Titre :

A comparative study of existing metrics for 3D-mesh segmentation evaluation

Auteur(s) :

Benhabiles, Halim [Auteur]
FOX MIIRE [LIFL]
Vandeborre, Jean Philippe [Auteur]
FOX MIIRE [LIFL]
Institut TELECOM/TELECOM Lille1
Lavoué, Guillaume [Auteur]
Geometry Processing and Constrained Optimization [M2DisCo]
Daoudi, Mohamed [Auteur]
FOX MIIRE [LIFL]
Institut TELECOM/TELECOM Lille1

Titre de la revue :

The Visual Computer

Pagination :

1451-1466

Éditeur :

Springer Verlag

Date de publication :

2010-12

ISSN :

0178-2789

Mot(s)-clé(s) en anglais :

3D-mesh segmentation
ground-truth
similarity metric
subjective tests
evaluation

Discipline(s) HAL :

Informatique [cs]/Vision par ordinateur et reconnaissance de formes [cs.CV]
Informatique [cs]/Multimédia [cs.MM]

Résumé en anglais : [en]

In this paper, we present an extensive experimental comparison of existing similarity metrics addressing the quality assessment problem of mesh seg- mentation. We introduce a new metric named the 3D Normalized Probabilistic ...
Lire la suite >In this paper, we present an extensive experimental comparison of existing similarity metrics addressing the quality assessment problem of mesh seg- mentation. We introduce a new metric named the 3D Normalized Probabilistic Rand Index (3D-NPRI) which outperforms the others in terms of properties and discriminative power. This comparative study includes a subjective experiment with human observers and is based on a corpus of manually segmented models. This corpus is an improved version of our previous one. It is composed of a set of 3D-mesh models grouped in different classes associated with several manual ground-truth segmentations. Finally the 3D-NPRI is applied to evaluate six recent segmentation algorithms using our corpus and the Chen's et al. corpus.Lire moins >

Langue :

Anglais

Vulgarisation :

Non

Collections :

• Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189

Source :

Harvested from HAL