turbulent structures inferred from the spectral broadening of the backscattered radar signal and direct, in situ measurements by the high-frequency response velocity sensor on the UAV. The data also enabled other turbulence parameters such as the temperature structure function parameter CT2 and refractive index structure function parameter Cn2 to be measured by sensors on the UAV, along with radar-inferred refractive index structure function parameter Cn, radar2. The comprehensive dataset collected during the campaign (from the radar, the UAV, the boundary layer lidar, the ceilometer, and radiosondes) is expected to help obtain a better understanding of turbulent atmospheric structures, as well as arrive at a better interpretation of the radar data., The Correction to this article has been published in Progress in Earth and Planetary Science 2018 5:48

International audience; In recent years, remote sensing of the Earth surface using images acquired from aircraft or satellites has gained a lot of attention. The acquisition technology has been evolving fast and, as a consequence, many different kinds of sensors (e.g., optical, radar, multispectral, and hyperspectral) are now available to capture different features of the observed scene. One of the main objectivesof remote sensing is to monitor changes on the Earth surface. Change detection has been thoroughly studied in the case of images acquired by the same sensors (mainly optical or radar sensors). However, due to the diversity and complementarity of the images, change detection between images acquired with different kinds of sensors (sometimes referred to as heterogeneous sensors) is clearly aninteresting problem. A statistical model and a change detection strategy were recently introduced in[J. Prendes, M. Chabert, F. Pascal, A. Giros, and J.-Y. Tourneret, Proceedings of the IEEE Inter-national Conference on Acoustics, Speech and Signal Processing, Florence, Italy, 2014; IEEE Trans.Image Process., 24 (2015), pp. 799-812] to deal with images captured by heterogeneous sensors. The main idea of the suggested strategy was to model the objects contained in an analysis window by mixtures of distributions. The manifold defined by these mixtures was then learned using trainingdata belonging to unchanged areas. The changes were finally detected by thresholding an appropriate distance to the estimated manifold. This paper goes a step further by introducing a Bayesian nonparametric framework allowing us to deal with an unknown number of objects in analysis windows without specifying an upper bound for this number. A Markov random field is also introduced to account for the spatial correlation between neighboring pixels. The proposed change detector is validated using different sets of synthetic and real images (including pairs of optical images and pairs of optical and radar images) showing a significant improvement when compared to existingalgorithms.

International audience; Retrieval of sea surface wind speed from Synthetic Aperture Radar (SAR) data is one of the most widely used methods, since it can give higher resolutions than the other available surface wind sources. For this approach, two principal methods can be found: one is based on electromagnetic (EM) models and the other is based on empirical (EP) ones. In both indicated ways, the Geophysical Model Functions (GMFs) are used to describe the dependency of radar scattering on wind speed and the geometry of observations. By knowing radar scattering and geometric parameters from SAR data, it is possible to invert the GMFs to retrieve sea wind speed. Wind speed estimated by two studied models is compared together and evaluated with measured data. Based on the comparisons, the advantages and limits of the studied models are analyzed and discussed.

The aim of this paper is to estimate geometric, water and physical surface soil parameters from typical semi-arid regions made over bare study area (North Africa) using multi-temporal X-band SAR images (TerraSAR-X). For spatial and temporal surface roughness estimation, empirical relationships between radar and soil roughness parameters (rms height “Hrms”, and Zg parameter) were proposed. Two roughness classes are identified through radar signal inversion (smooth and ploughed soils). For the retrieval of surface soil moisture at a high spatial resolution, an algorithm combing TerraSAR-X images with continuous thetaprobe measurements was proposed. Two assumptions were studied: (1) roughness variations during the radar acquisition campaigns were not accounted for; (2) a simple correction for temporal variations in roughness was included. Finally, an empirical relationship was established between the mean moisture values retrieved from the SAR images and the percentage of clay over several test fields. Results showed that highly accurate clay estimations can be achieved.

International audience; High Frequency radars usually use reflected waves from ionosphere to perform continuous surveillance of very large and far off areas. In the case of long term integration, the ionosphere instability affects the radar image and thus induces poor detection and high probability of false alarm. This problem appears in surveillance issues dealing with slow targets such as vessels. Such an issue is also present in surface wave radar due to the lack of directivity of the transmitting antennas. In this paper, we introduce probabilistic models of the ionosphere effects assuming that only the phase path fluctuation has a meaning in the radar point of view. A first model has been built and compared to a real case. It appears that the deviation in range and Doppler are meaningful. That model will be integrated in the radar processing in order to synthetize highly realistic radar data and estimate the radar parameters in case of turbulent ionosphere (i.e. range-Doppler shifts and uncertainties as well as detection and false alarm probabilities). The obtained results will be presented during the EuCAP 2016 conference.

International audience; This paper deals with the problem of HF surface wave radar. The goal is to integrate in a unique tool the antenna radiation and the propagation calculations in order to make the analysis consistent.

International audience; Among the different available wind sources, i.e. in situ measurements, numeric weather models, the retrieval of wind speed from Synthetic Aperture Radar (SAR) data is one of the most widely used methods, since it can give high wind resolution cells. For this purpose, one can find two principal approaches: via electromagnetic (EM) models and empirical (EP) models. In both approaches, the Geophysical Model Functions (GMFs) are used to describe the relation of radar scattering, wind speed, and the geometry of observations. By knowing radar scattering and geometric parameters, it is possible to invert the GMFs to retrieve wind speed. It is very interesting to compare wind speed estimated by the EM models, general descriptions of radar scattering from sea surface, to the one estimated by the EP models, specific descriptions for the inverse problem. Based on the comparisons, some ideas are proposed to improve the performance of the EM models for wind speed retrieval.

International audience; This paper investigates a new polarimetric device that produces reduced scale measurements for bistatic radar. The optical scale device is proposed to overcome the lack of real bistatic and full polarimetric radar acquisitions for forest of trunks. It is based on the scale invariant rule that is deduced from the Maxwell equations. From this principle, the electromagnetic (EM) response is kept at the reduced scale if the illumination wavelength is scaled and if the permittivity remains the same. This last condition needs the use of a carefully chosen material at optical scale. Nanoscale scenes are composed of carbon nanotubes (CNTs). They present several advantages: their arrangement is similar to the structure of a forest of trunks, their shape, and density can be controlled, it is possible to create very large scenes composed of thousands of elements, and finally their permittivity at optical wavelength is about the same order of magnitude than the permittivity of the tree trunks at radar wavelength. They are measured under the view of a microscope objective with a 633-nm wavelength laser playing the role of the EM source. The device images at once the forest polarimetric response, for a given transmitter location, and for the entire set of angular positions of reception. Two samples of forests with different densities are measured. They are analyzed using the Cloude and Pottier decomposition and the Lu and Chipman decomposition. The produced results are unprecedented on forest-like environment and it would be very helpful for the radar community.

The TanDEM-X mission allows generation of digital elevation models (DEMs) with high potential for glacier monitoring, but the radar penetration into snow and ice remains a main source of uncertainty. In this study, we generate five new DEMs of the Mont-Blanc area from TanDEM-X interferometric pairs acquired in 2012/2013. We conducted a multitemporal analysis of the DEMs in comparison with two high-resolution DEMs obtained from Pleiades stereo satellite images in 2012 and 2013. A vertical precision of 1–3 m of the radar DEMs is estimated over ice and snow free areas and slopes less than 40°. DEM-derived elevation changes are compared with outputs of the snowpack model Crocus and snow accumulation measurements. The results show that at altitudes below $\sim$ 2500-m a.s.l., the radar penetration is negligible in our study area. The DEM-derived elevation changes agree, within uncertainty, with the modeled and field snow height. At higher altitudes, the comparison between the radar and optical DEMs acquired only a few weeks apart allows estimating the interferometric bias of the X-band DEM in the dry snowpack. At 4000-m a.s.l, it reaches 4 m on average in October and February. A geodetic glacier mass balance calculated using the October radar DEM would be biased. For the least favorable case, the highly elevated Bossons glacier, the bias would correspond to 1.66-m w.e. This error is too large to derive significant annual mass balances, but similar to elevation or seasonality uncertainties if integrated over a 10-years period.

International audience; This paper aims to evaluate the contribution of multitemporal polarimetric synthetic aperture radar (SAR) data for winter wheat and rapeseed crops parameters [height, leaf area index, and dry biomass (DB)] estimation, during their whole vegetation cycles in comparison to backscattering coefficients and optical data. Angular sensitivities and dynamics of polarimetric indicators were also analyzed following the growth stages of these two common crop types using, in total, 14 radar images (Radarsat-2), 16 optical images (Formosat-2, Spot-4/5), and numerous ground data. The results of this study show the importance of correcting the angular effect on SAR signals especially for copolarized signals and polarimetric indicators associated to single-bounce scattering mechanisms. The analysis of the temporal dynamic of polarimetric indicators has shown their high potential to detect crop growth changes. Moreover, this study shows the high interest of using SAR parameters (backscattering coefficients and polarimetric indicators) for crop parameters estimation during the whole vegetation cycle instead of optical vegetation index. They particularly revealed their high potential for rapeseed height and DB monitoring [i.e., Shannon entropy polarimetry (r2=0.70) and radar vegetation index (r2=0.80), respectively]

International audience; The retrieval of buried objects is addressed through an innovative frequency-hopping imaging technique. The a-priori knowledge on the sparseness of the retrieved profiles (in a suitable representation basis) is exploited to process the spectral content of wideband signals collected in ground-penetrating radar (GPR) scenario. Towards this end, a frequency hopping procedure based on a customized version of the multi-task (MT) Bayesian Compressive Sensing is developed. Selected numerical results are reported to preliminarily assess the effectiveness of the presented strategy.

International audience; A wideband antenna array is developed for the land-mine detection in the desert with an on-board radar. The final array is very large and contains 128 Vivaldi antennas. The array is split into 16 sub-arrays of 8 elements. Each Vivaldi antenna is driven independently by electronic switches for SAR imaging purpose. In a first step, a sub-array of 8 elements has been studied, fabricated and measured in the anechoic chamber. In a second step, influences of the overall mechanical support and the protection radome are studied numerically and experimentally. This paper focuses on the second aspect with the minimization of both support and radome effects on the radiation performances of the Vivaldi elementary antenna.

International audience; the paper presents a novel “Through-The-Wall” radar devoted to the detection of people in motion behind a wall. It uses a new kind of 3D imaging associated with a new antenna as well as sophisticated information processing.

The Ku-band (13.8 GHz - 2.2 cm) RADAR instrument onboard the Cassini-Huygens spacecraft has revealed the richness of the surface of Titan, as numerous seas, lakes, rivers, cryo-volcanic flows and vast dune fields have been discovered. Linear dunes are a major geomorphological feature present on Titan, covering up to 17% of its surface, mainly in equatorial regions. However, the resolution of the RADAR instrument is not good enough to allow a detailed study of the morphology of these features. In addition, other linear wind-related landforms, such as mega-yardangs (linear wind-abraded ridges formed in cohesive rocks), are likely to present a comparable radar signature that could be confused with the one of dunes. We conducted a comparative study of the radar radiometry of both linear dunes and mega-yardangs, based on representative terrestrial analogues: the linear dunes located in the Great Sand Sea in western Egypt and in the Namib Desert in Namibia, and the mega-yardangs observed in the Lut Desert in eastern Iran and in the Borkou Desert in northern Chad. We analysed the radar scattering of both terrestrial linear dunes and mega-yardangs, using high-resolution radar images acquired by the X-band (9.6 GHz - 3.1 cm) sensor of the TerraSAR-X satellite. Variations seen in the radar response of dunes are the result of a contrast between the dune and interdune scattering, while for mega-yardangs these variations are the result of a contrast between ridges and erosion valleys. We tested a simple surface scattering model, with parameters derived from the local topography and surface roughness estimates, to accurately reproduce the radar signal variations for both landforms. It appears that we can discriminate between two types of dunes - bare interdunes as in Egypt and sand-covered interdunes as in Namibia, and between two types of mega-yardangs - young yardangs..., Comment: In press in Icarus (August 2015)

International audience; Modeling sea clutter is a difficult problem due to the interaction of the sea surface characteristics (wind speed and wind direction), the geometry of acquisition (grazing angle and azimuth angle), and radar parameters (frequency, polarization, and resolution). In synthetic aperture radar (SAR) imagery, the effect of coherent averaging and motion from the sea will also influence the statistics. An accurate description of the sea surface amplitude probability density function is important for robust target detection. The goal of this paper is to identify a common distribution which matches two different airborne SAR X-band data sets obtained from different locales and at different grazing angles. The first data set was collected by the French Aerospace Laboratory (ONERA) SETHI radar off the coast of France, at low grazing angles (3° and 10°). The second is the Ingara medium grazing angle (15°–45°) sea clutter data set collected by the Defence Science and Technology Organisation off the coast of Australia. In this paper, a number of recently developed distributions are considered, including the K+Rayleigh and Pareto+noise, both of which account for thermal noise. To measure the effectiveness of the model fit, the Bhattacharyya distance and the threshold error are computed with particular attention given to the tail region, where the threshold is determined in a detection scenario.

International audience; This paper proposes a vibration-induced micro-Doppler estimation method for oscillating targets in synthetic aperture radar (SAR) images using azimuth time-frequency tracking and a matched signal transform. The approach involves an azimuth defocusing of the SAR image in order to access the phase history. The tracking in azimuth is based on local polynomial phase modeling using as estimator for the polynomial coefficients the high-order ambiguity function. The vibration frequency is obtained from the spectrum of the tracked instantaneous frequency law, while the oscillation amplitude is estimated using a matched signal transform. The procedure is tested by simulations and on real SAR images acquired by the TerraSAR-X satellite over the Puylaurent water-dam in France.

International audience; The paper describes an operational method of rainfall now-casting based on ground radar acquisitions with high space and time resolution. The nowcast horizon is between 30 minutes and 1 hour as required by prevention measures of flash floods. The characteristics of the input data justify the design of an image-based method that estimates wind fields from image acquisitions and forecasts the location and quantity of rain in the near future. The estimation phase relies on an iterative data assimilation of the radar acquisitions with an evolution model of motion and image fields, while the forecast is obtained by simulating these fields at the chosen horizon. The research is done in the context of a collaboration with the french company Numtech and the data are obtained with radars of the company Weather Measures.

Land-cover geodatabases are key products for the understanding of environmental systems and for setting up national and international prevention and protection policies. However, their automatic generation and update remain complicated with high accuracy over large scales. In natural environments, most of the existing solutions are semi-automatic in order to achieve a suitable discrimation of the large number of forest and crop classes. A large amount of remote sensing possibilities is at the moment available and data fusion appears to be the most suitable solution for that purpose. The paper tackles the issue of land-cover mapping in such areas assuming the existence of a partly non-updated 5-class geodatabase: buildings, roads, water, crops, forests. Lidar point clouds and Radar images at two spatial resolutions and bands are merged at the feature level and fed into an efficient supervised classification framework. Results show that some classes benefit from the joint exploitation of multiple observations in terms of accuracy or recall.