A Probabilistic Neural Network-Based Module for Recognition of Objects from their 3-D Images

A Probabilistic Neural Network-Based Module for Recognition of Objects from their 3-D Images

Onsy A. Abdel Alim (Department of Electrical Engineering, Beirut University, Beirut, Lebanon), Amin Shoukry (Computer Department, Egyptian Japanese University of Science and Technology (EJUST), Alexandria, Egypt), Neamat A. Elboughdadly (Department of Electrical Engineering, Faculty of Engineering, Alexandria University, Alexandria, Egypt) and Gehan Abouelseoud (Department of Engineering, Alexandria University, Alexandria, Egypt)
Copyright: © 2013 |Pages: 14
DOI: 10.4018/ijsda.2013040105
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

In this paper, a pattern recognition module that makes use of 3-D images of objects is presented. The proposed module takes advantage of both the generalization capability of neural networks and the possibility of manipulating 3-D images to generate views at different poses of the object that is to be recognized. This allows the construction of a robust 3-D object recognition module that can find use in various applications including military, biomedical and mine detection applications. The paper proposes an efficient training procedure and decision making strategy for the suggested neural network. Sample results of testing the module on 3-D images of several objects are also included along with an insightful discussion of the implications of the results.
Article Preview

2. Basic Requirements For 3-D Object Recognition

Before proceeding with the description of the details of the proposed module, a discussion of the basic requirements of any object recognition module is presented. An object recognition module should possess the following main characteristics (other attributes can also be important dependent on the particular problem at hand and the type of measured signals to recognize the objects):

  • Invariance to rotation: If the module is presented with an image taken at a slightly different angle, it should still be able to recognize the object correctly;

  • Invariance to translation: If the module is presented with a slightly translated image (i.e. there is a shift in the position of the object), it should still be able to recognize the object correctly;

  • Invariance to slight deformation: To illustrate this point, the example of mine recognition is considered. A mine recognition module is liable to be trained on old images of mines. If these mines that are to be recognized have been buried several years ago, they will be probably hampered by inevitable deformations. The recognition module should be able to cope with this.

3. The Proposed Object Recognition Module

The previous arguments about recognition modules basic requirements motivated the proposal of the object recognition module that is to be detailed in this section.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 6: 4 Issues (2017)
Volume 5: 4 Issues (2016)
Volume 4: 4 Issues (2015)
Volume 3: 4 Issues (2014)
Volume 2: 4 Issues (2013)
Volume 1: 4 Issues (2012)
View Complete Journal Contents Listing