|
|
Publications of 1998
Result of the query in the list of publications :
5 Articles |
1 - Variational approach for edge preserving regularization using coupled PDE's.
S. Teboul and L. Blanc-Féraud and G. Aubert and M. Barlaud. IEEE Trans. Image Processing, 7(3): pages 387-397, March 1998.
@ARTICLE{lbf98,
|
| author |
= |
{Teboul, S. and Blanc-Féraud, L. and Aubert, G. and Barlaud, M.}, |
| title |
= |
{Variational approach for edge preserving regularization using coupled PDE's}, |
| year |
= |
{1998}, |
| month |
= |
{March}, |
| journal |
= |
{IEEE Trans. Image Processing}, |
| volume |
= |
{7}, |
| number |
= |
{3}, |
| pages |
= |
{387-397}, |
| keyword |
= |
{} |
| } |
|
2 - Combined constraints for efficient algebraic regularized methods.
I. Laurette and J. Darcourt and L. Blanc-Féraud and P.M. Koulibaly and M. Barlaud. Physics in Medicine and Biology, 34(4): pages 991-1000, 1998.
@ARTICLE{lbf98a,
|
| author |
= |
{Laurette, I. and Darcourt, J. and Blanc-Féraud, L. and Koulibaly, P.M. and Barlaud, M.}, |
| title |
= |
{Combined constraints for efficient algebraic regularized methods}, |
| year |
= |
{1998}, |
| journal |
= |
{Physics in Medicine and Biology}, |
| volume |
= |
{34}, |
| number |
= |
{4}, |
| pages |
= |
{991-1000}, |
| keyword |
= |
{} |
| } |
|
3 - A generalized sampling theory without bandlimiting constraints.
M. Unser and J. Zerubia. IEEE Trans. on Circuits And Systems II, 45(8): pages 959-969, 1998.
@ARTICLE{jz98b,
|
| author |
= |
{Unser, M. and Zerubia, J.}, |
| title |
= |
{A generalized sampling theory without bandlimiting constraints}, |
| year |
= |
{1998}, |
| journal |
= |
{IEEE Trans. on Circuits And Systems II}, |
| volume |
= |
{45}, |
| number |
= |
{8}, |
| pages |
= |
{959-969}, |
| keyword |
= |
{} |
| } |
|
4 - fMRI Signal Restoration Using an Edge Preserving Spatio-temporal Markov Random Field.
X. Descombes and F. Kruggel and Y. von Cramon. NeuroImage, 8: pages 340-349, 1998.
Keywords : fMRI, Restoration, Markov Fields.
Copyright : published in NeuroIMage by Elsevier
||http://www.elsevier.com/wps/find/homepage.cws_home
@ARTICLE{descombes98d,
|
| author |
= |
{Descombes, X. and Kruggel, F. and von Cramon, Y.}, |
| title |
= |
{fMRI Signal Restoration Using an Edge Preserving Spatio-temporal Markov Random Field}, |
| year |
= |
{1998}, |
| journal |
= |
{NeuroImage}, |
| volume |
= |
{8}, |
| pages |
= |
{340-349}, |
| pdf |
= |
{ftp://ftp-sop.inria.fr/ariana/Articles/1998_descombes98d.pdf}, |
| keyword |
= |
{fMRI, Restoration, Markov Fields} |
| } |
|
5 - Spatio-temporal fMRI analysis using Markov Random Fields.
X. Descombes and F. Kruggel and Y. Von Cramon. IEEE Trans. Medical Imaging, 17: pages 1028-1039, 1998.
Note : to appear.
Keywords : fMRI, Markov Random Fields.
@ARTICLE{descombes98,
|
| author |
= |
{Descombes, X. and Kruggel, F. and Von Cramon, Y.}, |
| title |
= |
{Spatio-temporal fMRI analysis using Markov Random Fields}, |
| year |
= |
{1998}, |
| journal |
= |
{IEEE Trans. Medical Imaging}, |
| volume |
= |
{17}, |
| pages |
= |
{1028-1039}, |
| keyword |
= |
{fMRI, Markov Random Fields} |
| } |
|
 top of the page
2 PhD Thesis and Habilitations |
1 - Modélisation de la Redondance d'Image, Etude et Application à la Classification.
E. Volden. PhD Thesis, Ecole des Mines de Paris, January 1998.
@PHDTHESIS{volden,
|
| author |
= |
{Volden, E.}, |
| title |
= |
{Modélisation de la Redondance d'Image, Etude et Application à la Classification}, |
| year |
= |
{1998}, |
| month |
= |
{January}, |
| school |
= |
{Ecole des Mines de Paris}, |
| keyword |
= |
{} |
| } |
|
2 - Extraction de Thèmes Cartographiques dans les Images Satellites ou Aériennes. Application à la Génération de Quick-Looks Adaptatifs et à la Compression des Images.
J.M. Benharrosh. PhD Thesis, Universite de Nice Sophia Antipolis, 1998.
@PHDTHESIS{benharrosh,
|
| author |
= |
{Benharrosh, J.M.}, |
| title |
= |
{Extraction de Thèmes Cartographiques dans les Images Satellites ou Aériennes. Application à la Génération de Quick-Looks Adaptatifs et à la Compression des Images}, |
| year |
= |
{1998}, |
| school |
= |
{Universite de Nice Sophia Antipolis}, |
| url |
= |
{http://www.inria.fr/rrrt/tu-0524.html}, |
| ps |
= |
{ftp://ftp.inria.fr/INRIA/publication/Theses/TU-0524.ps.gz}, |
| keyword |
= |
{} |
| } |
|
top of the page
14 Conference articles |
1 - Active contour models for segmentation and reconstruction on medical images.
S. Teboul and L. Blanc-Féraud and G. Aubert and M. Barlaud. In Asilomar Conference, USA, November 1998.
@INPROCEEDINGS{lbf98g,
|
| author |
= |
{Teboul, S. and Blanc-Féraud, L. and Aubert, G. and Barlaud, M.}, |
| title |
= |
{Active contour models for segmentation and reconstruction on medical images}, |
| year |
= |
{1998}, |
| month |
= |
{November}, |
| booktitle |
= |
{Asilomar Conference}, |
| address |
= |
{USA}, |
| keyword |
= |
{} |
| } |
|
2 - Image Retrieval and Indexing: A Hierarchical Approach in Computing the Distance between Textured Images.
R. Stoica and J. Zerubia and J.M. Francos. In Proc. IEEE International Conference on Image Processing (ICIP), Chicago, USA, October 1998.
@INPROCEEDINGS{stoica98a,
|
| author |
= |
{Stoica, R. and Zerubia, J. and Francos, J.M.}, |
| title |
= |
{Image Retrieval and Indexing: A Hierarchical Approach in Computing the Distance between Textured Images}, |
| year |
= |
{1998}, |
| month |
= |
{October}, |
| booktitle |
= |
{Proc. IEEE International Conference on Image Processing (ICIP)}, |
| address |
= |
{Chicago, USA}, |
| keyword |
= |
{} |
| } |
|
3 - Motion-based segmentation by means of active contours.
R. Ciampini and L. Blanc-Féraud and M. Barlaud and E. Salerno. In Proc. IEEE International Conference on Image Processing (ICIP), Chicago, USA, October 1998.
@INPROCEEDINGS{lbf98f,
|
| author |
= |
{Ciampini, R. and Blanc-Féraud, L. and Barlaud, M. and Salerno, E.}, |
| title |
= |
{Motion-based segmentation by means of active contours}, |
| year |
= |
{1998}, |
| month |
= |
{October}, |
| booktitle |
= |
{Proc. IEEE International Conference on Image Processing (ICIP)}, |
| address |
= |
{Chicago, USA}, |
| keyword |
= |
{} |
| } |
|
4 - Unsupervised deconvolution of satellite images.
M. Khoumri and L. Blanc-Féraud and J. Zerubia. In Proc. IEEE International Conference on Image Processing (ICIP), Chicago, USA, October 1998.
@INPROCEEDINGS{lbf98d,
|
| author |
= |
{Khoumri, M. and Blanc-Féraud, L. and Zerubia, J.}, |
| title |
= |
{Unsupervised deconvolution of satellite images}, |
| year |
= |
{1998}, |
| month |
= |
{October}, |
| booktitle |
= |
{Proc. IEEE International Conference on Image Processing (ICIP)}, |
| address |
= |
{Chicago, USA}, |
| keyword |
= |
{} |
| } |
|
5 - Segmentation of urban areas in SPOT images using MRF.
F. Richard and F. Falzon and J. Zerubia and G. Giraudon. In Proc. European Signal Processing Conference (EUSIPCO), Rhodes, September 1998.
@INPROCEEDINGS{jz98,
|
| author |
= |
{Richard, F. and Falzon, F. and Zerubia, J. and Giraudon, G.}, |
| title |
= |
{Segmentation of urban areas in SPOT images using MRF}, |
| year |
= |
{1998}, |
| month |
= |
{September}, |
| booktitle |
= |
{Proc. European Signal Processing Conference (EUSIPCO)}, |
| address |
= |
{Rhodes}, |
| keyword |
= |
{} |
| } |
|
6 - An unsupervised clustering method using the entropy minimization.
G. Palubinskas and X. Descombes and F. Kruggel. In Proc. International Conference on Pattern Recognition (ICPR), Brisbane, Australia, August 1998.
@INPROCEEDINGS{descombes98a,
|
| author |
= |
{Palubinskas, G. and Descombes, X. and Kruggel, F.}, |
| title |
= |
{An unsupervised clustering method using the entropy minimization}, |
| year |
= |
{1998}, |
| month |
= |
{August}, |
| booktitle |
= |
{Proc. International Conference on Pattern Recognition (ICPR)}, |
| address |
= |
{Brisbane, Australia}, |
| keyword |
= |
{} |
| } |
|
7 - Hyperparameter estimation of a variational model using a stochastic gradient.
J. Zerubia and L. Blanc-Féraud. In Proc. International Symposium on Optical Science, Engineering and Instrumentation, San Diego, USA, July 1998.
@INPROCEEDINGS{lbf98b,
|
| author |
= |
{Zerubia, J. and Blanc-Féraud, L.}, |
| title |
= |
{Hyperparameter estimation of a variational model using a stochastic gradient}, |
| year |
= |
{1998}, |
| month |
= |
{July}, |
| booktitle |
= |
{Proc. International Symposium on Optical Science, Engineering and Instrumentation}, |
| address |
= |
{San Diego, USA}, |
| keyword |
= |
{} |
| } |
|
8 - Real-Time 3D vizualization for medical images reconstruction and segmentation.
F. Diard and S. Teboul and L. Blanc-Féraud and M. Barlaud. In IEEE Image and MultiDimensional Signal Processing, Autriche, July 1998.
@INPROCEEDINGS{lbf98c,
|
| author |
= |
{Diard, F. and Teboul, S. and Blanc-Féraud, L. and Barlaud, M.}, |
| title |
= |
{Real-Time 3D vizualization for medical images reconstruction and segmentation}, |
| year |
= |
{1998}, |
| month |
= |
{July}, |
| booktitle |
= |
{IEEE Image and MultiDimensional Signal Processing}, |
| address |
= |
{Autriche}, |
| keyword |
= |
{} |
| } |
|
9 - On the use of nonlinear regularization in inverse methods for the solar tachocline profile determination.
T. Corbard and G. Berthomieu and J. Provost and L. Blanc-Féraud. In Structure and Dynamics of the Interior of the Sun and Sun-like Stars, Vol. ESA SP-418, Ed. S.G. Korzennik & A. Wilson, Publ. ESA Publications Division, Noordwijk, The Netherlands, July 1998.
@INPROCEEDINGS{lbf98h,
|
| author |
= |
{Corbard, T. and Berthomieu, G. and Provost, J. and Blanc-Féraud, L.}, |
| title |
= |
{On the use of nonlinear regularization in inverse methods for the solar tachocline profile determination}, |
| year |
= |
{1998}, |
| month |
= |
{July}, |
| booktitle |
= |
{Structure and Dynamics of the Interior of the Sun and Sun-like Stars}, |
| volume |
= |
{ESA SP-418}, |
| editor |
= |
{S.G. Korzennik & A. Wilson}, |
| publisher |
= |
{ESA Publications Division}, |
| address |
= |
{Noordwijk, The Netherlands}, |
| keyword |
= |
{} |
| } |
|
10 - Preprocessing of fMR Datasets.
F. Kruggel and X. Descombes and Y. von Cramon. In Proc. IEEE Workshop on Biomedical Image Analysi, pages 211-220, Los Alamitos, USA, June 1998.
@INPROCEEDINGS{descombes98c,
|
| author |
= |
{Kruggel, F. and Descombes, X. and von Cramon, Y.}, |
| title |
= |
{Preprocessing of fMR Datasets}, |
| year |
= |
{1998}, |
| month |
= |
{June}, |
| booktitle |
= |
{Proc. IEEE Workshop on Biomedical Image Analysi}, |
| pages |
= |
{211-220}, |
| address |
= |
{Los Alamitos, USA}, |
| keyword |
= |
{} |
| } |
|
11 - A step toward high resolution 3D SAR.
B. Pairault and M. Berthod. In Proc. European Conference on Synthetic Aperture Radar, Friedrichshafen, Germany, May 1998.
@INPROCEEDINGS{berthod98,
|
| author |
= |
{Pairault, B. and Berthod, M.}, |
| title |
= |
{A step toward high resolution 3D SAR}, |
| year |
= |
{1998}, |
| month |
= |
{May}, |
| booktitle |
= |
{Proc. European Conference on Synthetic Aperture Radar}, |
| address |
= |
{Friedrichshafen, Germany}, |
| keyword |
= |
{} |
| } |
|
12 - The two-dimensional Wold decomposition for segmentation and indexing in image libraries.
R. Stoica and J. Zerubia and J.M. Francos. In Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Seattle, USA, May 1998.
@INPROCEEDINGS{stoica98b,
|
| author |
= |
{Stoica, R. and Zerubia, J. and Francos, J.M.}, |
| title |
= |
{The two-dimensional Wold decomposition for segmentation and indexing in image libraries}, |
| year |
= |
{1998}, |
| month |
= |
{May}, |
| booktitle |
= |
{Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}, |
| address |
= |
{Seattle, USA}, |
| keyword |
= |
{} |
| } |
|
13 - Denoising by extracting fractional order singularities.
H. Shekarforoush and J. Zerubia and M. Berthod. In Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Seattle, USA, May 1998.
@INPROCEEDINGS{jz98a,
|
| author |
= |
{Shekarforoush, H. and Zerubia, J. and Berthod, M.}, |
| title |
= |
{Denoising by extracting fractional order singularities}, |
| year |
= |
{1998}, |
| month |
= |
{May}, |
| booktitle |
= |
{Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}, |
| address |
= |
{Seattle, USA}, |
| keyword |
= |
{} |
| } |
|
14 - Die Vorerarbeitung von fMRI-Daten.
F. Kruggel and X. Descombes and Y. von Cramon. In Bildverarbeitung für die Medizin, Algorithmen - Systeme - Anwendungen, Universitätsklinikum der RWTH Aachen, Germany, March 1998.
@INPROCEEDINGS{descombes98b,
|
| author |
= |
{Kruggel, F. and Descombes, X. and von Cramon, Y.}, |
| title |
= |
{Die Vorerarbeitung von fMRI-Daten}, |
| year |
= |
{1998}, |
| month |
= |
{March}, |
| booktitle |
= |
{Bildverarbeitung für die Medizin, Algorithmen - Systeme - Anwendungen}, |
| address |
= |
{Universitätsklinikum der RWTH Aachen, Germany}, |
| keyword |
= |
{} |
| } |
|
top of the page
7 Technical and Research Reports |
1 - Indexing and retrieval in multimedia libraries through parametric texture modeling using the 2D Wold decomposition.
R. Stoica and J. Zerubia and J.M. Francos. Research Report 3594, Inria, December 1998.
Keywords : Markov Fields, Texture, Segmentation, Indexation.
@TECHREPORT{stoica98,
|
| author |
= |
{Stoica, R. and Zerubia, J. and Francos, J.M.}, |
| title |
= |
{Indexing and retrieval in multimedia libraries through parametric texture modeling using the 2D Wold decomposition}, |
| year |
= |
{1998}, |
| month |
= |
{December}, |
| institution |
= |
{Inria}, |
| type |
= |
{Research Report}, |
| number |
= |
{3594}, |
| url |
= |
{http://www.inria.fr/rrrt/rr-3594.html}, |
| pdf |
= |
{ftp://ftp.inria.fr/INRIA/publication/publi-pdf/RR/RR-3594.pdf}, |
| ps |
= |
{ftp://ftp.inria.fr/INRIA/publication/publi-psgz/RR/RR-3594.ps.gz}, |
| keyword |
= |
{Markov Fields, Texture, Segmentation, Indexation} |
| } |
Résumé :
Ce rapport présente une méthode paramétrique permettant de faire de l'indexati- on et de la recherche dans une base de données multimédia. L'indexation (étiquetage) et la recherche de données multimédia sont réalisées grâce à la modélisation paramétrique de textures qui se trouvent dans les images de la base de données. Les textures sont caracterisées par des paramètres qui servent d'indices pour la recherche dans la base de données. Afin de pouvoir identifier les différentes régions texturées d'une image et estimer les paramètres correspondants, un algorithme de segmentation-estimatio- n est proposé dans ce rapport, qui fait appel à une décomposition de Wold 2D pour le modèle de texture et à un modèle markovien pour l'étiquetage. L'indexation nécessite de définir une distance entre les images. Une nouvelle distance, inspirée de la distance de Kullback, est décrite dans ce rapport. Elle utilise les paramètres estimés correspondants au modèle 2D de chaque texture. Les résultats obtenus relativement à la segmentation et à l'indexatio- n sont proches de ceux obtenus par un opérateur humain. |
Abstract :
This paper presents a parametric method for indexing and retrieval of multimedia data in digital libraries. %Indexing (labeling) and retrieval %of multimedia data, based on the properties %of the imagery components of the stored data record, are derived. Indexing (labeling) and retrieval of the multimedia data are performed using parametric modeling of the textured segments found in the data imagery components. The estimated parametric models of the textured segments serve as their indices, and hence as indices of the entire image, as well as of the multimedia record which the image is part thereof. To achieve the ability to identify textured image regions and estimate their parameters, a joint segmentation-estimation algorithm that combines the 2-D Wold decomposition based texture model with a Markovian labeling process, is derived. Ordering and indexing of images require a definition of a distance measure between images. Using the framework of the Kullback distance between probability distributions, a new rigorous distance measure between textures is derived. The distance between any two textured image segments is evaluated using their estimated parametric models. The proposed segmentation, distance evaluation, and indexing methods are shown to produce comparable results to those obtained by a human viewer. |
|
2 - Image Classification Using a Variational Approach.
C. Samson and L. Blanc-Féraud and G. Aubert and J. Zerubia. Research Report 3523, Inria, October 1998.
Keywords : Classification, Variational methods.
@TECHREPORT{samsonRR98,
|
| author |
= |
{Samson, C. and Blanc-Féraud, L. and Aubert, G. and Zerubia, J.}, |
| title |
= |
{Image Classification Using a Variational Approach}, |
| year |
= |
{1998}, |
| month |
= |
{October}, |
| institution |
= |
{Inria}, |
| type |
= |
{Research Report}, |
| number |
= |
{3523}, |
| url |
= |
{http://www.inria.fr/rrrt/rr-3523.html}, |
| pdf |
= |
{ftp://ftp.inria.fr/INRIA/publication/publi-pdf/RR/RR-3523.pdf}, |
| ps |
= |
{ftp://ftp.inria.fr/INRIA/publication/publi-psgz/RR/RR-3523.ps.gz}, |
| keyword |
= |
{Classification, Variational methods} |
| } |
Résumé :
Dans ce rapport nous présentons un modèle variationnel destiné à la classification d'images avec processus de régularisation préservant les contours. La notion de classification étant par nature discrète (i.e. attribuer un label à chaque pixel de l'image), il existe de nombreux modèles de classification par approche probabiliste, mais les modèles variationnels abordant ce sujet sont rares. Ces dernières années, l'approche variationnelle a montré sont efficacité dans le cadre de la restauration d'images avec prise en compte des discontinuités. Dans ce travail, nous ajoutons un processus de classification permettant d'obtenir une solution formée de régions homogènes dont les frontières sont régulières (une région étant définie par l'ensemble des pixels appartenant à la même classe). La justification théorique de notre modèle repose sur les travaux effectués dans le cadre des problèmes de transitions de phases en mécanique. L'algorithme que nous proposons est relativement rapide et facile à mettre en oeuvre. Nous comparons les résultats obtenus sur des images synthétiques et satellitaires avec ceux produits par un modèle stochastique avec régularisation de Potts. |
Abstract :
Herein, we present a variational model devoted to image classification coupled with an edge-preserving regularization process. The discrete nature of classification (i.e. to attribute a label to each pixel) has ledto the development of many probabilistic image classification models, but rarely to variational ones. In the last decade, the variational approach has proven its efficiency in the field of edge-preserving restoration. In this paper we add a classification capability which contributes to provide images compound of homogeneous regions with regularized boundaries, a region being defined as a set of pixels belonging to the same class. The soundness of our model is based on the works developed on the phase transitions theory in mechanics. The proposed algorithm is fast, easy to implement, and efficient. We compare our results on both synthetic and satellite images with the ones obtained by a stochastic model using a Potts regularization. |
|
3 - Mise en correspondance et recalage de graphes : application aux réseaux routiers extraits d'un couple carte/image.
C. Hivernat and X. Descombes and S. Randriamasy and J. Zerubia. Research Report 3529, Inria, October 1998.
Keywords : Markov Fields, Road network, Graph matching.
@TECHREPORT{hiv98,
|
| author |
= |
{Hivernat, C. and Descombes, X. and Randriamasy, S. and Zerubia, J.}, |
| title |
= |
{Mise en correspondance et recalage de graphes : application aux réseaux routiers extraits d'un couple carte/image}, |
| year |
= |
{1998}, |
| month |
= |
{October}, |
| institution |
= |
{Inria}, |
| type |
= |
{Research Report}, |
| number |
= |
{3529}, |
| url |
= |
{http://www.inria.fr/rrrt/rr-3529.html}, |
| pdf |
= |
{ftp://ftp.inria.fr/INRIA/publication/publi-pdf/RR/RR-3529.pdf}, |
| ps |
= |
{ftp://ftp.inria.fr/INRIA/publication/publi-psgz/RR/RR-3529.ps.gz}, |
| keyword |
= |
{Markov Fields, Road network, Graph matching} |
| } |
Résumé :
Nous considérons le problème de la mise en correspondance du réseau routier extrait d'une image SPOT avec celui fourni par une base de données cartographi- que. Cette mise en correspondance comprend deux étapes principales fondées sur des modélisations markoviennes. Dans la première étape, les pixels de l'image sont appariés aux segments cartographiques. Le résultat de cette étape permet de découper le réseau obtenu sur l'image sous forme de chaînes. Ces chaînes sont ensuite mises en correspondance avec les segments cartographiques. Pour finir, une étape de qualification des résultats permet de fournir les primitives fiables afin d'affiner le recalage initial. En bouclant l'algorithme sur la mise en correspondance nous obtenons un processus itératif permettant d'améliorer à la fois le recalage et la mise en correspondance. La qualification automatique des résultats est également une aide à l'interprétation pour la mise à jour cartographique. |
Abstract :
We consider herein the matching problem between the road network extracted from a SPOT image and the roads contained in a cartographic database. This matching consists of two main steps based on a Markovian modelling. During the first step, the image road pixels are associated to the map segments. the derived result allows us to split the image network into chains. These chains are matched with the map segments. Finally, an automatic validation procedure provides matched chains/segments which are used to improve the initial registration. An iterative scheme is obtained by performin- g a new matching. The automatic result validation is also helpful for map updating. |
|
4 - Estimation d'hyperparamètres pour la restauration d'images satellitaires par une méthode MCMCML.
A. Jalobeanu and L. Blanc-Féraud and J. Zerubia. Research Report 3469, Inria, August 1998.
Keywords : Markov Fields, Regularization, Variational methods, Likelihood maximum.
@TECHREPORT{jaloRR98,
|
| author |
= |
{Jalobeanu, A. and Blanc-Féraud, L. and Zerubia, J.}, |
| title |
= |
{Estimation d'hyperparamètres pour la restauration d'images satellitaires par une méthode MCMCML}, |
| year |
= |
{1998}, |
| month |
= |
{August}, |
| institution |
= |
{Inria}, |
| type |
= |
{Research Report}, |
| number |
= |
{3469}, |
| url |
= |
{http://www.inria.fr/rrrt/rr-3469.html}, |
| pdf |
= |
{ftp://ftp.inria.fr/INRIA/publication/publi-pdf/RR/RR-3469.pdf}, |
| ps |
= |
{ftp://ftp.inria.fr/INRIA/publication/publi-psgz/RR/RR-3469.ps.gz}, |
| keyword |
= |
{Markov Fields, Regularization, Variational methods, Likelihood maximum} |
| } |
Résumé :
Le problème que nous abordons ici est la déconvolution d'images satellitaires, qui sont dégradées par l'optique et l'électronique utilisées pour leur acquisition. Les dégradations sont connues : les images sont convoluées par un opérateur H, et la variance du bruit N additif, blanc et gaussien, est connue. Nous utilisons un modèle de régularisation introduisant une fonction de potentiel phi, qui interdit l'amplification du bruit lors de la restauration tout en préservant les discontinuités. Ce modèle admet deux hyperparamètres lambda et delta. Nous nous intéressons ici à l'estimation des hyperparamètres optimaux afin d'effectuer la déconvolution de manière automatique. Nous proposons pour cela d'utiliser l'estimateur du maximum de vraisemblance appliqué à l'image observée. Cet estimateur constitue le critère que nous allons optimiser. Pour évaluer ses dérivées, nous devons estimer des espérances calculées sur des échantillon- s, tenant compte des données observées et de l'a priori imposé. Cette probabilité faisant intervenir l'opérateur de convolution, il est très difficile d'utiliser un échantillonneur classique. Nous avons développé un algorithme de type Geman-Yang modifié, utilisant une variable auxiliaire, ainsi qu'une transformée en cosinus. Nous présentons à cette occasion un nouvel algorithme de déconvolution, rapide, qui est dérivé de cette méthode d'échantillonnage. Nous proposons un algorithme "MCMCML" permettant d'effectuer simultanément l'estimation des hyperparamètres lambda et delta et la restauration de l'image dégradée. Une étude des échantillonneurs (y compris ceux de Gibbs et Metropolis), portant sur la vitesse de convergence et les difficultés de calcul liées à l'attache aux données, a également été réalisée. |
Abstract :
This report deals with satellite image restoration. These images are corrupted by an optical blur and electronic noise, due to the physics of the sensors. The degradation model is known : blurring is modeled by convolution, with a linear operator H, and the noise is supposed to be additive, white and Gaussian, with a known variance. The recovery problem is ill-posed and therefore must be regularized. We use a regularization model which introduces a phi function, which avoids noise amplification while preserving image discontinuities (ie. edges) of the restored image. This model exhibits two hyperparameters (lambda and delta). Our goal is to estimate the optimal parameters in order to reconstruct images automatically. Herein, we propose to use the Maximum Likelihood estimator, applied to the observed image. To optimize this criterion, we must estimate expectations by sampling (samples are extracted from a Markov chain) to evaluate its derivatives. These samples are images whose probability takes into account the convolution operator. Thus, it is very difficult to obtain them directly by using a standard sampler. We have developped a modified Geman-Yang algorithm, using an auxiliary variable and a cosine transform. We also present a new reconstruc- tion method based on this sampling algorithm. We detail the MCMCML algorithm which ables to simultaneously estimate lambda and delta parameters, and to reconstruct the corrupted image. An experimental study of samplers (including Gibbs and Metropolis algorithms), with respect to the rate of convergence and the difficulties of dependent data sampling, is also presented in this report. |
|
5 - Extraction des zones urbaines fondée sur une analyse de la texture par modélisation markovienne.
A. Lorette and X. Descombes and J. Zerubia. Research Report 3423, Inria, May 1998.
Keywords : Texture, Markov Fields, Urban areas, Entropy.
@TECHREPORT{loretteRR98,
|
| author |
= |
{Lorette, A. and Descombes, X. and Zerubia, J.}, |
| title |
= |
{Extraction des zones urbaines fondée sur une analyse de la texture par modélisation markovienne}, |
| year |
= |
{1998}, |
| month |
= |
{May}, |
| institution |
= |
{Inria}, |
| type |
= |
{Research Report}, |
| number |
= |
{3423}, |
| url |
= |
{http://hal.inria.fr/inria-00073267}, |
| pdf |
= |
{http://hal.inria.fr/docs/00/07/32/67/PDF/RR-3423.pdf}, |
| ps |
= |
{http://hal.inria.fr/docs/00/07/32/67/PS/RR-3423.ps}, |
| keyword |
= |
{Texture, Markov Fields, Urban areas, Entropy} |
| } |
Résumé :
Pour délimiter un masque urbain précis à partir d'une image satellitaire la seule information du niveau de gris est insuffisante. Laplupart des méthodes font donc appel à une analyse de la texture de l'image. Nous nous sommes placés dans ce cadre. Dans une première étape, nous avons défini un nouveau paramètre de texture à partir d'un modèle markovien gaussien. Nous obtenons ce nouveau paramètre en calculant la variance conditionnelle de l'image dans huit directions. Ainsi, nous éliminons la mauvaise classification d'objets ayant une orientation privilégiée tels que les vignes et les serres par exemple. Dans une seconde étape, nous proposons un algorithme de emphfuzzy Cmeans modifié incluant un terme d'entropie et pour lequel le nombre de classes n'est pas fixé a priori. Cet algorithme nous permet d'obtenir une première classification de l'image. Enfin, nous régularisons l'image ainsi obtenue grâce à une modélisation par champs de Markov. Des résultats obtenus sur des simulations d'images SPOT5 fournies par le CNES sont présentés. |
Abstract :
Urban areas cannot be extracted from satellite images through only grey level information. Hence most methods analyze the texture of the image to discriminate between urban areas and non urban areas. We define a new texture parameter derived from a Markovian Gaussian model. This new parameter takes into account the variance of the image in eight directions- . Consequently it copes with the misclassification of objects with a privileged orientation like vineyards or greenhouses for instance. Afterwards we develop a modified fuzzy Cmeans algorithm including an entropy term. The advantage of such an algorithm is that the number of classes does not need to be known a priori. By applying this modified fuzzy Cmeans algorithm on the parameter image we obtain a first classification. Finally we regularize the segmented image by using a Markov random field modelling. Some results on SPOT5 simulated images are presented. These images are provided by the CNES (French Space Agency). |
|
6 - An elementary proof of the equivalence between 2D and 3D classical snakes and geodesic active contours.
G. Aubert and L. Blanc-Féraud. Research Report 3340, Inria, January 1998.
Keywords : Geodesic active contours, Partial differential equation.
@TECHREPORT{lbf98i,
|
| author |
= |
{Aubert, G. and Blanc-Féraud, L.}, |
| title |
= |
{An elementary proof of the equivalence between 2D and 3D classical snakes and geodesic active contours}, |
| year |
= |
{1998}, |
| month |
= |
{January}, |
| institution |
= |
{Inria}, |
| type |
= |
{Research Report}, |
| number |
= |
{3340}, |
| url |
= |
{http://www.inria.fr/rrrt/rr-3340.html}, |
| pdf |
= |
{ftp://ftp.inria.fr/INRIA/publication/publi-pdf/RR/RR-3340.pdf}, |
| ps |
= |
{ftp://ftp.inria.fr/INRIA/publication/publi-psgz/RR/RR-3340.ps.gz}, |
| keyword |
= |
{Geodesic active contours, Partial differential equation} |
| } |
Résumé :
Les équations aux dérivées partielles (EDP) définissant l'évolution de courbe plane permettent, avec une implantation par ligne de niveau, un changement de topologie par rapport à la courbe initiale et sont de ce fait un outil puissant pour la segmentation d'objet dans une image. Récemment, Caselles et al. ont montrés l'équivalence entre les modèles de contours actifs classiques de Kass et al. et de contours actifs géodésiques qui définissent une EDP particulière d'évolution de courbe. Cependant la preuve proposée par Caselles n'est valable que pour la segmentation d'objet d'une image 2D avec des courbes 1D et fait appel à des concepts de la théorie hamiltonienne, sans aucun sens physique pour les contours actifs. Ce papier propose une preuve utilisant uniquement des calculs élémentaires d'analyse mathématique, valables aussi pour la segmentation d'image 3D à l'aide de surface. |
Abstract :
Recently, Caselles et al. have shown in the equivalence between a classical snake problem of Kass et al. and a geodesic active contour model. The PDE derived from the geodesic problem gives an evolution equation for active contours which is very powerfull for image segmentation since changes of topology are allowed using the level set implementation. However in Caselles' paper the equivalence with classical snake is only shown for 2D images with 1D curves, by using concepts of Hamiltonian theory which have no meanings for active contours. This paper propose a proof using only elementary calculus of mathematical analysis. This proof is also valid in the 3D case for active surfaces. |
|
7 - Estimation-segmentation du flot optique et contours actifs.
R. Ciampini and L. Blanc-Féraud and M. Barlaud and E. Salerno. Research Report 98-11, Laboratoire I3S, 1998.
@TECHREPORT{lbf98j,
|
| author |
= |
{Ciampini, R. and Blanc-Féraud, L. and Barlaud, M. and Salerno, E.}, |
| title |
= |
{Estimation-segmentation du flot optique et contours actifs}, |
| year |
= |
{1998}, |
| institution |
= |
{Laboratoire I3S}, |
| type |
= |
{Research Report}, |
| number |
= |
{98-11}, |
| keyword |
= |
{} |
| } |
|
top of the page
These pages were generated by 
|
get the bibtex
