Skip to Main content Skip to Navigation
New interface
Conference papers

Reconnaissance d'objets 3D par réseau d'automates

Abstract : A lot of vision applications, for instance in robotics, require identification of 3D objects. Use of complex methods, based on model matching, is not always necessary, and may be too computationally expensive. We propose a fast and simple method for recognition of 3D objects. The method takes profit of the learning capabilities of a neural network. The idea is to train a neural network on some views of each object. In order to reduce the amount of data, the object is characterized by its silhouette. At the end of the learning phase, the generalization capabilities of the network allow it to recognize non-learned views. After a description of the proposed method, we will present experimental results obtained on a data base of 216 images.
Document type :
Conference papers
Complete list of metadata
Contributor : Gilles Burel Connect in order to contact the contributor
Submitted on : Monday, May 10, 2021 - 1:50:30 PM
Last modification on : Friday, May 14, 2021 - 9:05:03 AM


  • HAL Id : hal-03222644, version 1



Gilles Burel. Reconnaissance d'objets 3D par réseau d'automates. RFIA, Nov 1991, Lyon-Villeurbanne, France. ⟨hal-03222644⟩



Record views