Artificial neural networks-based face recognition

Dissertant

Hadhood, Sabah Finjan

Thesis advisor

Abbas, Iyad Ibrahim

University

University of Technology

Faculty

-

Department

Electronic Engineering Department

University Country

Iraq

Degree

Master

Degree Date

2011

English Abstract

-Face recognition plays a major role in several applications, such as security, access to restricted areas, banking identity verification and recognition of wanted people at airport.

This work implements the face recognition system in two stages, the first stage is feature extraction stage and the second stage is the classification stage.

The feature extraction stage consists of Discrete Cosine Transform (DCT) and the Principle Component Analysis (PCA).

The DCT transform expresses the image by few coefficients and thus image compaction and hence simpler neural network structure .

The PCA is used for dimensionality reduction by finding patterns in data, and expressing the data in such a way to highlight their similarities and differences.

The second stage is the artificial neural network and it is used for input image classification by using the extracted features from the previous stage to find the similar image to the input image.

The research work implements a system that is capable of successful face recognition in the presence of image rotation, occlusion, and change in facial expression.

The proposed method is implemented using MATLAB 7.10.0.499 and Olivetti Research Laboratory (ORL) face images database, 40 face images are used for the training set and 8 images per person, 10 face images for test set with 2 images per person.

Three types of neural networks are used in this work as follows: 1.

Feed-forward Neural Network with Back-Propagation Algorithm.

2.

Probabilistic Neural Network (PNN).

3.

Pattern Recognition Neural Network.

The results obtained of face recognition system are good for the three types of neural networks and it is around 90%.

Main Subjects

Information Technology and Computer Science

Topics

• Neural networks(Computer science)

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-304968