Les Reseaux Associatifs constituent un modele de calcul parallele original permettant d'expri... more Les Reseaux Associatifs constituent un modele de calcul parallele original permettant d'exprimer aisement des calculs sur des donnees distribuees sur un reseau. Ce reseau, represente par un graphe, generalement sous forme de maille, sert a vehiculer les valeurs entre les sommets. Par des mecanismes d'activation ou de desactivation des arcs de ces graphes, il est alors possible de former des structures irregulieres sur lesquelles peuvent etre appliquees des primitives de calcul globales appelees associations. Le traitement et l'analyse d'image sont les principales applications de ce modele. Notre etude porte sur la mise en œuvre d'un modele de programmation des Reseaux Associatifs sur machines paralleles a usage general, afin de proposer un environnement de programmation adapte a ce support d'execution. Nous etudions dans un premier temps les specifications du modele de programmation et nous proposons une premiere approche d'extension pour la representation simplifiee de regions d'une image par les Reseaux Virtuels. Nous nous interessons ensuite aux conditions d'une parallelisation efficace des primitives de calcul sur reseaux de stations ou sur machines a memoire partagee. Nous montrons enfin, au travers de deux applications classiques d'analyse d'image, en quoi ce modele de programmation constitue une solution elegante et compacte pour l'expression des algorithmes d'analyse d'image et la parallelisation automatique efficace de programmes complets. Ce travail a conduit au developpement d'ANET, ensemble de bibliotheques de programmation pour l'analyse d'image sur machines paralleles a memoire partagee.
IEEE International Conference on Image Processing 2005, 2005
Many basic computations can be done by means of iterative neighborhood-based calculations, includ... more Many basic computations can be done by means of iterative neighborhood-based calculations, including threshold, optimum, distance transform, contour closing, mathematical morphology, etc. Some of them can be performed using rows-per-rows scans (A. Rosenfeld and J.-L. Pfaltz, 1966) (G. Borgefors, 1986). Such regular computations allow to optimize the use of caches on standard architecture, and to achieve computations in good times.
Seventh International Workshop on Computer Architecture for Machine Perception (CAMP'05)
Several hard problems have to be addressed in order to parallelize image analysis algorithms. Ind... more Several hard problems have to be addressed in order to parallelize image analysis algorithms. Indeed, at the region level, these algorithms handle irregular (and sometimes strongly dynamic) data-structures. Moreover, they often lead to an unbalanced amount of computations, which is quite impossible to foresee offline. This paper focus on the parallelization of the ANET image analysis programming environment. Thanks to
Fuzzy tree mining has been recently introduced in order to extract frequent subtrees from databas... more Fuzzy tree mining has been recently introduced in order to extract frequent subtrees from databases of labeled trees. It has many applications, especially for handling semi-structured data (e.g., XML). In this framework, soft approaches have been proposed, also known as fuzzy tree mining. They allow the methods to better recognize patterns that are embedded in the database, even if the
Proceedings Fifth IEEE International Workshop on Computer Architectures for Machine Perception, 2000
In this paper we present the programming environment Anet for image analysis, that aims to bridge... more In this paper we present the programming environment Anet for image analysis, that aims to bridge the gap between programmability requirements and parallel efficiency. It is based on the graph based associative nets computing model, and allows irregular data manipulation. As it is intrinsically a parallel model, parallel execution can be quite naturally considered, and as the number of primitives
We introduce two different representation approaches and propose two techniques to estimate the p... more We introduce two different representation approaches and propose two techniques to estimate the position of vanishing points in an image, one bused on a probabilistic strategy and the other focused on a deterministic analysis. Unlike most of the methods so far developed, which exploit the Gaussian sphere, the new techniques operate in the (ρ, θ) polar parameter space and in
Les Reseaux Associatifs constituent un modele de calcul parallele original permettant d'expri... more Les Reseaux Associatifs constituent un modele de calcul parallele original permettant d'exprimer aisement des calculs sur des donnees distribuees sur un reseau. Ce reseau, represente par un graphe, generalement sous forme de maille, sert a vehiculer les valeurs entre les sommets. Par des mecanismes d'activation ou de desactivation des arcs de ces graphes, il est alors possible de former des structures irregulieres sur lesquelles peuvent etre appliquees des primitives de calcul globales appelees associations. Le traitement et l'analyse d'image sont les principales applications de ce modele. Notre etude porte sur la mise en œuvre d'un modele de programmation des Reseaux Associatifs sur machines paralleles a usage general, afin de proposer un environnement de programmation adapte a ce support d'execution. Nous etudions dans un premier temps les specifications du modele de programmation et nous proposons une premiere approche d'extension pour la representation simplifiee de regions d'une image par les Reseaux Virtuels. Nous nous interessons ensuite aux conditions d'une parallelisation efficace des primitives de calcul sur reseaux de stations ou sur machines a memoire partagee. Nous montrons enfin, au travers de deux applications classiques d'analyse d'image, en quoi ce modele de programmation constitue une solution elegante et compacte pour l'expression des algorithmes d'analyse d'image et la parallelisation automatique efficace de programmes complets. Ce travail a conduit au developpement d'ANET, ensemble de bibliotheques de programmation pour l'analyse d'image sur machines paralleles a memoire partagee.
IEEE International Conference on Image Processing 2005, 2005
Many basic computations can be done by means of iterative neighborhood-based calculations, includ... more Many basic computations can be done by means of iterative neighborhood-based calculations, including threshold, optimum, distance transform, contour closing, mathematical morphology, etc. Some of them can be performed using rows-per-rows scans (A. Rosenfeld and J.-L. Pfaltz, 1966) (G. Borgefors, 1986). Such regular computations allow to optimize the use of caches on standard architecture, and to achieve computations in good times.
Seventh International Workshop on Computer Architecture for Machine Perception (CAMP'05)
Several hard problems have to be addressed in order to parallelize image analysis algorithms. Ind... more Several hard problems have to be addressed in order to parallelize image analysis algorithms. Indeed, at the region level, these algorithms handle irregular (and sometimes strongly dynamic) data-structures. Moreover, they often lead to an unbalanced amount of computations, which is quite impossible to foresee offline. This paper focus on the parallelization of the ANET image analysis programming environment. Thanks to
Fuzzy tree mining has been recently introduced in order to extract frequent subtrees from databas... more Fuzzy tree mining has been recently introduced in order to extract frequent subtrees from databases of labeled trees. It has many applications, especially for handling semi-structured data (e.g., XML). In this framework, soft approaches have been proposed, also known as fuzzy tree mining. They allow the methods to better recognize patterns that are embedded in the database, even if the
Proceedings Fifth IEEE International Workshop on Computer Architectures for Machine Perception, 2000
In this paper we present the programming environment Anet for image analysis, that aims to bridge... more In this paper we present the programming environment Anet for image analysis, that aims to bridge the gap between programmability requirements and parallel efficiency. It is based on the graph based associative nets computing model, and allows irregular data manipulation. As it is intrinsically a parallel model, parallel execution can be quite naturally considered, and as the number of primitives
We introduce two different representation approaches and propose two techniques to estimate the p... more We introduce two different representation approaches and propose two techniques to estimate the position of vanishing points in an image, one bused on a probabilistic strategy and the other focused on a deterministic analysis. Unlike most of the methods so far developed, which exploit the Gaussian sphere, the new techniques operate in the (ρ, θ) polar parameter space and in
Uploads
Papers by Nicolas Sicard