@Conference{Beau23ep,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Edge Preserving Bi-Level Set SAR Image Filter},
booktitle = {Advanced SAR Workshop 2023, Canadian Space Agency},
volume = {},
publisher = {},
url = {http://www.asc-csa.gc.ca},
isbn = {},
doi = {},
address = {Montreal (Saint-Hubert)},
pages = {},
year = {2023},
month = {Nov. 27-30},
abstract = {},
mypdf = {6},
slide = {https://BeaulieuJM.ca/slide/slideBeau23ep.pdf},
keywords = {}
}

@book{Beau23HS,
author = {Beaulieu, Jean-Marie},
title = {Hierarchical Image Segmentation by Stepwise Optimization: New edition of 1984 Thesis},
editor = {Jean-Marie Beaulieu},
url = {https://BeaulieuJM.ca/pupli/Beau23HS},
isbn = {978-1-7388812-0-8},
doi = {},
pages = {145},
publisher = {Jean-Marie Beaulieu},
address = {Quebec (Canada)},
year = {2023},
abstract = {The survey of image segmentation considers four different approaches: pixel classification, pixel linking and region growing, hierarchical segmentation, and segmentation optimization. A new Hierarchical Stepwise Optimization (HSO) algorithm is proposed, which combines these last two approaches. The algorithm employs a sequence of optimization processes to produce a hierarchical segmentation. Starting with an initial image partition, two segments are then merged at each iteration by using an optimization process to select the segment pair that minimizes a “stepwise criterion.” The algorithm is then employed for piecewise image approximation where the stepwise criterion is derived from the global criterion, the overall approximation error. The stepwise criterion is then related to statistical hypothesis testing, and it is shown how the probability of error can be minimized in a stepwise fashion. It is also shown experimentally how convenient stopping points in the hierarchy can be found from the criterion values. Different criteria are tested on Landsat and SAR imagery.},
mypdf = {8},
keywords = {Hierarchical segmentation; Similarity measures; Clustering}
}

@Conference{Beau23pa,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Préservation des Arêtes dans le Filtrage des Images SAR avec les ensembles à deux niveaux},
booktitle = {Congr{\`e}s 2023 de l'Association Qu{\'e}b{\'e}coise de T{\'e}l{\'e}d{\'e}tection},
volume = {},
publisher = {},
url = {http://laqt.ca},
isbn = {},
doi = {},
address = {Université du Québec à Trois-Rivières},
pages = {},
year = {2023},
month = {23-25 oct.},
abstract = {},
mypdf = {6},
slide = {https://BeaulieuJM.ca/slide/slideBeau23pa.pdf},
keywords = {}
}

@Conference{Beau21sh,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Segmentation hiérarchique avec Julia},
booktitle = {Congrès 2021 de L’Association Québécoise de Télédétection},
volume = {},
publisher = {},
url = {https://event.fourwaves.com/fr/aqt2021/pages},
isbn = {},
doi = {},
address = {},
pages = {},
year = {2021},
month = {2-3 juin},
abstract = {},
mypdf = {6},
keywords = {},
slide = {https://BeaulieuJM.ca/slide/slideBeau21sh.pdf},
keywords = {}
}

@Conference{Bea2019b,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Efficient Hierarchical Clustering for Polsar Image Analysis},
booktitle = {Advanced SAR Workshop 2019, Canadian Space Agency},
volume = {},
publisher = {},
url = {http://www.asc-csa.gc.ca/eng/events/2019/asar-2019-workshop-on-synthetic-aperture-radar.asp},
isbn = {},
doi = {},
address = {Montreal (Saint-Hubert)},
pages = {},
year = {2019},
month = {Oct. 1-3},
abstract = {},
mypdf = {6},
slide = {https://BeaulieuJM.ca/slide/slideBea2019b.pdf},
keywords = {}
}

@Conference{Bea2019a,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Contr{\^o}le du Voisinage pour un Regroupement Hi{\'e}rarchique Efficace},
booktitle = {Colloque AQT/RHQ 2019: La t{\'e}l{\'e}d{\'e}tection et l'eau dans tous leurs {\'e}tats},
volume = {},
publisher = {},
url = {https://aqtrhq2019.sciencesconf.org},
isbn = {},
doi = {},
address = {Campus de l'U. Bishop's, Sherbrooke},
pages = {1},
year = {2019},
month = {15-17 mai},
abstract = {},
mypdf = {6},
slide = {https://BeaulieuJM.ca/slide/slideBea2019a.pdf},
keywords = {}
}

@Conference{Bea2017a,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Mean-Shif Polsar Image Denoising with Position Tensor},
booktitle = {Earth Observation Summit 2017},
volume = {},
publisher = {},
url = {https://crss-sct.ca/conferences/csrs2017},
isbn = {},
doi = {},
address = {Montreal},
pages = {1},
year = {2017},
month = {June 20-22},
abstract = {},
mypdf = {6},
slide = {https://BeaulieuJM.ca/slide/slideBea2017a.pdf},
keywords = {},
openpdf = {https://sommetot2017-eosummit2017.exordo.com/files/papers/60/initial_draft/SOT-2017-Polarimetry-Beaulieu.pdf},
openid = {ExOrdo}
}

@Conference{Bea2015b,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Filtrage d'Image Polsar par Mean-Shift avec Tenseur
/ Tensor Based Mean-Shift Polsar Image Enhancement},
booktitle = {Advanced SAR Workshop 2015, Canadian Space Agency},
volume = {},
publisher = {},
url = {http://www.asc-csa.gc.ca},
isbn = {},
doi = {},
address = {Montreal (Saint-Hubert)},
pages = {},
year = {2015},
month = {Oct. 20-22},
abstract = {},
mypdf = {6},
slide = {https://BeaulieuJM.ca/slide/slideBea2015a.pdf},
keywords = {}
}

@Conference{Bea2015a,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Filtrage d'Image Polsar par Mean-Shift avec Tenseur},
booktitle = {XVIe Congr{\`e}s de l'Association Qu{\'e}b{\'e}coise de T{\'e}l{\'e}d{\'e}tection},
volume = {},
publisher = {},
url = {http://laqt.org},
isbn = {},
doi = {},
address = {INRS, Quebec},
pages = {},
year = {2015},
month = {28-30 oct.},
abstract = {},
mypdf = {6},
slide = {https://BeaulieuJM.ca/slide/slideBea2015a.pdf},
keywords = {}
}

@Conference{Bea2014a,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Tensor Based Mean-Shift Polsar Image Enhancement},
booktitle = {IEEE International Geoscience and Remote Sensing Symposium},
volume = {IGARSS 2014},
publisher = {},
url = {http://ieeexplore.ieee.org/xpl/login.jsp?tp=\&arnumber=6947503},
isbn = {},
doi = {10.1109/IGARSS.2014.6947503},
address = {Quebec City, QC},
pages = {4544-4547},
year = {2014},
month = {July 13-18,},
abstract = {The mean-shift approach uses a local estimation of the pdf and moves every data points toward the modes. The direction is calculated from the mean value of surrounding points weighted by a Gaussian kernel. An advantage of the technique is that both radiometric and spatial information could be used in the weighted mean calculation. For polarimetric SAR images, we use likelihood ratio as radiometric similarity or distance measure. The spatial distance between pixels is also used with a Gaussian weight. Contours are well preserved because pixels on one side are dissimilar to pixels on the other side. To improve contour preservation, we examine how the tensor of pixel position can be integrated into the weight calculation. The tensor is calculated from weighted pixel position inside a window. Good PolSAR image smoothing is obtained.},
mypdf = {11},
slide = {https://BeaulieuJM.ca/slide/slideBea2014a.pdf},
keywords = {}
}

@phdthesis{ElM2012,
author = {El Mabrouk, Abdelhai},
title = {Segmentation, Regroupement et Classification pour l'Analyse d'Image Polarim{\'e}trique Radar},
school = {Universit{\'e} Laval},
dept = {D{\'e}partement d'Informatique et de G{\'e}nie Logiciel},
degree = {Msc},
thesis = {Master},
address = {Universit{\'e} Laval, Qu{\'e}bec, Qu{\'e}bec, Canada},
pages = {},
year = {2012},
month = {},
publisher = {Universit{\'e} Laval},
url = {https://library-archives.canada.ca/eng/services/services-libraries/theses/Pages/item.aspx?idNumber=803363632},
isbn = {},
doi = {},
wdown = {},
mypdf = {13},
openpdf = {https://www.collectionscanada.ca/obj/thesescanada/vol2/QQLA/TC-QQLA-28912.pdf},
openid = {Library and Archives Canada},
abstract = {},
keywords = {}
}

@Conference{Bea2011b,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
editor = {},
title = {Mean-Shift Clustering and Hierarchical Segmentation for Polsar Image Analysis},
booktitle = {Advanced SAR Workshop 2011, Canadian Space Agency},
volume = {},
publisher = {},
url = {http://www.asc-csa.gc.ca},
isbn = {},
doi = {},
address = {Montreal (Saint-Hubert)},
pages = {6},
year = {2011},
month = {June},
abstract = {Image segmentation and unsupervised classification are difficult problems. We propose to combine both. A clustering process is applied over segment mean values. Only large segments are considered. The clustering is composed of a mean-shift step and a hierarchical clustering step. The hierarchical grouping is based upon a powerful segmentation technique previously developed [1]. The approach is applied on a 9-look polarimetric SAR image. Textured and non- textured image regions are considered. The K and Wishart distributions are used respectively. The unsupervised classification results can be very useful for image analysis and further supervised classification. The obtained region groups constitute an important simplification of the image.},
mypdf = {7},
keywords = {}
}

@Conference{ElM2011,
author = {El Mabrouk, Abdelhai and Beaulieu, Jean-Marie},
editor = {},
title = {Segmentation hi{\'e}rarchique par optimisation s{\'e}quentielle pour la classification H/A/Alpha d'image polarim{\'e}trique {SAR}},
booktitle = {32e Symposium canadien sur la t{\'e}l{\'e}d{\'e}tection et 14e Congr{\`e}s de l'AQT},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {Université Bishop’s, Sherbrooke (Québec)},
pages = {1114},
year = {2011},
month = {13-16 juin},
abstract = {Dans une image polarimétrique SAR multi-vue, chaque pixel est représenté par une matrice Hermitienne 3x3, soit la matrice de cohérence. À partir des valeurs propres et des vecteurs propres de cette matrice, on peut calculer l'entropie H, l'anisotropie A et l'angle α. En définissant des intervalles (seuils) pour chaque paramètre, on obtient 16 zones définissant 16 classes de rétrodiffusion. Chaque pixel sera assigné à une de ces classes. On effectue préalablement un filtrage pour réduire le bruit dans la classification. Il y a un aspect arbitraire dans le choix des seuils. Il faut chercher une approche qui s'adapte aux données. Par exemple, la technique de regroupement des K-centres a été utilisée. Nous proposons d'utiliser la segmentation hiérarchique et le regroupement hiérarchique pour faire un premier regroupement des pixels. L'image est ainsi simplifiée tout en préservant l'information spatiale. Ceci remplace le filtrage et permet d'obtenir une meilleure image de classification.},
mypdf = {7},
keywords = {}
}

@Conference{Bea2011a,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
editor = {},
title = {Segmentation/Classification des Images Polsar par Regroupement Hierarchique et Mean-Shift},
booktitle = {32e Symposium Canadien sur la T{\'e}l{\'e}d{\'e}tection et 14e Congr{\`e}s de l'AQT, Sherbrooke},
volume = {},
publisher = {},
url = {https://crss-sct.ca},
isbn = {},
doi = {},
address = {Campus de l'U. Bishop's, Sherbrooke},
pages = {1-7},
year = {2011},
month = {13-16 juin},
abstract = {Nous avons d{\'e}velopp{\'e} une approche de segmentation hi{\'e}rarchique performante pour les images polarim{\'e}triques SAR. Cependant, la segmentation et la classification non supervis{\'e}e demeurent des probl{\`e}mes difficiles. Dans cet article, nous proposons de combiner les deux.
En t{\'e}l{\'e}d{\'e}tection, la t{\^a}che principale est l'interpr{\'e}tation de l'image. Nous devons d{\'e}velopper des outils qui facilitent l'accomplissement de cette t{\^a}che complexe. Ceci est l'objectif des techniques automatiques de classification, qu'on nomme techniques de regroupement (clustering). Nous examinerons les relations entre les techniques it{\'e}ratives de regroupement, le regroupement hi{\'e}rarchique et la segmentation de l'image. Nous regarderons comment nous pouvons passer d'une {\`a} l'autre.},
mypdf = {7},
keywords = {}
}

@Conference{Bea2010a,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
editor = {},
title = {Mean-shift and Hierarchical Clustering for Textured Polarimetric {SAR} Image Segmentation/Classification},
booktitle = {IEEE International Geoscience and Remote Sensing Symposium},
volume = {IGARSS 2010},
publisher = {},
url = {http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=5653919},
isbn = {978-1-4244-9565-8},
doi = {10.1109/IGARSS.2010.5653919},
address = {Honolulu, HI},
pages = {2519-2522},
year = {2010},
month = {July 25-30,},
abstract = {Image segmentation and unsupervised classification are difficult problems. We propose to combine both. A clustering process is applied over segment mean values. Only large segments are considered. The clustering is composed of a mean-shift step and a hierarchical clustering step. The hierarchical grouping is based upon a powerful segmentation technique previously developed. The approach is applied on a 9-look polarimetric SAR image. Textured and non-textured image regions are considered. The K and Wishart distributions are used respectively. The unsupervised classification results can be very useful for image analysis and further supervised classification. The obtained region groups constitute an important simplification of the image.},
mypdf = {11},
slide = {https://BeaulieuJM.ca/slide/slideBea2010a.pdf},
keywords = {9-look polarimetric SAR image; hierarchical clustering; hierarchical grouping; image analysis; image classification; image segmentation; image texture; K distribution; mean-shift step; nontextured image region; pattern clustering; radar imaging; radar polarimetry; segment mean value; statistical distributions; synthetic aperture radar; textured polarimetric SAR image; unsupervised classification; Wishart distribution}
}

@Conference{Bom2009a,
author = {Bombrun, Lionel and Beaulieu, Jean-Marie and Vasile, G and Ovarlez, J P and Pascal, F and Gay, M},
editor = {},
title = {Hierarchical Segmentation of Polarimetric {SAR} Images using Heterogeneous Clutter Models},
booktitle = {IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009},
volume = {III},
publisher = {IEEE},
url = {https://ieeexplore.ieee.org/abstract/document/5418271},
isbn = {978-1-4244-3394-0},
doi = {10.1109/IGARSS.2009.5418271},
address = {Cape Town, South Africa},
pages = {5-8},
year = {2009},
month = {12-17 July},
abstract = {In this paper, heterogeneous clutter models are introduced to describe Polarimetric Synthetic Aperture Radar (PolSAR) data. Based on the Spherically Invariant Random Vectors (SIRV) estimation scheme, the scalar texture parameter and the normalized covariance matrix are extracted. If the texture parameter is modeled by a Fisher PDF, the observed target scattering vector follows a KummerU PDF. Then, this PDF is implemented in a hierarchical segmentation algorithm. Segmentation results are shown on high resolution PolSAR data at L and X band.},
mypdf = {13},
keywords = {Backscatter; Clutter; Covariance matrix; Data mining; Fisher PDF; geophysical image processing; geophysical techniques; heterogeneous clutter models; hierarchical image segmentation; hierarchical segmentation algorithm; image segmentation; image texture; KummerU PDF; L band high resolution PolSAR data; L-band; Maximum likelihood estimation; normalized covariance matrix; Parameter estimation; polarimetric SAR images; polarimetric synthetic aperture radar data; PolSAR data; radar clutter; radar polarimetry; Radar scattering; remote sensing by radar; scalar texture parameter; Segmentation; Spherically Invariant Random Vectors; spherically invariant random vectors estimation scheme; synthetic aperture radar; target scattering vector; Testing; X band high resolution PolSAR data},
openpdf = {https://hal.archives-ouvertes.fr/hal-00398923/},
openid = {HAL archives-ouvertes}
}

@Conference{Bom2008b,
author = {Bombrun, Lionel and Beaulieu, Jean-Marie},
editor = {},
title = {Segmentation of Polarimetric {SAR} Data based on the Fisher Distribution for Texture Modeling},
booktitle = {IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2008},
volume = {V},
publisher = {IEEE},
url = {https://ieeexplore.ieee.org/abstract/document/4780100},
isbn = {978-1-4244-2807-6},
doi = {10.1109/IGARSS.2008.4780100},
address = {Boston, MA, USA},
pages = {350-353},
year = {2008},
month = {July 7-11},
abstract = {The Polarimetric Synthetic Aperture Radar (PolSAR) covariance matrix is generally modeled by a complex Wishart distribution. For textured scenes, the product model is used and the texture component is often modeled by a Gamma distribution. In this paper, authors propose to use the Fisher distribution for texture modeling. From a Fisher distributed texture component, we derive the distribution of the complex covariance matrix and we propose to implement the KummerU distribution in a hierarchical segmentation and a hierarchical clustering algorithm. Segmentation and classification results are shown on synthetic images and on ESAR L-band PolSAR data over the Oberpfaffenhofen test-site.},
mypdf = {13},
keywords = {Classification; Clustering algorithms; complex Wishart distribution; covariance matrices; covariance matrix; Electromagnetic scattering; ESAR L-band PolSAR data; Fisher distribution; Gamma distribution; geophysical techniques; geophysics computing; hierarchical clustering algorithm; hierarchical segmentation; image classification; image segmentation; image texture; KummerU; KummerU distribution; L-band; Layout; Oberpfaffenhofen test-site; Polarimetric SAR images; Polarimetric Synthetic Aperture Radar data; Polarization; radar polarimetry; Radar scattering; Receiving antennas; remote sensing by radar; Segmentation; Speckle; synthetic aperture radar; Texture; texture component; texture modeling},
openpdf = {https://hal.archives-ouvertes.fr/hal-00369374/},
openid = {HAL archives-ouvertes}
}

@Article{Bom2008a,
author = {Bombrun, Lionel and Beaulieu, Jean-Marie},
title = {Fisher Distribution for Texture Modeling of Polarimetric {SAR} Data},
journal = {IEEE Geoscience and Remote Sensing Letters},
volume = {5},
number = {3},
pages = {512--516},
year = {2008},
month = {July},
abstract = {The multilook polarimetric synthetic aperture radar (PolSAR) covariance matrix is generally modeled by a complex Wishart distribution. For textured areas, the product model is used, and the texture component is modeled by a Gamma distribution. In many cases, the assumption of Gamma-distributed texture is not appropriate. The Fisher distribution does not have this limitation and can represent a large set of texture distributions. As an example, we examine its advantage for an urban area. From a Fisher-distributed texture component, we derive the distribution of the complex covariance matrix for multilook PolSAR data. The obtained distribution is expressed in terms of the KummerU confluent hypergeometric function of the second kind. Those distributions are related to the Mellin transform and second-kind statistics (Log-statistics). The new KummerU-based distribution should provide in many cases a better representation of textured areas than the classic K distribution. Finally, we show that the new model can discriminate regions with different texture distribution in a segmentation experiment with synthetic textured PolSAR images.},
publisher = {IEEE},
url = {https://ieeexplore.ieee.org/abstract/document/4554026},
issn = {1545-598X},
doi = {10.1109/LGRS.2008.923262},
mypdf = {13},
address = {},
keywords = {Classification; complex covariance matrix distribution; covariance matrices; Fisher distributed texture component; Fisher distribution; geophysical signal processing; geophysical techniques; image segmentation; image texture; KummerU; log statistics; Mellin transform; multilook PolSAR covariance matrix; polarimetric SAR data texture modeling; polarimetric synthetic aperture radar; polarimetric synthetic aperture radar (PolSAR) images; radar polarimetry; radar signal processing; remote sensing by radar; second kind KummerU confluent hypergeometric function; second kind statistics; segmentation; statistical distributions; synthetic aperture radar; synthetic textured PolSAR image segmentation; texture; texture distributions; urban area},
openpdf = {https://hal.archives-ouvertes.fr/hal-00350055/},
openid = {HAL archives-ouvertes}
}

@Conference{Bea2008a,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
editor = {},
title = {Classification of Polarimetric {SAR} Images using Radiometric and Texture Information},
booktitle = {IEEE International Geoscience and Remote Sensing Symposium IGARSS 2008},
volume = {IV},
publisher = {IEEE},
url = {https://ieeexplore.ieee.org/document/4779648},
isbn = {978-1-4244-2807-6},
doi = {10.1109/IGARSS.2008.4779648},
address = {Boston, MA, USA},
pages = {29-32},
year = {2008},
month = {July 7-11},
abstract = {},
mypdf = {11},
slide = {https://BeaulieuJM.ca/slide/slideBea2008a.pdf},
keywords = {classification; classification map; clustering; Clustering algorithms; clustering process; Covariance matrix; geophysical techniques; hierarchical clustering; hierarchical segmentation; image classification; image segmentation; Iterative algorithms; K distribution; mean shift clustering; mean-shift; Merging; Partitioning algorithms; pattern clustering; Pixel; polarimetric SAR image; Probability; radar polarimetry; radiometry; Remote sensing; remote sensing by radar; scalar texture component; synthetic aperture radar; texture; texture information; Wishart distribution}
}

@Conference{Bea2008b,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
editor = {},
title = {Aller--Retour Segmentation/Classification des Images Polarim{\'e}triques {SAR}},
booktitle = {13e Congr{\`e}s de l'Association Qu{\'e}b{\'e}coise de T{\'e}l{\'e}d{\'e}tection},
volume = {},
publisher = {},
url = {http://laqt.org},
isbn = {},
doi = {},
address = {Trois-Rivi{\`e}res, Canada},
pages = {1-6},
year = {2008},
month = {30 avril - 2 mai},
abstract = {Nous avons d{\'e}velopp{\'e} une technique efficace de segmentation hi{\'e}rarchique et l'avons appliqu{\'e} aux images polarim{\'e}triques SAR. La segmentation et la classification non-supervis{\'e}e d'image sont des probl{\`e}mes difficiles. On peut simplifier le probl{\`e}me en acceptant un nombre {\'e}lev{\'e} de seg- ments (r{\'e}gions) ou de classes. Il est reconnu que la classification bas{\'e}e sur la valeur des seg- ments est moins affect{\'e}e par le bruit que la classification bas{\'e}e sur la valeur des pixels. Nous pouvons utilisez une partition avec beaucoup de r{\'e}gions (sur-segmentation) simplifiant ainsi la t{\^a}che de la segmentation. Cependant, la classification non-supervis{\'e}e de segment demeure un probl{\`e}me difficile. Pour simplifier, nous utilisons seulement un sous ensemble des segments et nous produisons une classification avec beaucoup de classes. Chaque segment de la sur- segmentation est alors assign{\'e} {\`a} une des nombreuses classes. Nous pouvons utiliser cette infor- mation de classe pour poursuivre la segmentation en fusionnant les r{\'e}gions et r{\'e}duire {\`a} une valeur convenable le nombre de r{\'e}gions.},
mypdf = {7},
slide = {https://BeaulieuJM.ca/slide/slideBea2008b.pdf},
keywords = {}
}

@Conference{Bea2006a,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Pseudo-Convex Contour Criterion for Hierarchical Segmentation of {SAR} Images},
booktitle = {The 3rd Canadian Conference on Computer and Robot Vision},
volume = {},
publisher = {IEEE},
url = {http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=1640384},
isbn = {0-7695-2542-3},
doi = {10.1109/CRV.2006.58},
address = {Laval University, Canada},
pages = {29-29},
year = {2006},
month = {June 07-09,},
abstract = {The hierarchical segmentation of SAR (Synthetic Aperture Radar) images is greatly complicated by the presence of coherent speckle. We are exploring the utilization of spatial constraints and contour shapes in order to improve the segmentation results. With standard merging criterion, the high noise level of SAR images results in the production of regions that have variable mean and variance values and irregular shapes. If the first segments are not correctly delimited then the following steps will merge segments from different fields. In examining the evolution of the initial segments, we see that the merging should take into account spatial aspects. Particularly, the segment contours should have good shapes. In this paper, we examine how the pseudo-convex envelope of a region can be used to evaluate the region contour. We present a pseudo-convex measure adapted to the geometry of image lattice. We show how the pseudo-convex envelope can be calculated. We present measures comparing contour shapes and using the perimeter, the area and the boundary length of segments. We use a hierarchical segmentation algorithm based upon stepwise optimization. A stepwise merging criterion is derived from the multiplicative speckle noise model. The shape measures are combined with the merging criterion in order to guide correctly the segment merging process. The new criterion produces good segmentation of SAR images. This is illustrated by synthetic and real image results.},
mypdf = {11},
keywords = {Area measurement; Geometry; Image segmentation; Lattices; Merging; Noise level; Production; Shape measurement; Speckle; Synthetic aperture radar}
}

@Conference{Bea2005a,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {{\'E}valuation des Mesures de Dissimilarit{\'e} entre R{\'e}gions dans les Images {SAR}},
booktitle = {12{\`e}me Congr{\`e}s de l'Association Qu{\'e}b{\'e}coise de T{\'e}l{\'e}d{\'e}tection},
volume = {},
publisher = {},
url = {http://laqt.org},
isbn = {},
doi = {},
address = {Chicoutimi (Qu{\'e}bec) Canada},
pages = {1-8},
year = {2005},
month = {10-12 mai},
abstract = {Nous montrons comment les courbes de la probabilit{\'e} de d{\'e}tection vs la probabilit{\'e} de fausse alarme (courbes ROC) peuvent {\^e}tre utilis{\'e}es pour comparer diff{\'e}rentes mesures ou crit{\`e}res de d{\'e}tection d'ar{\^e}tes ou de segmentation d'images radar. Les 3 crit{\`e}res {\'e}tudi{\'e}s sont 1) le logarithme du rapport de vraisemblance, 2) le rapport des moyennes et 3) une adaptation du crit{\`e}re de Ward pour les images SAR. Les 3 crit{\`e}res donnent des r{\'e}sultats identiques lorsque nous utilisons 2 r{\'e}- gions de m{\^e}me taille. Lorsque les r{\'e}gions sont petites et ont une diff{\'e}rence de taille importante, nous obtenons des courbes diff{\'e}rentes selon que la r{\'e}gion d'intensit{\'e} la plus faible est plus petite ou plus grande que l'autre r{\'e}gion. Nous observons alors une diff{\'e}rence entre les 3 crit{\`e}res. Nous notons alors un l{\'e}ger avantage pour le logarithme du rapport de vraisemblance. La similarit{\'e} des r{\'e}sultats sugg{\`e}re que nous pourrions utiliser indiff{\'e}remment une mesure ou l'autre dans plusieurs applications. Nous avons examin{\'e} comment le crit{\`e}re du rapport des moyennes peut {\^e}tre utilis{\'e} dans la segmentation hi{\'e}rarchique et nous avons compar{\'e} le r{\'e}sultat obtenu sur une image de synth{\`e}se avec les 2 autres crit{\`e}res.},
mypdf = {7},
slide = {https://BeaulieuJM.ca/slide/slideBea2005a.pdf},
keywords = {}
}

@Conference{Bea2005b,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
editor = {},
title = {Segmentation of Polarimetric Sar Images Composed of Textured and Non-Textured Fields},
booktitle = {Advanced SAR Workshop 2005, Canadian Space Agency},
volume = {},
publisher = {},
url = {http://www.asc-csa.gc.ca},
isbn = {},
doi = {},
address = {Montreal (Saint-Hubert)},
pages = {6},
year = {2005},
month = {15-17 Nov.},
abstract = {},
mypdf = {7},
slide = {https://BeaulieuJM.ca/slide/slideBea2005b.pdf},
keywords = {}
}

@Article{Bea2004a,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
title = {Segmentation of Textured Polarimetric {SAR} Scenes by Likelihood Approximation},
journal = {IEEE Transactions on Geoscience and Remote Sensing},
volume = {42},
number = {10},
pages = {2063-2072},
year = {2004},
month = {Oct.},
abstract = {A hierarchical stepwise optimization process is developed for polarimetric synthetic aperture radar image segmentation. We show that image segmentation can be viewed as a likelihood approximation problem. The likelihood segment merging criteria are derived using the multivariate complex Gaussian, the Wishart distribution, and the K-distribution. In the presence of spatial texture, the Gaussian-Wishart segmentation is not appropriate. The K-distribution segmentation is more effective in textured forested areas. The validity of the product model is also assessed, and a field-adaptable segmentation strategy combining different criteria is examined.},
publisher = {},
url = {https://ieeexplore.ieee.org/document/1344159},
issn = {0196-2892},
doi = {10.1109/TGRS.2004.835302},
mypdf = {11},
address = {},
keywords = {}
}

@Article{Bea2004b,
author = {Beaulieu, Jean-Marie},
title = {Utilisation of Contour Criteria in Micro-Segmentation of {SAR} Images},
journal = {International Journal of Remote Sensing},
volume = {25},
number = {17},
pages = {3497-3512},
year = {2004},
month = {Sept.},
abstract = {The segmentation of SAR (Synthetic Aperture Radar) images is greatly complicated by the presence of coherent speckle. To carry out this process a hierarchical segmentation algorithm based on stepwise optimization is used. It starts with each individual pixel as a segment and then sequentially merges the segment pair that minimizes the criterion. In a hypothesis testing approach, we show how the stepwise merging criterion is derived from the probability model of image regions. The Ward criterion is derived from the Gaussian additive noise model. A new criterion is derived from the multiplicative speckle noise model of SAR images. The first merging steps produce micro-regions. With standard merging criteria, the high noise level of SAR images results in the production of micro-regions that have unreliable mean and variance values and irregular shapes. If the micro-segments are not correctly delimited then the following steps will merge segments from different fields. In examining the evolution of the initial segments, we see that the merging should take into account spatial aspects. In particular, the segment contours should have good shapes. We present three measures based on contour shapes, using the perimeter, the area and the boundary length of segments. These measures are combined with the SAR criterion in order to guide correctly the segment merging process. The new criterion produces good micro-segmentation of SAR images. The criterion is also used in the following merges to produce larger segments. This is illustrated by synthetic and real image results.},
publisher = {},
url = {http://www.tandfonline.com/doi/abs/10.1080/01431160310001647714},
issn = {0143-1161},
doi = {10.1080/01431160310001647714},
mypdf = {11},
address = {},
keywords = {}
}

@Conference{Bea2003b,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Utilisation of Segment Border Information in Hierarchical Segmentation},
booktitle = {25rd Canadian Symposium on Remote Sensing & 11e Congrès de l’Association québécoise de télédétection},
volume = {},
publisher = {},
url = {https://crss-sct.ca},
isbn = {},
doi = {},
address = {Université de Montréal, Montréal QC},
pages = {},
year = {2003},
month = {Oct. 14-16},
abstract = {},
mypdf = {7},
keywords = {}
}

@Conference{Bea2003c,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
editor = {},
title = {Segmentation of Textured Areas using Polarimetric {SAR}},
booktitle = {25rd Canadian Symposium on Remote Sensing & 11e Congrès de l’Association québécoise de télédétection},
volume = {},
publisher = {},
url = {https://crss-sct.ca},
isbn = {},
doi = {},
address = {Université de Montréal, Montréal QC},
pages = {},
year = {2003},
month = {Oct. 14-16},
abstract = {},
mypdf = {7},
slide = {https://BeaulieuJM.ca/slide/slideBea2003c.pdf},
keywords = {}
}

@Conference{Bea2003e,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
editor = {},
title = {Classification and Segmentation of Radar Polarimetric Images},
booktitle = {Classification Society of North America Annual Meeting, 2003},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {Tallahassee, Florida},
pages = {},
year = {2003},
month = {June 25-15},
abstract = {},
mypdf = {6},
slide = {https://BeaulieuJM.ca/slide/slideBea2003e.pdf},
keywords = {}
}

@Conference{Bea2003d,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
editor = {},
title = {Segmentation of Polarimetric {SAR} Images: a Best Estimate Partitioning Approach},
booktitle = {Advanced SAR Workshop 2003, Canadian Space Agency},
volume = {},
publisher = {},
url = {http://www.asc-csa.gc.ca},
isbn = {},
doi = {},
address = {Montreal (Saint-Hubert)},
pages = {1-7},
year = {2003},
month = {June 25-27},
abstract = {},
mypdf = {7},
slide = {https://BeaulieuJM.ca/slide/slideBea2003d.pdf},
keywords = {}
}

@Conference{Bea2003a,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
editor = {},
title = {Segmentation of Textured Scenes using Polarimetric {SAR}s},
booktitle = {IEEE International Geoscience and Remote Sensing Symposium, GARSS'03},
volume = {1},
publisher = {},
url = {https://ieeexplore.ieee.org/document/1293804},
isbn = {},
doi = {10.1109/IGARSS.2003.1293804},
address = {Toulouse, France},
pages = {446-448},
year = {2003},
month = {21-25 July},
abstract = {},
mypdf = {11},
keywords = {}
}

@Conference{Bea2002,
author = {Beaulieu, Jean-Marie and Touzi, Ridha},
editor = {},
title = {Hierarchical Segmentation of Polarimetric {SAR} Images},
booktitle = {IEEE International Geoscience and Remote Sensing Symposium, IGARSS'02},
volume = {5},
publisher = {},
url = {https://ieeexplore.ieee.org/document/1026710},
isbn = {},
doi = {10.1109/IGARSS.2002.1026710},
address = {Toronto, ON, Canada},
pages = {2590-2592},
year = {2002},
month = {24-28 June},
abstract = {A hierarchical stepwise optimization process is used for polarimetric SAR image segmentation. The process starts with small sets of pixels as segments and then sequentially merges the segment pair that minimises a stepwise criterion. The polarimetric information could be represented by a covariance matrix. The proposed criterion is based upon the testing of the equality of covariance matrices of adjacent regions. The segmentation of SAR images is greatly complicated by the presence of coherent speckle. We are using spatial constraints and contour shapes in order to improve the segmentation results.},
mypdf = {11},
slide = {https://BeaulieuJM.ca/slide/slideBea2002.pdf},
keywords = {}
}

@Conference{Bea2001a,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Utilisation of Contour Criteria in Micro-Segmentation of {SAR} Images},
booktitle = {23rd Canadian Symposium on Remote Sensing & 10e Congrès de l’Association québécoise de télédétection},
volume = {},
publisher = {},
url = {https://crss-sct.ca},
isbn = {},
doi = {},
address = {},
pages = {91-100},
year = {2001},
month = {August},
abstract = {},
mypdf = {7},
slide = {https://BeaulieuJM.ca/slide/slideBea2001a.pdf},
keywords = {}
}

@Conference{Bea2001b,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {{SAR} Image Enhancement: Combining Image Filtering and Segmentation},
booktitle = {The 2001 International Conference on Imaging Science, Systems, and Technology, CISST'2001},
volume = {},
publisher = {},
url = {http://www2.ift.ulaval.ca/~beaulieu/home/paper/322CT_CISST_Beaulieu.pdf},
isbn = {},
doi = {},
address = {Las Vegas, NV, United States},
pages = {327-333},
year = {2001},
month = {25-28 June},
abstract = {},
mypdf = {7},
slide = {https://BeaulieuJM.ca/slide/slideBea2001b.pdf},
keywords = {},
openpdf = {https://pdfs.semanticscholar.org/0a0f/1f0e8e1399851573fed0973d73e4343c67a2.pdf?_ga=2.156291896.2051265828.1566578008-697872498.1566578008},
openid = {from Semantics}
}

@Conference{Bea2000a,
author = {Beaulieu, Jean-Marie and Mineau, Guy},
editor = {},
title = {Hierarchical Segmentation of {SAR} Images with Shape Criteria},
booktitle = {Classification Society of North America Annual Meeting, 2000},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {},
pages = {35},
year = {2000},
month = {June},
abstract = {},
mypdf = {6},
slide = {https://BeaulieuJM.ca/slide/slideBea2000a.pdf},
keywords = {}
}

@Conference{Bea2000b,
author = {Beaulieu, Jean-Marie and Bouzkraoui, Mohammed},
editor = {},
title = {D{\'e}tection des Arbres Individuels dans des Images de Haute R{\'e}solution},
booktitle = {Vision Interface 2000},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {},
pages = {311-317},
year = {2000},
month = {mai},
abstract = {},
mypdf = {7},
slide = {https://BeaulieuJM.ca/slide/slideBea2000b.pdf},
keywords = {}
}

@Conference{Bea1999,
author = {Beaulieu, Jean-Marie and Mineau, Guy},
editor = {},
title = {Evaluation of a Least Commitment Approach for Feature Preserving in {SAR} Image Filtering},
booktitle = {Classification Society of North America Annual Meeting, 1999},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {},
pages = {8},
year = {1999},
month = {August},
abstract = {},
mypdf = {6},
keywords = {}
}

@incollection{Min1998,
author = {Mineau, Guy W and Lahboub, Mounsif and Beaulieu, Jean-Marie},
title = {An object indexing methodology as support to object recognition},
booktitle = {Advances in Artificial Intelligence, Canadian AI 1998},
editor = {},
publisher = {Springer Berlin Heidelberg},
address = {Universit{\'e} Laval},
pages = {72-85},
year = {1998},
month = {},
url = {https://link.springer.com/chapter/10.1007/3-540-64575-6_41},
isbn = {978-3-540-69349-9},
doi = {10.1007/3-540-64575-6_41},
mypdf = {5},
abstract = {This paper presents an object recognition methodology which uses a step-by-step discrimination process. This process is made possible by the use of a classification structure built over examples of the objects to recognize. Thus, our approach combines numerical vision (object recognition) with conceptual clustering, showing how the latter helps the former, giving another example of useful synergy among different AI techniques. It presents our application domain: the recognition of road signs, which must support semi-autonomous vehicles in their navigational task. The discrimination process allows appropriate actions to be taken by the recognizer with regard to the actual data it has to recognize the object from: light, angle, shading, etc., and with regard to its recognition capabilities and their associated cost. Therefore, this paper puts the emphasis on this multiple criteria adaptation capability, which is the novelty of our approach.},
keywords = {}
}

@Conference{Naj1997,
author = {Najeh, Maher and Beaulieu, Jean-Marie},
editor = {},
title = {A Common Evaluation Approach to Smooting and Feature Preservation in {SAR} Image Filtering},
booktitle = {International Symposium: Geomatics in the era of radarsat},
volume = {},
publisher = {Canadian Space Agency},
url = {https://ostrnrcan-dostrncan.canada.ca/handle/1845/272042},
isbn = {},
doi = {https://doi.org/10.4095/331914},
address = {Ottawa, Canada},
pages = {},
year = {1997},
month = {May 25-30},
abstract = {},
mypdf = {9},
keywords = {}
}

@Conference{Bea1997a,
author = {Beaulieu, Jean-Marie and Najeh, Maher},
editor = {},
title = {A Least Commitment Approach to {SAR} Image Filtering},
booktitle = {International Symposium: Geomatics in the era of radarsat},
volume = {},
publisher = {Canadian Space Agency},
url = {https://ostrnrcan-dostrncan.canada.ca/handle/1845/272042},
isbn = {},
doi = {https://doi.org/10.4095/331914},
address = {Ottawa, Canada},
pages = {},
year = {1997},
month = {May 25-30},
abstract = {},
mypdf = {9},
keywords = {}
}

@Conference{Bea1996,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Filtrage des Images Radar par D{\'e}tection des R{\'e}gions Homog{\`e}nes},
booktitle = {9e Congr{\`e}s de l'Association Qu{\'e}b{\'e}coise de T{\'e}l{\'e}d{\'e}tection},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {},
pages = {},
year = {1996},
month = {30 avril - 3 mai},
abstract = {},
mypdf = {9},
slide = {https://BeaulieuJM.ca/slide/slideBea1996.pdf},
keywords = {}
}

@Conference{Jao1992,
author = {Jaoua, A and Beaulieu, Jean-Marie and Belkhiter, N and Deshernais, J and Reguig, M},
editor = {},
title = {Optimal rectangular decomposition of a finite binary relation},
booktitle = {International Conference on Discrete Mathematics (sixth conference)},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {},
pages = {},
year = {1992},
month = {},
abstract = {},
mypdf = {5},
keywords = {}
}

@Article{Bel1992,
author = {Belaid, M Ait and Edwards, Geoffrey and Jaton, A and Thomson, K P B and Beaulieu, Jean-Marie},
title = {Post-segmentation classification of images containing small agricultural fields},
journal = {Geocarto International},
volume = {7},
number = {3},
pages = {53--60},
year = {1992},
month = {},
abstract = {This paper presents results from applying a hierarchical segmentation algorithm to two agricultural data sets characterised by small fields. Several new techniques were developed over the course of this project. These include a new supervised classification technique for identifying segments and the inclusion of other information derived from the segmentation process. In addition, a technique for including cartographic information to help structure the segmentation is also described. The results indicate that significant improvement in classification accuracy can be achieved.
A number of problems which arise when working with segmentation data are also reported and discussed. These problems appear to be different enough from those encountered in pixel classifications to be worth describing in greater detail. The paper concludes with the lines of research being pursued to circumvent these problems and to further increase the fidelity of the segmentation results.},
publisher = {Taylor \& Francis},
url = {https://www.tandfonline.com/doi/abs/10.1080/10106049209354380},
issn = {},
doi = {10.1080/10106049209354380},
mypdf = {5},
address = {},
keywords = {}
}

@Conference{Bea1991,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Programming of Application Interface and Image Access Made Simple},
booktitle = {Canadian Conference on Electrical and Computer Engineering},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {Quebec, Quebec, Canada},
pages = {23.1.1-4},
year = {1991},
month = {Sept.},
abstract = {A good user interface is an important aspect of software products. We should also consider the effort and time spent for its programming. We present a "C" language extension facility which allows the definition of menus and fill-in forms. All the flexibility of the "C" programming language is preserved while making user interfaces programming easier. Following an object-oriented approach, we have also defined a set of functions and macros for image manipulation.},
mypdf = {9},
keywords = {}
}

@Conference{Bea1990b,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Hierarchical Segmentation of {SAR} Picture},
booktitle = {Image'Com 90, Bordeaux},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {},
pages = {392-397},
year = {1990},
month = {Nov.},
abstract = {The segmentation of SAR (Synthetic Aperture Radar) pictures is greatly complicated by the presence of coherent speckle in the image. The complex structure of the SAR pictures requires the utilization of a composite criterion for the segmentation. This paper takes advantage of a powerful hierarchical segmentation technique based upon step-wise optimization. The algorithm could easily be adapted to complex criterion. We present a two stage approach. A constant approximation criterion is first employed to yield an initial partition of the image. Then, a composite criterion is employed to continue the merging. The segment means and variances are then exploited in the step-wise criterion (segment similarity measure). Moreover, the segment shape is employed to reduce the formation of random contours. Good segmentation results are obtained, and they compare advantageously with other segmentation approaches. The algorithm produces a good separation of regions, and in the same time, yields accurate boundary location.},
mypdf = {7},
keywords = {}
}

@Conference{Bea1990a,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Versatile And Efficient Hierarchical Clustering For Picture Segmentation},
booktitle = {International Geoscience and Remote Sensing Symposium, IGARSS'90},
volume = {},
publisher = {IEEE},
url = {https://ieeexplore.ieee.org/document/688832},
isbn = {},
doi = {10.1109/IGARSS.1990.688832},
address = {The University of Maryland, College Park,Maryland},
pages = {1663-1663},
year = {1990},
month = {May 20-24},
abstract = {},
mypdf = {11},
keywords = {Approximation algorithms; Approximation error; Bismuth; Clustering algorithms; Corporate acquisitions; Data structures; Image segmentation; Partitioning algorithms; Probability; Shape}
}

@Conference{Edw1990,
author = {Edwards, G and Ait-Belaid, M and Thomson, K P B and Cauchon, G and Beaulieu, Jean-Marie},
editor = {},
title = {Cartographic Information as a Structuring Principle for Image Segmentation},
booktitle = {ISPRS Commission II/VII International Workshop},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {University of Main, Orono},
pages = {},
year = {1990},
month = {},
abstract = {Image segmentation is a relatively new image processing procedure which breaks an image into a set of contiguous regions characterized by some criteria of homogeneity and/or continuity. Segmentation is carried out before classification, and hence allows for classification which is based on the average spectral cha.racteristics of the componente ntities. Furthermore,s uch classificationc an also benefit from shape and context information. These advantages make image segmentation a powerful tool for automated image interpretation. However, when segmentation algorithms arc applied to remotely sensed images they often yield image partitions which do not correspond well to meaningful structures in the scene. This paper presents a scenario for introducing auxiliary cartographic information into image segmentation. In particular, we have found that the introduction of partially complete cartographic infonnation, based on the more stable elements of a scene, serves as a "structuring principle" for the segmentation algorithm. The resulting partitions contain structures which correspond better with meaningful entities in the image, even in regions for which no cartographic information was introduced, and hence greatly facilitate the task of automated image interpretation. This technique is shown to be particularly useful when crop covers are small compared to the pixel size. Examples are presented where cadastral information has been used as such a structuring principle with SPOT multispectral imagery and with airborne SAR imagery of agricultural scenes},
mypdf = {5},
keywords = {}
}

@Conference{Kal1990,
author = {Kaliaguine, Nicolas and Beaulieu, Jean-Marie},
editor = {},
title = {Segmentation of {SAR} Picture},
booktitle = {Canadian Conference on Electrical and Computer Engineering},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {Ottawa, Ontario, Canada},
pages = {69.5.1-4},
year = {1990},
month = {},
abstract = {This paper presents a new and simple filter useful for the segmentation of SAR (Synthetic Apertur Radar) picture. This filter corresponds to an adiabatic transformation, It is reversible and does not produce information lose. The purpose of the filter is to transform the noise model of the image from multip1icative gaussian to an additive gaussian one which is easier to process.},
mypdf = {7},
keywords = {}
}

@Conference{Val1988,
author = {Velarde, Cesar and Beaulieu, Jean-Marie},
editor = {},
title = {Quantitative Evaluation of Image Segmentation Techniques},
booktitle = {Canadian Conference on Electrical and Computer Engineering},
volume = {},
publisher = {},
url = {},
isbn = {0-9694170-0-4},
doi = {},
address = {Montreal, Canads},
pages = {314-317},
year = {1989},
month = {September 17-20},
abstract = {The objective evaluation of picture segmentation techniques is an import ant and difficult topic. This paper presents an objective evaluation approach, based upon the comparison of the results of a given segmentation technique to either the ground truth, or to the results of another technique. Three performance criteria are defined : 1) the ability to extract the structure of the images, 2) the sensitivity to noise , and 3) the consistency between the results of two techniques. These performance criteria are measured for 4 segmentation techniques, over a set of artificial images. The results show that one of the techniques outperforms the others in retrieving the structure of the images. They also show that two techniques are less sensitive to noise than the others . Finally, we note that the four techniques produced essentially different picture partitions.},
mypdf = {9},
keywords = {}
}

@Conference{Bel1989,
author = {Belaid, M Ait and Thomson, K P B and Edwards, Geoffrey and Beaulieu, Jean-Marie},
editor = {},
title = {Sementation d'image spot integree a l'information cartographique en vu de l'etablissment de la carte d'utilization de sol au maroc},
booktitle = {IEEE International Geoscience and Remote Sensing Symposium, IGARSS'89},
volume = {1},
publisher = {IEEE},
url = {https://ieeexplore.ieee.org/document/567151},
isbn = {},
doi = {10.1109/IGARSS.1989.567151},
address = {Vancouver, Canada},
pages = {56-59},
year = {1989},
month = {July 10-14},
abstract = {This paper is concerned with the integration of remote sensing and conventional data. It presents a purely digital method of merging a multispectral SPOT image and field boundaries. This yields a product which is a new image registered to the national grid of Morocco, having four channels with images resampled to 10 m. The fourth channel contains the field boundaries which are digitized using the spatial information system PAMAP.
A Hierarchical Step-Wise Optimization (HSWO) algorithm developed by Beaulieu is applied to the new four band ``image'' to test the capability of the segmentation to map the land use and to provide the crop inventory in small areas of land.},
mypdf = {11},
keywords = {}
}

@Conference{Edw1989,
author = {Edwards, Geoffrey and Beaulieu, Jean-Marie},
editor = {},
title = {Segmentation of {SAR} Imagery Containing Forest Clear Cuts},
booktitle = {IEEE International Geoscience and Remote Sensing Symposium, IGARSS'89},
volume = {3},
publisher = {IEEE},
url = {https://ieeexplore.ieee.org/document/576042},
isbn = {},
doi = {10.1109/IGARSS.1989.576042},
address = {Vancouver, Canada},
pages = {1195-1197},
year = {1989},
month = {July 10-14},
abstract = {A Hierarchical Step-Wise Optimisation (HSWO) algorithm has been adapted to the problem of identifying and mapping forest clear cuts in synthetic aperture radar (SAR) C-band imagery.
Preliminary results are presented. The mean grey level of a segment is the most useful segment discriminator, especially for recent clear cuts, but relative segment size and the ratio of perimeter length to surface area (P/A) appear to be useful secondary discriminators. A filtered image which is segmented appears to be the most reliable for locating clear cuts, whereas the unfiltered image, when segmented, yields better boundary information. A method for combining both segment partitions is presented. All clear cuts in the sample were identified. Surface areas concord with manually estimated values.},
mypdf = {13},
keywords = {Clouds; Forestry; Image segmentation; Information filtering; Information filters; Layout; Partitioning algorithms; Pixel; Satellites; Synthetic aperture radar}
}

@incollection{Bea1989c,
author = {Beaulieu, Jean-Marie},
title = {Segmentation Hi{\'e}rarchique de l'Image par Optimisation S{\'e}quentielle},
booktitle = {T{\'e}l{\'e}d{\'e}tection et Gestion des Ressources},
volume = {Vol. VI},
editor = {Bernier and Bonn and Gagnon},
publisher = {Association Qu{\'e}b{\'e}coise de T{\'e}l{\'e}d{\'e}tection, Qu{\'e}bec},
address = {Ste-Foy, Qu{\'e}bec},
pages = {245-251},
year = {1989},
month = {},
url = {},
isbn = {},
doi = {},
mypdf = {9},
abstract = {A hierarchical segmentation algorithm based upon step-wise optimization is presented. The algorithm starts with an initial picture partition, and at each iteration, the two most similar segments are merged by optimizing a "step-wise criterion". This yields a hierarchical decomposition of the picture, which is data driven with no restriction on segment shapes. The algorithm is designed so as to reduce the computing
time. Hence, the fact that a segment merge affects only the surrounding segments is exploited. Good results are produced because of the global and gradual manner of the sequential optimization.},
keywords = {}
}

@incollection{Bea1989b,
author = {Beaulieu, Jean-Marie and Boulanger, Pierre},
title = {Segmentation of Range Images by Piecewise Approximation with Shape Constraints},
booktitle = {Computer Vision and Shape Recognition},
editor = {},
publisher = {World Scientific Pub Co Inc},
address = {},
pages = {87},
year = {1989},
month = {},
url = {https://www.worldscientific.com/doi/abs/10.1142/9789814434362_0004},
isbn = {978-981-4434-36-2},
doi = {10.1142/9789814434362_0004},
mypdf = {5},
abstract = {Piecewise functional approximation of picture is shown to be a useful tool for the segmentation of range (3D) image. A hierarchical step-wise optimization algorithm is employed to transform the global optimization problem into one of sequential optimization. The step-wise criterion then corresponds to the increase of the approximation error produced by the merge of two segments. The segmentation results of range image of polyhedra are shown with the utilization of a planar approximation model. Constraints on segment contour length and segment shape is then added to improve the results.},
keywords = {Three-dimensionalhierarchical, segmentation, picture approximation, segment shape}
}

@Article{Bea1989a,
author = {Beaulieu, Jean-Marie and Goldberg, Moris},
title = {Hierarchy in Picture Segmentation: a Stepwise Optimization Approach},
journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
volume = {11},
number = {2},
pages = {150--163},
year = {1989},
month = {},
abstract = {A segmentation algorithm based on sequential optimization which produces a hierarchical decomposition of the picture is presented. The decomposition is data driven with no restriction on segment shapes. It can be viewed as a tree, where the nodes correspond to picture segments and where links between nodes indicate set inclusions. Picture segmentation is first regarded as a problem of piecewise picture approximation, which consists of finding the partition with the minimum approximation error. Then, picture segmentation is presented as an hypothesis-testing process which merges only segments that belong to the same region. A hierarchical decomposition constraint is used in both cases, which results in the same stepwise optimization algorithm. At each iteration, the two most similar segments are merged by optimizing a stepwise criterion. The algorithm is used to segment a remote-sensing picture, and illustrate the hierarchical structure of the picture},
publisher = {},
url = {https://ieeexplore.ieee.org/document/16711},
issn = {0162-8828},
doi = {10.1109/34.16711},
mypdf = {12},
address = {},
keywords = {computerised picture processing; data structure; hierarchical decomposition; iterative methods; optimisation; picture segmentation; sequential optimization; stepwise optimization; tree; trees (mathematics); nw-05}
}

@Conference{Bea1988a,
author = {Beaulieu, Jean-Marie and Boulanger, Pierre},
editor = {},
title = {Segmentation of Range Image by Piecewise Approximation with Shape Constraints},
booktitle = {Vision Interface'88},
volume = {},
publisher = {Proceedings Vision Interface'88},
url = {},
isbn = {},
doi = {},
address = {Edmonton, Canada},
pages = {19-14},
year = {1988},
month = {June},
abstract = {Piecewise functional approximation of picture is shown to be a useful tool for the segmentation of range (3D) image. A hierarchical step-wise optimization algorithm is employed to transform the global optimization problem into one of sequential optimization. The step-wise criterion then corresponds to the increase of the approximation error produced by the merge of two segments. The segmentation results of range image of polyhedra are shown with the utilization of a planar approximation model. Constraints on segment contour length and segment shape is then added to improve the results.},
mypdf = {9},
keywords = {}
}

@Conference{Bea1988b,
author = {Beaulieu, Jean-Marie},
editor = {},
title = {Segmentation Hi{\'e}rarchique de l'Image par Optimisation S{\'e}quentielle},
booktitle = {6e Congr{\`e}s de l'Association Qu{\'e}b{\'e}coise de T{\'e}l{\'e}d{\'e}tection},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {Sherbrooke, Quebec},
pages = {},
year = {1988},
month = {May},
abstract = {A hierarchical segmentation algorithm based upon step-wise optimization is presented. The algorithm starts with an initial picture partition, and at each iteration, the two most similar segments are merged by optimizing a "step-wise criterion". This yields a hierarchical decomposition of the picture, which is data driven with no restriction on segment shapes. The algorithm is designed so as to reduce the computing
time. Hence, the fact that a segment merge affects only the surrounding segments is exploited. Good results are produced because of the global and gradual manner of the sequential optimization.},
mypdf = {9},
keywords = {}
}

@Conference{Bea1985b,
author = {Beaulieu, Jean-Marie and Goldberg, Morris},
title = {Selection of Segment Similarity Measures for Hierarchical Picture Segmentation},
booktitle = {Proceedings Graphics Interface 1985},
volume = {},
publisher = {Canadian Information Processing Soc, Toronto, Ont, Can},
address = {Montreal, Que, Can},
issn = {07135424},
year = {1985},
month = {May 27-31},
pages = {179-186},
url = {},
doi = {},
abstract = {The problem of defining appropriate segment similarity measures for picture segmentation is examined. In agglomerative hierarchical segmentation, two segments are compared and merged if found similar. The proposed Hierarchical Step-Wise Optimization (HSWO) algorithm finds and then merges the two most similar segments, on a step-by-step basis. By considering picture segmentation as a piece-wise picture approximation problem, the similarity measure (or the step-wise criterion) is related to the overall approximation error. The measure then corresponds to the increase of the approximation error resulting from merging two segments. Similarity measures derived from constant approximations (zeroth order polynomials) and planar approximations (first order polynomials) are applied to a Landsat picture, and the results are presented.},
mypdf = {9},
keywords = {}
}

@incollection{Bea1985,
author = {Beaulieu, Jean-Marie and Goldberg, Morris},
title = {Selection of Segment Similarity Measures for Hierarchical Picture Segmentation},
booktitle = {Computer-Generated Images},
editor = {},
publisher = {Springer},
address = {},
pages = {87-97},
year = {1985},
month = {},
url = {https://link.springer.com/chapter/10.1007/978-4-431-68033-8_8},
isbn = {978-4-431-68033-8},
doi = {10.1007/978-4-431-68033-8_8},
abstract = {The problem of defining appropriate segment similarity measures for picture segmentation is exmained. In agglomerative hierqarchical segmentation, two segments are coamapared and merged if found similar. The propesed Hierarchical Step-Wise Optimization (HSWO) algorithm finds and then merges the two most similar segements, on a step-by-step basis. By considering picture segmentation as a piece-wise picture approximation problem, the similarity measure (or the step-wise criterion) is related to the overall approximation error. The measure then corresponds to the increase of the approximation error resulting from merging two segments. Similarity measures derived from constant approximations (zeroth order polynomials) and planar approximations (first order polynomials). An adaptive measurebased upon local variance is also used. The advantages of combining similarity measures (or cirteria) are also stressed. Different picture areas can require different measures which must therefore be combined in order to obtain good overall results. Moreover, in hierarchical segmentation, simple measures can be used for the first merging steps, while, at a higher level of the segment hierarchy, more complex measures can be employed.},
mypdf = {5},
keywords = {Hierarchical segmentation; Similarity measures; Clustering}
}

@phdthesis{Bea1984,
author = {Beaulieu, Jean-Marie},
title = {Hierarchical Picture Segmentation by Step-Wise Optimization},
school = {University of Ottawa (Canada)},
dept = {Electrical Engineering},
degree = {Ph.D.},
thesis = {PhD},
address = {},
pages = {220},
year = {1984},
month = {},
publisher = {University of Ottawa (Canada)},
url = {https://ruor.uottawa.ca/bitstream/10393/4824/1/NL20894.PDF},
isbn = {},
doi = {},
wdown = {},
mypdf = {14},
abstract = {},
keywords = {}
}

@Conference{Bea1983,
author = {Beaulieu, Jean-Marie and Goldberg, Morris},
editor = {},
title = {Step-Wise Optimization for Hierarchical Picture Segmentation},
booktitle = {Conference on Computer Vision and Pattern Recognition},
volume = {59},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {Washington, D.C.},
pages = {64},
year = {1983},
month = {},
abstract = {Hierarchical picture segmentations are very useful in picture analysis. We present a sequential segment merging algorithm for picture segmentation. Each iteration merges two segments which optimize a step-wise criterion. We relate picture segmentation to optimization problems. The implementation of the segmentation algorithm is examined, and results are presented and discussed.},
mypdf = {7},
keywords = {}
}

@Conference{Bea1982,
author = {Beaulieu, Jean-Marie and Goldberg, Morris},
editor = {},
title = {Hierarchical Picture Segmentation by Approximation},
booktitle = {Proc. Can. Commun. Energy Conf},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {Montreal, Canada},
pages = {393-396},
year = {1982},
month = {},
abstract = {Piecewise functional approximation of picture is shown to be a useful tool for primitive and segment extraction. We present an algorithm based on a hierarchical segmentation structure. Functional approximation criteria are employed to direct the sequential merging of segments. A criteria related to the mean squared error is used to guide the merging. The algorithm yields solutions to the picture segmentation and approximation problem. Techniques described have been applied to a LANDSAT picture of an agricultural region (32 multiplied by 32 pixels).},
mypdf = {7},
keywords = {}
}

@Article{Coh1980,
author = {Cohen, Paul and Beaulieu, Jean-Marie},
title = {The Modeling and Generation of Visual Textures},
journal = {Canadian Electrical Engineering Journal},
volume = {5},
number = {3},
pages = {5--8},
year = {1980},
month = {},
abstract = {Describes two methods of generating artificial textures, based on a second order statistical model. By choosing the correct model parameters, these methods make it possible to obtain textures with given statistical properties (granularity, homogeneity, periodicity, desired directions). Such artificial textures are a useful tool both in analyzing the stochastic structure of real images and in studying the discriminatory power of the eye.},
publisher = {},
url = {https://ui.adsabs.harvard.edu/abs/1980CEEJ....5....5C/abstract},
isbn = {},
doi = {},
mypdf = {},
address = {},
keywords = {}
}

@Conference{Bea1979,
author = {Beaulieu, Jean-Marie and Cohen, Paul and Adoul, Jean-Pierre},
editor = {},
title = {Digital Picture Generation by Texture and Contour Modeling},
booktitle = {22nd Midwest Symposium on Circuits and Systems},
volume = {},
publisher = {},
url = {},
isbn = {},
doi = {},
address = {Philadelphia},
pages = {344-348},
year = {1979},
month = {June 17-19},
abstract = {This paper describes some efficient techniques for visual stochastic field generation, based on statistical models of textures and contours. Computer simulation of these techniques yields pictures which show controllable properties of granularity, clustering and symmetry depending on the specified model parameters.},
mypdf = {7},
keywords = {}
}