Deformable 3D Shape Classification Using 3D Racah Moments and Deep Neural Networks
| Autor: | Aissam Berrahou, Abderrahim Mesbah, Abdelmajid El Alami, Zouhir Lakhili, Hassan Qjidaa |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Artificial neural network
Computer science business.industry Property (programming) Cognitive neuroscience of visual object recognition 020206 networking & telecommunications Pattern recognition 02 engineering and technology Image (mathematics) Pattern recognition (psychology) 0202 electrical engineering electronic engineering information engineering General Earth and Planetary Sciences Deep neural networks 020201 artificial intelligence & image processing Artificial intelligence business General Environmental Science |
| Zdroj: | Procedia Computer Science. 148:12-20 |
| ISSN: | 1877-0509 |
| DOI: | 10.1016/j.procs.2019.01.002 |
| Popis: | In this paper, we propose a new model for 3D shape classification based on 3D image Racah moments and deep neural networks to enhance the classification accuracy and reduce the computational complexity of 3D object recognition. The proposed model is derived by introducing 3D image Racah moments as an input vector in deep neural network (DNN), ordinarily utilized in many applications of pattern recognition. Discrete Racah moments have the property to extract pertinent features from an image in lower orders, and with the effectiveness of the DNN, we can make up the proposed model. This work aims to investigate the classification capabilities of the proposed model on non-rigid 3D datasets. Experiment simulations are conducted on SHREC 2011 database to evaluate the performance of our proposed method. The obtained results indicate that the proposed model achieves high performance classification rates. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|