Distanceless Label Propagation: an Efficient Direct Connected Component Labeling Algorithm for GPUs

Abstract : Modern computer architectures are mainly composed of multi-core processors and GPUs. Consequently, solely providing a sequential implementation of algorithms or comparing algorithm performance without regard to architecture is no longer pertinent. Today, algorithms have to address parallelism, multithreading and memory topology (private/shared memory, cache or scratchpad, ...). Most Connected Component Labeling (CCL) algorithms are sequential, direct and optimized for processors. Few were designed specifically for GPU architectures and none were designed to be adapted to different architectures. The most efficient GPU implementations are iterative; in order to manage synchronizations between processing units, but the number of iterations depends on the image shape and density. This paper describes the DLP (Distanceless Label Propagation) algorithms, an adaptable set of algorithms usable both on GPU and multi-core architectures, and DLP-GPU, an efficient direct CCL algorithm for GPU based on DLP mechanisms.
Type de document :
Communication dans un congrès
IPTA2017 - International Conference on Image Processing Theory, Tools and Applications, Nov 2017, Montreal, Canada
Liste complète des métadonnées

Littérature citée [17 références]  Voir  Masquer  Télécharger

https://hal-centralesupelec.archives-ouvertes.fr/hal-01656756
Contributeur : Laurent Cabaret <>
Soumis le : mercredi 13 décembre 2017 - 16:11:02
Dernière modification le : vendredi 31 août 2018 - 09:25:56
Document(s) archivé(s) le : mercredi 14 mars 2018 - 12:31:36

Fichier

IPTA_79.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01656756, version 1

Citation

Laurent Cabaret, Lionel Lacassagne, Daniel Etiemble. Distanceless Label Propagation: an Efficient Direct Connected Component Labeling Algorithm for GPUs. IPTA2017 - International Conference on Image Processing Theory, Tools and Applications, Nov 2017, Montreal, Canada. 〈hal-01656756〉

Partager

Métriques

Consultations de la notice

419

Téléchargements de fichiers

80