2D/3D Object Recognition and Categorization Approaches for Robotic Grasping

Abstract : Object categorization and manipulation are critical tasks for a robot to operate in the household environment. In this paper, we propose new methods for visual recognition and categorization. We describe 2D object database and 3D point clouds with 2D/3D local descriptors which we quantify with the k-means clustering algorithm for obtaining the Bag of Words (BOW). Moreover, we develop a new global descriptor called VFH-Color that combines the original version of Viewpoint Feature Histogram (VFH) descriptor with the color quantization histogram, thus adding the appearance information that improves the recognition rate. The acquired 2D and 3D features are used for training Deep Belief Network (DBN) classifier. Results from our experiments for object recognition and categorization show an average of recognition rate between 91% and 99% which makes it very suitable for robot-assisted tasks.
Type de document :
Chapitre d'ouvrage
Advances in Soft Computing and Machine Learning in Image Processing, 730, pp.567-593, 2017, 978-3-319-63754-9
Liste complète des métadonnées

Littérature citée [58 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01676387
Contributeur : Haris Ahmad Khan <>
Soumis le : vendredi 5 janvier 2018 - 15:22:23
Dernière modification le : lundi 8 octobre 2018 - 21:00:02
Document(s) archivé(s) le : samedi 5 mai 2018 - 01:59:05

Fichier

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

Identifiants

  • HAL Id : hal-01676387, version 1

Collections

Citation

Nabila Zrira, Mohamed Hannat, El Houssine Bouyakhf, Haris Ahmad Khan. 2D/3D Object Recognition and Categorization Approaches for Robotic Grasping. Advances in Soft Computing and Machine Learning in Image Processing, 730, pp.567-593, 2017, 978-3-319-63754-9. 〈hal-01676387〉

Partager

Métriques

Consultations de la notice

63

Téléchargements de fichiers

314