Article: 3D object classification based on deep belief networks and point clouds Journal: International Journal of Computational Vision and Robotics (IJCVR) 2019 Vol.9 No.6 pp.527 - 558 Abstract: Since the discovery of 3D sensors such as Kinect camera, 3D object models, and point clouds become frequently used in many areas. The most important one is the 3D object recognition and classification in robotic applications. This type of sensors, like the human vision, allows generating the object model from a field of view or even a complete 3D object model by combining several individual Kinect frames. In this work, we propose a new feature learning-based object classification approach using point cloud library (PCL) detectors and descriptors and deep belief networks (DBNs). Before developing the classification approach, we evaluate 3D descriptors by proposing a new pipeline that uses the L2-distance and the recognition threshold. 3D descriptors are computed on different datasets, in order to achieve the best descriptors. Subsequently, these descriptors are used to learn robust features in the classification approach using DBNs. We evaluate the performance of these contributions on two datasets; Washington RGB-D and our real 3D object datasets. The results show that the proposed approach outperforms advanced methods by approximately 5% in terms of accuracy. Inderscience Publishers - linking academia, business and industry through research

Title: 3D object classification based on deep belief networks and point clouds

Authors: Fatima Zahra Ouadiay; Nabila Zrira; Mohamed Hannat; El Houssine Bouyakhf; Majid Mohamed Himmi

Addresses: Faculty of Sciences, Mohammed V University in Rabat, Rabat, Morocco ' Faculty of Sciences, Mohammed V University in Rabat, Rabat, Morocco ' Faculty of Sciences, Mohammed V University in Rabat, Rabat, Morocco ' Faculty of Sciences, Mohammed V University in Rabat, Rabat, Morocco ' Faculty of Sciences, Mohammed V University in Rabat, Rabat, Morocco

Abstract: Since the discovery of 3D sensors such as Kinect camera, 3D object models, and point clouds become frequently used in many areas. The most important one is the 3D object recognition and classification in robotic applications. This type of sensors, like the human vision, allows generating the object model from a field of view or even a complete 3D object model by combining several individual Kinect frames. In this work, we propose a new feature learning-based object classification approach using point cloud library (PCL) detectors and descriptors and deep belief networks (DBNs). Before developing the classification approach, we evaluate 3D descriptors by proposing a new pipeline that uses the L2-distance and the recognition threshold. 3D descriptors are computed on different datasets, in order to achieve the best descriptors. Subsequently, these descriptors are used to learn robust features in the classification approach using DBNs. We evaluate the performance of these contributions on two datasets; Washington RGB-D and our real 3D object datasets. The results show that the proposed approach outperforms advanced methods by approximately 5% in terms of accuracy.

Keywords: Kinect; 3D object classification; point cloud library; PCL; recognition threshold; deep belief networks; DBNs; Washington RGB-D; Robotic; Computational Vision.

DOI: 10.1504/IJCVR.2019.104037

International Journal of Computational Vision and Robotics, 2019 Vol.9 No.6, pp.527 - 558

Received: 01 Aug 2018
Accepted: 19 Oct 2018
Published online: 09 Dec 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: 3D object classification based on deep belief networks and point clouds

Keep up-to-date