Appearance Based Facial Recognition System Using Dhmm with Linear Discriminant Analysis
Islam, Md. Rabiul
Rahman, Md. Fayzur
Khan, M. Abdul Goffar
MetadataShow full item record
Abstract: The aim of this work is to enhance the performance of the face recognition system where appearance based facial features have been used. To locate and extract the facial features from the human face image, Active Shape Model (ASM) has been applied. To reduce the dimension of the feature vector, Linear Discriminant Analysis (LDA) based dimension reduction technique has been used. Finally, these reduced feature vectors are feed to the Discrete Hidden Markov Model (DHMM) using Baum-Welch algorithm and Baum’s Forward-Backward algorithm is used for the recognition model. Extended Yale face database has used to measure the performance of the proposed face recognition system with various dimensions. Finally proposed system performance has been reported according to LDA and existing PCA based facial feature reduction technique.