About

ABSTRACT

In the thesis, three (3) computational models is design to recognize an unknown face image. Principal Component Analysis (PCA) also known as eigenfaces is used to extract the feature vectors. In the PCA, Jacobi’s method is utilized to obtain the eigenvalues and eigenvectors. Model A is designed using euclidean distance to measure the distance between feature vectors. Backpropagation neural network is used for other two (2) model namely as B and C to classify the unknown face image. Model B is developed using a single neural network with multiple output neurons. Conversely, model C used multiple neural networks with single output neuron. The performance of these models is tested by using 15 persons which consists 150 face images from Olivetti Research Laboratory (ORL) face database. Results are compared with respects to the effects of changes in face images combination, different normalization technique, number of eigenvectors, recognize of new face images, size of face images and using different combinations of neural network parameter. The experiments result demonstrated model C is the best design model for face recognition with more than 90% of recognition rate, using Improve Linear Scaling (ILS) to normalize the feature vectors. 


Muhammad Firdaus Hashim, Puteh Saad, Mohamed Rizon  and Shahrul Nizam Yaakob, An Approach of Feature Extraction for Face Recognition, Proceeding of Conference on Intelligent Systems and Robotics (iCISAR2005): Malaysia, 2005.

Muhammad Firdaus Hashim, Puteh Saad , Mohamed Rizon , Shahrul Yaakob, Haniza Yazid and M.Karthigayan, Dimension Reduction for Face Recognition using PCA, Proceeding of the Eleventh International Symposium on Artificial Life and Robotics (AROB 11th):Japan,2006.

Muhammad Firdaus Hashim, Puteh Saad , Mohamed Rizon , Shahrul Yaakob , Haniza Yazid and M.Karthigayan, Face Recognition using Neural Network,  Proceeding of International Conference on Intelligent System (ICIS):Malaysia, 2005.

Muhammad Firdaus Hashim, Puteh Saad, Mohamed Rizon Mohamed Juhari  and Shahrul Nizam Yaakob, A Face Recognition System Using Template Matching and Neural Network Classifier, Proceeding of First international Workshop Artificial Life & Robotics 2005(AROB 05): Malaysia, pp. 1-6, May 2005.

Muhammad Firdaus Hashim, Puteh Saad and Mohamed Rizon Mohamed Juhari, Face Recognition using Eigenfaces, The 2nd Regional Conference on Artificial Life and Robotics: Songkhla, Thailand, 14-15 July 2006.


Muhammad Firdaus Hashim, Puteh Saad and Mohamed Rizon Mohamed Juhari, Face Recognition using Eigenfaces, Seminar Penyelidikan KUKUM Pertama, 15 June 2006, KUKUM, Malaysia.

Muhammad Firdaus Hashim, Puteh Saad and Mohamed Rizon Mohamed Juhari, Face Recognition using Eigenfaces and Neural Network, International Conference on Man-Machine System 2006 (ICOMM 2006), 15-16 September 2006, Langkawi-Kedah.

Author: firdaush