Search

In order to satisfy vital train-to-wayside transmissions for metro applications, IEEE802.11× modems are deployed along the tracks on an Ethernet network in tunnel. This paper presents possible enhancements of the PHY layer of these systems based on MIMO techniques with and without channel state information at transmitter in order to increase system performance without increasing the number of deployed access points and the transmitted power. The tunnel scenario with a masking train in the case of low and high spatial correlation is modeled thanks to the Kronecker model obtained with a 3D ray tracing tool. We consider the following MIMO schemes: spatial multiplexing, space-time block codes, max-dmin and P-OSM precoders at 5.8 GHz with a targeted Frame Error Rate equals to 3.10−2 for a frame length of 100 bytes. The MIMO algorithms are compared in terms of achieved signal to noise ratio (SNR) versus spectral efficiency.

National audience; Ce papier illustre le gain de performance potentiel apporté par l'utilisation des techniques MIMO dans le cadre des communications radio en tunnel ferroviaire. Différents schémas de transmission sont testés et évalués dans le cas de canaux faiblement puis fortement corrélés.

International audience; BACKGROUND: Epidemiologic studies indicate that chronic cough and sputum production are associated with increased mortality and disease progression in COPD subjects. Our objective was to identify features associated with chronic cough and sputum production in COPD subjects. METHODS: Cross-sectional analysis of data were obtained in a multicenter (17 university hospitals in France) cohort of COPD patients. The cohort comprised 433 COPD subjects (65 +/- 11 years; FEV(1), 50 +/- 20% predicted). Subjects with (n = 321) and without (n = 112) chronic cough and sputum production were compared. RESULTS: No significant difference was observed between groups for age, FEV(1), body mass index, and comorbidities. Subjects with chronic cough and sputum production had increased total mean numbers of exacerbations per patient per year (2.20 +/- 2.20 vs 0.97 +/- 1.19, respectively; p < 0.0001), moderate exacerbations (1.80 +/- 2.07 vs 0.66 +/- 0.85, respectively; p < 0.0001), and severe exacerbations requiring hospitalizations (0.43 +/- 0.95 vs 0.22 +/- 0.56, respectively; p < 0.02). The total number of exacerbations per patient per year was the only variable independently associated with chronic cough and sputum production. Frequent exacerbations (two or more per patient per year) occurred in 55% vs 22% of subjects, respectively, with and without chronic cough and sputum production (p < 0.0001). Chronic cough and sputum production and decreased FEV(1) were independently associated with an increased risk of frequent exacerbations and frequent hospitalizations. CONCLUSIONS: Chronic cough and sputum production are associated with frequent COPD exacerbations, including severe exacerbations requiring hospitalizations.

International audience

International audience; In image processing, contours can be extracted by linear convolution of the image with first-order derivative filters. The oriented half Gaussian filters are directed in all desired directions to analyze the images. Thus, they are useful for detecting contours or extracting precise orientations. The process of filter orientation requires the application of a rotation technique in a two-dimensional discrete domain of different scientific strategies: interpolation (which can modify image information) or discrete geometry (without modification of the image values). In this paper, different methods of discrete rotating the half Gaussian filter are compared and evaluated to detect contours in synthetic and real images. Results are qualitatively and quantitatively compared, validating which rotation technique is most beneficial for edge detection.

International audience; This paper presents both a recursive scheme to perform Geometric Algebra operations over a prefix tree, and Garamon, a C++ library generator implementing these recursive operations. While for low dimension vector spaces, precomputing all the Geometric Algebra products is an efficient strategy, it fails for higher dimensions where the operation should be computed at run time. This paper describes how a prefix tree can be a support for a recursive formulation of Geometric Algebra operations. This recursive approach presents a much better complexity than the usual run time methods. This paper also details how a prefix tree can represent efficiently the dual of a multivector. These results constitute the foundations for Garamon, a C++ library generator synthesizing efficient C++ / Python libraries implementing Geometric Algebra in both low and higher dimensions, with any arbitrary metric. Garamon takes advantage of the prefix tree formulation to implement Geometric Algebra operations on high dimensions hardly accessible with state-of-the-art software implementations. Garamon is designed to produce easy to install , easy to use, effective and numerically stable libraries. The design of the libraries is based on a data structure using precomputed functions for low dimensions and a smooth transition to the new recursive products for higher dimensions. Mathematics Subject Classification (2010). Primary 99Z99; Secondary 00A00.

International audience; In this paper we are proposing a new way to compute a discrete bisector function, which is an important tool for analyzing and filtering Euclidean skeletons. From a continuous point of view, a point that belongs to the medial axis is the center of a maximal ball that hits the background in more than one point. The maximal angle between those points is expected to be high for most of the object points and corresponds to the bisector angle. This logic is not really applicable in the discrete space since in some configurations we miss some background points leading sometimes to small bisector angles. In this work we use annuli to find the background points in order to compute the bisector angle. The main advantage of this approach is the possibility to change the thickness and therefore to be more flexible while computing the bisector angle.

International audience; Les Algèbres Géométriques constituent un outil intuitif et performant pour représenter et manipuler des objets géométriques ainsi que pour calculer les intersections entre plusieurs objets géométriques. Parmi les Algèbres Géométriques les plus courantes figure l'Algèbre Géométrique Conforme qui permet de représenter des objets ronds (cercles et sphères) et plats (droites et plans) de façon compacte et élégante. Dans l'Algèbre Géométrique Conforme, les objets plats se construisent comme un objet rond dont on a remplacé un point de construction par un point à l'infini. Cela a pour effet de voir les objets plats comme des objets ronds avec une courbure nulle. La structure de ces objets plats n'est pas différentes de celles des objets ronds (une droite et un cercle de courbure nulle). Tous ces objets se construisent avec le nombre minimum de points de contrôle nécessaires à leur définition, ou bien par leur formule implicite communément utilisée en algèbre linéaire. Mais, contrairement à l'algèbre linéaire, ces objets se manipulent (transformations rigides) tous avec exactement les mêmes opérateurs, indifféremment de leur type. Enfin, l'intersection entre ces objets se calcule elle aussi avec un unique opérateur. En revanche, la diversité des objets exprimés dans cette algèbre se limite à quelques objets (sphères, plans, cercles, droites, ...). Dans cet article, nous nous intéressons à une extension de l'Algèbre Géométrique Conforme qui permet de représenter et manipuler les surfaces quadriques, l'Algèbre Géométrique Conforme Quadrique. Les surfaces quadratiques suscitent de plus en plus d'attention dans la communauté des Algèbres Géométriques et plusieurs algèbres ont été proposées pour représenter et manipuler ces surfaces. En nous plaçant dans l'Algèbre Géométrique Conforme Quadrique, nous étudions les objets ronds et plats de cette algèbre. L'objectif est, dans un premier temps, de retrouver les objets ronds et plats de l'Algèbre Géométrique Conforme et ensuite d'explorer quels sont les nouveaux objets ronds et plats que l'on peut définir et manipuler avec l'Algèbre Géométrique Conforme Quadrique.

International audience; With the widespread use of image processing technologies, objective image quality metrics are a fundamental and challenging problem. In this paper, we present a new No-Reference Image Quality Assessment (NR-IQA) algorithm based on visual attention modeling and a multivariate Gaussian distribution to predict the final quality score from the extracted features. Computational modeling of visual attention is performed to compute saliency maps at three resolution levels. At each level, distortions of the input image are extracted and weighted by the saliency maps in order to highlight degradations of visually attracting regions. The generated features are used by a probabilistic model to predict the final quality score. Experimental results demonstrate the effectiveness of the metric and show better performance when compared to well known NR-IQA algorithms.

The paper deals with conics in a rational Bézier representation based on mass points where the weights are complex numbers here. A special representation of conics using weighted points and vectors offers a calculus flexibility in the handle elementary geometrical transformations as rotations, homotheties and direct similarity transformations. Some examples are proposed to the reader., L'article traite du mode de représentation des coniques écrites sous forme Bézier rationnelle à l'aide de points massiques où les masses sont des nombres complexes. Une représentation spéciale des coniques reposant sur des points pondérés et des vecteurs purs offre une souplesse de calcul dans les transformations géométriques élémentaires comme les rotations, homothéties et similitudes directes. Des exemples sont proposés au lecteur prouvant cet usage.

Modelling polynomial curves or arcs with Bezier curves can be seen as a basis conversion not so easy for the rational curves. The classical representation of Rational curves based on controlled points with non negative weights as in NURBS does not cover all rational curves. This can be fixed by using the rational Bezier representation by mass points that are weighted points with negative or null weights. The curve of any rational function includes arcs denoted as connex components. These curves and their asymptotic lines are here modelled by the use of mass control points. The asymptotic lines are described by a point that are one weighted point or a vector. An algorithm proposes to represent any arc of a rational function f = P Q where P, Q are polynomials. The algorihtm starts by the rational Bezier representation by mass points with the parameter running on [0, 1], isolates any connex component of the curve by obtaining the zeroes of the denominator. The real line isthus split in subintervals. The next step consists in a homographic parameter change that transforms the [0, 1] interval into any previous interval. The parameter change does not modify the degree curve and points out the mass control points Bézier representation including the asymptotic line., La modélisation de courbes ou d'arcs de courbe par des courbes de Bézier, qui peut être vu comme un simple changement de base pour les courbes polynomiales, est moins évidente pour les courbes rationnelles. En effet, les courbes rationnelles ne se sont pas complètement capturées par le modèle classique des courbes de Bézier rationnelles à points pondérés de contrôle (i.e. les poids sont strictement positifs comme dans les NURBS), il est nécessaire d'introduire les courbes de Bézier rationnelles à points massiques de contrôle (i.e. les poids peuvent être nuls ou négatifs). La courbe représentative C f d'une fraction rationnelle f est généralement constituée de plusieurs branches (composantes connexes de la courbe). Dans cet article, nous montrons comment chaque branche est modélisée par une courbe de Bézier rationnelle à points massiques de contrôle. Pour ces courbes, la difficulté est de pouvoir modéliser les asymptotes. Cette difficulté est résolue en utilisant un vecteur (i.e. un point massique de poids nul) qui donne la direction de cette asymptote et un point pondéré (i.e. un point massique de poids non nul) qui fixe la position de cette asymptote. Pour déterminer les courbes de Bézier rationnelles (à points massiques de contrôle) modélisant chaque branche de la courbe représentative C f d'une fraction rationnelle f = P Q où P et Q sont des polynômes dépendant de la variable x, la méthode proposée commence par déterminer la courbe de Bézier ration-nelle représentant C f sur [0; 1], puis détermine, pour chaque composante connexe, la courbe de Bézier rationnelle en découpant la droite réelle R en intervalles où les extrémités sont les racines de Q(x) auxquelles sont ajoutées −∞ et +∞. Les points massiques de contrôle sont alors déterminés par un changement de paramètre homographique qui envoie l'intervalle [0; 1] sur l'un des intervalles précités.

International audience; – L'échographie est la modalité d'imagerie la plus répandue car la moins nocive. Cependant, la présence de bruit de « speckle » dans l'imagerie médicale ultrasonographique reste un inconvénient majeur. La principale limite des méthodes de réduction de bruits proposées est qu'une réduction efficace du bruit s'effectue au détriment de la qualité (lissage excessif, effet de flou, un aspect artificiel, ...). Dans cet article, nous proposons un nouveau filtre Multi-Sorties basé sur la Décomposition Multiplicative Multirésolutions (MOF-MMD). Cette méthode multi-échelles, particulièrement efficace dans le cas de bruit multiplicatif, améliore distinctement trois sorties: bords, texture, et image globale. Le filtre multi-sorties vise à offrir aux praticiens des images améliorées en fonction des besoins de diagnostic. Les différentes structures, textures et bordures sont filtrées selon l'image contour obtenue par des opérateurs de morphologie mathématique. Les images résultantes du filtrage sont évaluées et validées par comparaison avec deux méthodes de l'état de l'art. Abstract – Ultrasonographic examination, either as visual inspection or quantitative analysis, is less effective than other medical imaging systems due to speckle noise. In this paper, a new Multi-Output Filter based on a Multiplicative Multiresolution Decomposition (MOF-MMD) is proposed. This multiscale based method, particularly efficient in the case of multiplicative noise, enhances distinctively three outputs: edges, texture and the global image. Ultrasound is the most widespread imaging modality. However, it suffers from a major disadvantage as it is corrupted by speckle noise. The state-of-the-art speckle reduction methods often offers an effective speckle reduction at the expense of the quality (oversmoothig, blurring effect, artificial appearance...). In this paper, a new Multi-Output Filter based on a Multiplicative Multiresolution Decomposition (MOF-MMD) is proposed. This multiscale based method, particularly efficient in the case of multiplicative noise, enhances distinctively three outputs: edges, texture and the global image. The multi-output filter aims at offering an enhanced images according to the features desired by radiologists. The different structures, textures and edges are filtered according to the contour image obtained by morphological operators.The resulting images are evaluated and validated by comparison with two state-of-the-art methods.

2D Visual saliency has been widely explored for decades. Several comprehensive and well performing models have been proposed, but they are not totally adapted to stereoscopic 3D content. To date only few tentatives of 3D saliency prediction can be found in the literature and most of them rely on binocular depth/disparity. The latter information cannot be correctly obtained in the case of asymmetric processing of the stereo-pair, exploiting the phenomenon of binocular suppression. Based on this aspect, we propose in this paper a new saliency model for stereoscopic 3D images. The proposed model considers two features: (1) spatial feature based on the characteristics of interest points and (2) depth feature based on monocular cues. The latter feature is adapted to asymmetric content and uses occlusions for predicting depth order of the image objects. A tunable fusion strategy is proposed in order to take advantage of different modalities of combining conspicuity maps. For the needs of performance evaluation, an eye-tracking database is created using stereo-pairs with different content. The proposed model gives very good performance in comparison to the literature. The results show that the use of monocular cues outperforms the use of disparity. HighlightsA saliency model for stereoscopic 3D images (asymmetric or symmetric) is proposed.Interest point is exploited for the construction of the spatial conspicuity map.Monocular cues (occlusion) are used for the construction of depth conspicuity map.Different fusion strategies are applied to combine spatial and depth features.An eye-tracking experiment is conducted for the validation of the proposed model.

In this paper, we investigate a new cross layer PHYsical/APPlication (PHY-APP) communication strategy for scalable JPEG 2000 wireless image transmission over a realistic Multiple-Input Multiple-Output (MIMO) system. To exploit the channel diversity, we use a closed-loop MIMO-OFDM scheme. In the proposed scheme, the MIMO channel is decomposed into several hierarchical Single-Input Single-Output (SISO) subchannels by using a precoding approach. The scalable bitstream is divided into hierarchical quality layers which are passing through these SISO subchannels. In this paper, we propose a Joint Source-Channel (JSC) decoding approach based on soft-inputs decoding techniques to decrease the error rates at the reception without introducing extra redundancy. This scheme involves the serial concatenation of a soft-input soft-output Reed-Solomon (RS) decoder and a soft-input arithmetic decoder that were integrated into the JPEG 2000 wireless decoder. The objective of our approach is to guarantee the Quality of Service (QoS) required by the user for varying channel states. To this end, a link adaption strategy adjusting all the systems parameters of each SISO sub-channel (number of used subchannels, modulation order, Forward Error Correction (FEC) code capability, source coding rate) is also adopted in order to maximize the image visual quality at the reception. Thus, Unequal Error Protection (UEP), Unequal Power Allocation (UPA), adaptive modulation and source coding rate are provided for each quality layer. Simulation results of the optimized adaptive strategy illustrate good image quality improvements at the receiver side compared to a non adaptive strategy, with significant Peak Signal to Noise Ratio (PSNR) gains, especially for a realistic noisy channel provided by a 3D ray-tracing software. HighlightsNew transmission and coding strategies for JPEG 2000 wireless com-pressed images are proposed.A joint implementation of UEP, UPA, adaptive modulation and soft decoding techniques are also considered.Soft-inputs decoding techniques are used to decrease the error rates at the reception.A realistic time-varying MIMO channel provided by a 3D ray-tracing software is adopted.Results show a significant improvement in quality compared to non adaptive strategy.

International audience; Today's advanced simulators facilitate thorough studies on Vehicular Ad hoc NETworks (VANETs). However the choice of the physical layer model in such simulators is a crucial issue that impacts the results. A solution to this challenge might be found with a hybrid model. In this paper, we propose a semi-deterministic channel propagation model for VANETs called UM-CRT. It is based on CRT (Communication Ray Tracer) and SCME--UM (Spatial Channel Model Extended--Urban Micro) which are, respectively, a deterministic channel simulator and a statistical channel model. It uses a process which adjusts the statistical model using relevant parameters obtained from the deterministic simulator. To evaluate realistic VANET transmissions, we have integrated our hybrid model in fully compliant 802.11 p and 802.11 n physical layers. This framework is then used with the NS-2 network simulator. Our simulation results show that UM-CRT is adapted for VANETs simulations in urban areas as it gives a good approximation of realistic channel propagation mechanisms while improving significantly simulation time.

International audience; In this paper, the decomposition of nD-rotations is studied. Using this decomposition, nD-rotations are classified and properties are underlined. A generalization of the algorithm previously presented by the authors to decompose nD-rotation into planar rotations is proposed. Since our framework includes experimental applications, we designed a method that is somewhat robust to noise. An alternate algorithm based on the Schur decomposition is investigated. A comparison between both methods is finally provided.

International audience; This article proposes a study of the recent quaternionic wavelet transform (QWT) from a practical point of view through a digital image analysis application. Based on a theoretic 2D generalization of the analytic signal leading to a strong 2D signal modeling, this representation uses actual 2D analytic wavelets and yields subbands having a shift-invariant magnitude and a 3-angle phase, using the quaternion algebra. Our experiment furthers the understanding of this quite sophisticated tool, and shows its practical interest by a clear improvement of a famous wavelet application: texture classification. Thanks to coherent multiscale analysis brought by the QWT we obtain better classification results than with standard wavelets in a similar process.

International audience; We present a new type of point process, called Grouping/Degrouping Point Process (GDPP), which aim is to select a set of regions of a volume associated to an object or a Region of Interest (ROI). These regions can be obtained from a first low-level region-based segmentation for example. Geometrical and topological information of regions as localisation, adjacency, number of holes ..., are introduced in potentials which computation is done from a population of points which fall in these regions. Thus, a population of points can iteratively converge using Simulated Annealing and therefore select an optimal set of regions. In the paper, we provide the definition of region based potentials and birth and death moves used in a Reversible Jump Monte Carlo Markov Chain method. We also propose special birth and death moves using adjacency of regions. Simulations are done on Positron Emission Tomography. They show the possibility to estimate coherent sets of regions using GDPP as these sets make sense with ROIs defined by a clinician. Implementation of the process implies manipulation of 3D regions. Topological maps have been used as they permit an efficient computation of geometrical and topological properties of 3D regions and they provide a basis that allows further developments.

International audience; In this paper, we use Conley index theory to develop necessary conditions for stability of equilibrium and periodic solutions of nonlinear continuous-time systems. The Conley index is a topological generalization of the Morse theory which has been developed to analyze dynamical systems using topological methods. In particular, the Conley index of an invariant set with respect to a dynamical system is defined as the relative homology of an index pair for the invariant set. The Conley index can then be used to examine the structure of the system invariant set as well as the system dynamics within the invariant set, including system stability properties. Efficient numerical algorithms using homology theory have been developed in the literature to compute the Conley index and can be used to deduce the stability properties of nonlinear dynamical systems.

International audience; In this paper we propose a novel scheme for real time SVC-based video transmission over MIMO channels in the context of Joint Source Channel Coding (JSCC). This scheme compares the transmission of the H.264/SVC video over four precoder solutions, namely Max-SNR, WF, QoS and E-dmin. We exploit the high flexibility of the QoS precoder to minimize the total distortion of the received video. The proposed adaptive QoS precoder takes into account the scalability of the H.264/SVC standard jointly with the instantaneous MIMO channel statue. Finally, the proposed scheme is evaluated over both statistical and time Optimal power allocation varying realistic MIMO channels. This study provides the performance of these four precoder designs in term of BER, ML decoder complexity and the quality of the received video. We show that the precoder solutions providing the best BER MIMO channels performance are not usually the most appropriate for real time video transmission. However, the adaptive QoS precoder which uses three configurations, by considering both the importance of the video bit-QoS precoder stream and the channel statue, provides the best Rate-Distortion performance regardless the channel conditions. We assess the accuracy of these four precoder solutions against channel estimation errors over time varying realistic MIMO channel. The results shows that the adaptive QoS precoder remains robust against channel estimation errors even at high mobility speed.

International audience; This article presents a new method for fitting a digital line or plane to a given set of points in a 2D or 3D image in the presence of noise by maximizing the number of inliers, namely the consensus set. By using a digital model instead of a continuous one, we show that we can generate all possible consensus sets for model fitting. We present a deterministic algorithm that efficiently searches the optimal solution with time complexity O(Nd log N) for dimension d, where d = 2,3, together with space complexity O(N) where N is the number of points.

International audience; The uprising number of applications that involve very large images with resolutions greater than 30\,000$\times$30\,000 raises major memory management issues. Firstly, the amount of data usually prevents such images from being processed globally and therefore, designing a global image partition raises several issues. Secondly, a multi-resolution approach is necessary since an analysis only based on the highest resolution may miss global features revealed at lower resolutions. This paper introduces the tiled top-down pyramidal framework which addresses these two main constraints. Our model provides a full representation of multi-resolution images with both geometrical and topological relationships. The advantage of a top-down construction scheme is twofold: the focus of attention only refines regions of interest which results in a reduction of the amount of required memory and in a refinement process that may take into account hierarchical features from previous segmentations. Moreover, the top-down model is combined with a decomposition in tiles to provide an accurate memory bounding while allowing global analysis of large images.

International audience; This paper is concerned with the study of observability properties of systems without inputs via homogeneous approximations. This approximation is induced by a filtration on the space of observation. A corresponding filtration on a Lie algebra of vector fields is defined and allows to construct the approximation that preserve observability properties. An explicit construction is given.

International audience; This article presents a comparison of different color spaces including RGB, IHLS and L*a*b* for color texture characterization. This comparison is based on the fusion of the independent spatial structure and color feature cues. In IHLS and L*a*b*, two channel complex color images are created from the luminance and the chrominance values. For such images, two dimensional complex multichannel linear prediction models are used to perform parametric power spectrum estimation and the structure feature cues are computed from this estimated power spectrum. Quantitative comparison of auto spectra of luminance and combined chrominance channels for different color spaces is done. This comparison is based on the degree of decorrelation between luminance and chrominance information provided by different color space transformations. Three dimensional histograms are used as color feature cues. Then, to classify color textures, Kullback-Leibler divergence based symmetric distance measures are calculated for pure color, luminance structure and chrominance structure feature cues. Individual as well as combined effect of information from all feature cues on classification results is then compared for different color spaces and different color texture data sets. The proposed color texture classification method performs better than the state of the art methods in certain cases. The L*a*b* color space gives us a better characterization of the chrominance spatial structure as well as the overall spatial structure for all of the chosen data sets. Experimental results on pixel classification of color textures are also presented and discussed.

International audience; Cosine functions have been widely used for representing Bidirectional Reflection Distribution Functions (BRDF) such as Lambert, Phong and Lafortune models. They are well suited to represent both high and low frequency signals. However, they are difficult to use with visibility and incident radiance. In most systems, the rendering equation terms are thus estimated using various methods. Several interactive rendering systems rather rely on the projection of each term onto orthonormal basis functions such as spherical harmonics or wavelets. These methods are easier to handle since the integration becomes a dot product. However, these functions are also subject to several drawbacks. For instance the number of coefficients is high for the representation of high frequency phenomena; the precomputation time required for projecting each term of the rendering equation cannot be neglected. This paper demonstrates that cosine lobes can be generalized to visibility and incoming radiance with several advantages. First, cosine lobes do not form an orthonormal basis of functions and the number of parameters remains naturally adapted to the signal. This is very interesting for complex and high frequency functions: glossy BRDF or small light sources for instance. We also use this property for reducing the number of parameters as the computation goes along. Second, Lambert, Phong and Lafortune BRDF models are already used in many rendering systems. Since they already rely on this representation, no transformation into other types of model is necessary. This paper shows how it is possible to rapidly integrate the product of cosine lobes. As a demonstration of our methodology, we propose an interactive rendering system for direct lighting, including soft shadows and spatially varying materials.

International audience; Topological properties of the domain of attraction for dynamical systems are investigated. The main purpose of this paper is to prove that a compact, asymptotically stable attractor of a dynamical system defined on a locally compact metric space is a deformation retract of its domain of attraction, in a weak sense that is made precise. Under additional local assumptions, the attractor can be shown to be a retract, a deformation retract, or a strong deformation retract. The well known result that the domain of attraction of an asymptotically stable equilibrium is contractible follows as a corollary.

Vision is a complex process that integrates multiple aspects of an image: spatial frequencies, topology and colour. Unfortunately, so far, all these elements were independently took into consideration for the development of image and video quality metrics, therefore we propose an approach that blends together all of them. Our approach allows for the analysis of the complexity of colour images in the RGB colour space, based on the probabilistic algorithm for calculating the fractal dimension and lacunarity. Given that all the existing fractal approaches are defined only for gray-scale images, we extend them to the colour domain. We show how these two colour fractal features capture the multiple aspects that characterize the degradation of the video signal, based on the hypothesis that the quality degradation perceived by the user is directly proportional to the modification of the fractal complexity. We claim that the two colour fractal measures can objectively assess the quality of the video signal and they can be used as metrics for the user-perceived video quality degradation and we validated them through experimental results obtained for an MPEG-4 video streaming application; finally, the results are compared against the ones given by unanimously-accepted metrics and subjective tests.

International audience; Several wireless communication systems are developed for communication needs between train and ground and between trains in the railway or mass transit domains. They are developed for operational needs for security and comfort. In order to deploy these systems in specific environments, such as tunnels, straight or curved, rectangular or arch-shaped section, specific propagation models have to be developed. A modelisation of the radio wave propagation in straight arch-shaped tunnels is realized by using asymptotic methods, such as Ray Tracing and Ray Launching, combined with the tessellation of the arched section. A method of interpolation of the facets' normals was implemented in order to minimize the error made when using the tessellation. Results obtained are validated by comparison to the literature and to measurement results.

International audience; An important factor in electromagnetic wave propagation simulation, performed by a 3D ray-tracing method is the modeling of the environment. Results being sensitive to descriptive accuracy, this paper presents a study on the effect of indoor environment modeling precision on multiple input multiple output (MIMO) channel characterization. In order to give some diversity to the results, the investigation takes into account two unfurnished indoor environments. The first environment investigated is the entrance hall of a building at the University of Poitiers, while the second is a more confined environment and is represented by the first floor of the authors' laboratory. For these two indoor environments, four levels of description are proposed in order to establish the geometrical and electrical modeling impact on MIMO channel characterization. Results are obtained by analyzing the capacity, variation in correlation, multipath richness and eigenvalues in relation to the polarization, the presence of line-of-sight (LOS) or non-LOS (NLOS) configurations. The effects of the spacing between antennas and the number of transmitter and receiver antennas are also investigated.

International audience; In this paper we study the relationship between the Euclidean and the discrete space. We study discrete operations based on Euclidean functions: the discrete smooth scaling and the discrete-continuous rotation. Conversely, we study Euclidean oper- ations based on discrete functions: the discrete based simplification, the Euclidean- discrete union and the Euclidean-discrete co-refinement. These operations operate partly in the discrete and partly in the continuous space. Especially for the discrete smooth scaling operation, we provide error bounds when different such operations are chained.

10 pages; International audience; In current parametric systems, the persistent naming issue (based on edge mappings of intersecting surfaces) is not as fully supported as it should be. Unpredictability and ambiguity of models often happen during design reevaluation within systems. This reference deficiency is widely treated in the literature, especially about non-planar entities during design construction. Although related works ensure the uniqueness of the references to topological entities, they often neglect the shape characteristics of surfaces and give results different from those expected during design reevaluation. We propose in this paper a method to add some additional information about surfaces to improve such works. We compute those information by decomposing surfaces according to hump(s) and/or hollow(s). More precisely, our method use local extremums and inflexion curves to obtain one hump or hollow per sub-surface. The existing matching processes replace every surface with their corresponding subsurfaces, leading to the right edge mappings.

International audience; This work is about an operational conception toolkit for real time applications. Temporal validation is usually based on tools with binary validation results. In this paper, we show how to build some quality indicators able to quantify invalidity numerically. First, we present an automata-based model and then its geometrical adaptation, which permits to compute a measure of invalidity

International audience

International audience

International audience; The physical validation of computer-generated images (CGIs) has received a lot of attention from the computer graphics community, as opposed to the assessment of these images' psychovisual quality. The field indeed lacks the proper tools to quantify the perceptual quality of a CGI; this paper engages in the construction of such a metric. A psychovisual experiment was submitted to a representative panel of observers, where participants were asked to score the overall quality and aspects of this quality for several CGIs. An analytical quality model, fit to the data, next gives insight into the relative perceptual importances of these aspects. Accuracy in the simulation of shadows, good contrast and absence of noise were found to have a major impact on the perceived quality, rather than precise anti-aliasing and faithfull color bleeding.

International audience; Textured surface analysis is essential for many applications. In this paper, we present a three-dimensional (3D) recovery approach for real textured surfaces based on photometric stereo. The aim is to be able to reconstruct the textured surfaces in 3D with a high degree of accuracy. For this, the proposed method uses a sequence of six images and a Lambertian bidirectional reflectance distribution function (BRDF) to recover the surface height map. A hierarchical selection of these images is employed to eliminate the effects of shadows and highlights for all surface facets. To evaluate the performances of our method, we compare it to other traditional photometric stereo methods on real textured surfaces using six or more images.

Présentation invitée de l'article du même nom publié en 2012 dans la revue Computer Graphics Forum.; International audience; This paper presents a novel approach to compute high quality and noise-free soft shadows using exact visibility computations. This work relies on a theoretical framework allowing to group lines according to the geometry they intersect. From this study, we derive a new algorithm encoding lazily the visibility from a polygon. Contrary to previous works on from-polygon visibility, our approach is very robust and straightforward to implement. We apply this algorithm to solve exactly and efficiently the visibility of an area light source from any point in a scene. As a consequence, results are not sensitive to noise, contrary to soft shadows methods based on area light source sampling. We demonstrate the reliability of our approach on different scenes and configurations.

La problématique centrale de mes travaux est l'étude de méthodes de décomposition discrète Temps-échelle. Ils s'inscrivent pleinement dans l'équipe projet ICONES, dont j'assure l'animation. Dans ce cadre, je me suis concentré, dès mon intégration dans le laboratoire sur les décompositions atomiques discrètes reconstructibles définies dans le domaine général des ondelettes. Toutes ces méthodes possèdent maintenant un cadre général et notamment la transformée en ondelettes standard. Cependant, il existe encore certaines limites : construction d'une représentation efficace d'éléments géométriques 2-D de l'image, généralisation de ces transformées à des signaux à valeurs vectorielles reste. L'objectif recherché est donc une représentation liée à une structure de données sous forme de primitives, facilement calculable par un algorithme, telle que l'ensemble de ces primitives permettent la reconstruction de la donnée d'origine. Il faut par ailleurs que l'on puisse ordonner ces primitives et que l'on puisse donner un sens à ces primitives. Dans ce cadre, mon travail va se découper suivant trois axes de recherche et un axe applicatif fort. Approche géométrique. Nous avons étudié un outil qui permet la représentation d'images d'une manière optimale (au sens des ruptures linéaires) : la transformée Ridgelet. Grâce à une collaboration avec les chercheurs en géométrie discrète du département XLIM-SIC nous avons proposé une transformée Ridgelet/Curvelets paramétrable, rapide et inversible, obtenant des résultats en termes de restauration de données 2-D très satisfaisants, notamment en vidéo. De plus, la simplicité et la rigueur dans la définition apportée par la géométrie discrète nous a permis d'étendre ce travail à toute dimension. Suivant ce même but, à savoir capter la géométrie pour adapter la transformée, nous avons proposé dans le cadre d'un partenariat avec OrangeLab une représentation en ondelettes géodésiques qui se propose d'introduire une modélisation de la topologie dans un schéma numérique de calcul du type Lifting Scheme. Il s'appuie sur la construction d'un graphe qui doit capter l'organisation structurelle de l'image à traiter. De ce graphe, une distance entre chaque pixel de l'image est estimée et introduite dans l'algorithme de décomposition. Couleur et MultiCanaux. L'objet de ce second axe a été d'examiner des approches plus globales pour l'extension des briques élémentaires issues du signal dans le cadre des images couleur. Suivant cette idée, j'ai débuté dès 2001 une étude concernant l'utilisation des quaternions. Cette recherche a concerné la redéfinition des traitements classiques couleur en termes quaternionique, comme le filtrage, la modification des caractéristiques couleur ... Ceci a fait l'objet d'une thèse, notamment à travers la redéfinition des manipulations couleur avec un formalisme lié à l'algèbre géométrique. Approche orientée. Le simple découpage horizontal, vertical et selon les deux diagonales, lié à l'approche classique, n'est pas suffisant à la fois en termes de description du signal mais aussi par rapport à ce que l'on connait du système visuel humain. L'objectif est de définir une représentation associée à des directions d'analyse plus adaptatives, invariante soit à la translation soit à la rotation. La transformée en ondelettes complexes suivant l'arbre dual constitue une réponse possible. A partir de ces travaux, nous étudions une approche alternative reposant sur les travaux de Bulow et al. et Choi et al. sur le signal analytique 2-D à partir des nombres quaternioniques. Ce travail permet de mettre en place une transformée en ondelettes quaternioniques pour la compression et la classification introduisant une nouvelle information : une notion de phase. La généralisation de ce travail nous mène à la définition d'ondelettes monogéniques qui peuvent être considérées en termes signalistique, comme les décompositions les plus abouties. Cependant de nombreuses questions restent ouvertes, et la première d'entre-elles étant la définition d'un schéma numérique stable. Sécurité. Depuis plusieurs années maintenant, j'ai développé un travail sur la protection électronique des images et vidéo, qui a donné lieu à 2 brevets sur le tatouage basé ondelettes, et intégrant des données psychovisuelles de perception des couleurs. Ces travaux se sont maintenant poursuivis dans le cadre d'une collaboration avec le département DMI de XLIM. L'objectif visé est l'intégration des différentes familles de codes correcteurs d'erreur afin d'améliorer la robustesse d'une stratégie de tatouage. Les résultats montrent effectivement une amélioration de la robustesse de l'algorithme breveté. Nous poursuivons ce travail sous un angle un peu différent à travers la définition d'algorithme de stétaganalyse en utilisant les compétences de l'équipe Icones autour des métriques de qualité.

International audience; The fusion of regions from segmentation is often a required step in order to analyze objects in the scene. It allows to group under a same closure property the partitions of an image, which denote the same object. In this way, several studies have been made. Most of the algorithms are based on global information to evaluate the relevance of region combinations and fusion. In this paper, we propose a merging process based on Gestalt vision's theory. The particularity of our method is to preserve photographer's region of interest thanks to the definition of a new criterion based on Di Zenzo gradient. Thus, region fusion preserves photographer's relevant objects of the scene. The performance of the proposed method is evaluated using various natural images, including Berkeley image database and its accuracy is shown.

International audience; Many research on VANETs present new propagation models. This paper focus on the last optimisation of the UM-CRT model proposed by the XLIM-SIC Laboratory. The model is simplified in order to obtain a faster computation time without reducing the realism of the UM-CRT model and new criteria computed from the impulse response are also shown.

International audience; This paper presents a generalization of the notion of circumcenter as the intersection of perpendicular bisectors. We define Generalized Perpendicular Bisectors between two regions as an area where each point is the center of at least one circle crossing both regions. This allows us to determine all the possible discrete circle centers that cross a given set of pixels. The possible radii can then easily be determined. This exhaustive digital circle parameter computation is adapted to various types of circles/digitization schemes such as Naïve, Pythagorean and standard/supercover circles.

International audience; One of the advantages of the Discrete Wavelet Transform (DWT) compared to Fourier Transform (e.g. Discrete Cosine Transform DCT) is its ability to provide both spatial and frequency localization of image energy. However, WT coefficients, like DCT coefficients, are defined by magnitude as well as sign. While algorithms exist for the coding of wavelet coefficients magnitude, there are no efficient for coding their sign. In this paper, we propose a new method based on separate entropy coding of sign and magnitude of wavelet coefficients. The proposed method is applied to the standard color test images Lena, Peppers, and Mandrill. We have shown that sign information of wavelet coefficients as well for the luminance as for the chrominance, and the refinement information of the quantized wavelet coefficients may not be encoded by an estimated probability of 0.5. The proposed method is evaluated; the results obtained are compared to JPEG2000 and SPIHT codec. We have shown that the proposed method has significantly outperformed the JPEG2000 and SPIHT codec as well in terms of PSNR as in subjective quality. We have proved, by an original mathematical analysis of the entropy, that the proposed method uses a minimum bit allocation in the sign information coding.

National audience; Ce poster porte sur la conception et le développement d'une nouvelle technique de modélisation de la propagation d'ondes radioélectriques dans des tunnels non rectilignes de section non droite rencontrés en milieu ferroviaire pour des applications métros. Les solutions développées sous forme de modules appelés " plugin ", ont été ajoutées au noyau de l'outil de modélisation de la propagation d'ondes, RaPSor, issu des recherches menées par le laboratoire XLIM-SIC de l'Université de Poitiers. Dans ce travail, nous montrons que les techniques classiques mises en œuvre afin de modéliser la propagation des ondes en tunnel rectiligne infini de section rectangulaire, ne peuvent pas être transposées au cas des tunnels courbes de section courbe. Ainsi, nous avons développé une nouvelle méthode qui se fonde sur une représentation analytique spécifique des surfaces courbes associée à un lancer de rayons et à une optimisation originale des trajets reçus. La répartition uniforme des rayons lancés à l'émission repose sur l'utilisation de séquences quasi-aléatoires de Hammersley. Une sphère adaptative est utilisée à la réception. L'optimisation des rayons est réalisée avec la technique de minimisation de Levenberg-Marquardt. Il s'agit de minimiser la distance totale des trajets afin d'approcher les propriétés des trajets réels au sens de l'Optique Géométrique. L'ensemble des méthodes développées sont testées et validées par comparaisons avec des résultats de la littérature mais aussi avec des résultats de mesures.

National audience; Ce papier présente une comparaison des résultats de modélisation du canal radio en milieu industriel indoor, issu du modèle statistique WINNER d'une part, et d'un modèle déterministe tridimensionnel basé rayons allié à des méthodes asymptotiques enfréquence d'autre part. Cette comparaison est réalisée à partir des paramètres larges bandes que sont le retard moyen et la dispersion des retards calculés sur un grand nombre de réponses impulsionnelles issues des deux modèles. Les premiers résultats montrent que le comportement est similaire d'un modèle à l'autre et que WINNER est plus pessimiste.

National audience; Cet exposé est centré sur la modélisation et la simulation du canal radioélectrique en s'appuyant sur des méthodes à rayon. Les méthodes développées ont pour principal objectif d'allier à la précision de la prédiction, un temps de simulation acceptable pour les applications de télécommunication visées. Ces dernières sont très souvent associées à des environnements assez complexes tant du point de vue géométrique (grand nombre d'obstacles de formes variées) qu'électrique de part la diversité des matériaux (permittivité et conductivité). Ainsi, pour modéliser la propagation des ondes radio, nous nous sommes tout d'abord appuyés sur une approche 3D vectorielle et déterministe. Elle repose sur l'optique géométrique et la Théorie Uniforme de la Diffraction associés à une technique de tracé de rayons. Cette approche a été validée expérimentalement et par comparaison à des méthodes à formulation rigoureuse dans plusieurs configurations outdoor et indoor entre 1 et quelques GHz. Sur cette base, des travaux complémentaires ont été développés pour traiter des structures plus complexes qu'un obstacle canonique. Le cas de surfaces courbes permettant d'analyser la propagation dans des tunnels sera présenté ainsi qu'une hybridation avec la FDDT pour étudier des objets de formes complexes et de dimensions voisines de celles de la longueur d'onde. De façon complémentaire, comme les approches précédentes ne peuvent être appliquées pour étudier un grand nombre de liaisons d'autant plus dans un contexte de mobilité, des travaux basés sur une approche " semi déterministe " sont développés. Ils consistent à combiner une méthode à rayon avec une modélisation statistique de la propagation des ondes pour tirer profit du caractère " spécifique au site étudié " d'une méthode déterministe et de la rapidité d'un modèle statistique. Des exemples d'études dans le cas des réseaux VANETS (Vehicular ad'hoc network) et des communications satellite/mobile terrestre seront présentés.

International audience; This paper shows the information criteria (IC) performances in feature selection framework. Feature selection aims to select a representative subset among a wide set of features. We apply this approach to classify an hand segmented image. The performance is tested using various feature selection schemes (SFS, SBS, SFFS and SBFS) to select the candidate subsets. The accuracy of the approach is based on a good quality of the joint probability density approximation of the combined features. They are obtained using histogram optimized thanks to the adaptive arithmetic coding principles. Our approach is tested on different reference data. The subsets quality is evaluated using correct classification rate computed on multiple classifiers. Results show stability and convergence properties of this tool and its ability to select representative subsets (in the sense that the subset of feature is a good characterization of the classes in which the data belong). Information Criteria could be used for feature selection as a good alternative to other criteria.

International audience; This paper presents a power amplifier (PA) cascade model, using two polynomial functions for amplitude and phase variations and multiple delay taps. This model allows an accurate modeling of the output envelope distortion and spectral asymmetries. To extract the model coefficients and delays for each non linearity order, an analytical method is developed. This method uses the Fourier transform of the output gain and phase to calculate all the parameters. To demonstrate the performances of this modeling approach, comparison between measured and simulated data has been performed.

Cette présentation rappelle les biais inhérents à l'estimation de densité lors de l'utilisation de cartes de photons et propose des solutions pour les réduire ou les supprimer.

International audience; To face QoS constraints of multimedia applications, many researches focus on improving the standard MANET routing protocols. Nevertheless, we find few works about the comparison of enhancement heuristics. In the literature dealing with this question, the comparison parameters are measured through simulations performed mainly under unrealistic conditions. In this paper, we compare BER-based enhancements of AODV and OLSR. Our performance evaluation heavily relies on simulation conducted under realistic wave propagation and mobility models. Mobility and multi-communication effects are thus evaluated. We point out that link quality, influenced by obstacles in the propagation field, should be taken into account in MANET routing protocols enhancement approaches.

To satisfy the smart grid electrical network, communication systems in high-voltage substations have to be installed in order to control equipments. Considering that those substations were not necessarily designed for adding communication networks, one of the most appropriate solutions is to use wireless sensor network (WSN). However, the high voltage transported through the station generates a strong and specific radio noise. In order to prepare for such a network, the electromagnetic environment has to be characterized and tests in laboratories have to be performed to estimate the communication performances. This paper presents a method for measuring the noise due to high voltage and more particularly the impulsive noise. In the laboratory, we generate the impulsive noise using two specimens, and we show that these laboratory measurements validate the field measurements of Pakala et al. For the two specimens, it aims to link the noise characteristics (magnitude and frequency) with the specimen parameters (power supply and geometric dimensions) to predict the environments where wireless communications can be troublesome. By using different sets of this measured noise, we show that the statistical model of Middleton Class A can be used to model the impulsive noise in high-voltage substations better than the Gaussian model. We consider a cooperative multiple-input-multiple-output (MIMO) system to achieve the wireless sensor communication. This system uses recent MIMO techniques based on precoding like max-d min and P-OSM precoders. The MIMO precoder-based cooperative system is a potential candidate for energy saving in WSN since energy efficiency optimization is a very important critical issue. Since MIMO precoders are with Gaussian noise assumption, we evaluate the performance of several MIMO precoders in the presence of impulsive noise using estimated parameters from the measured noise.