Face recognition using LDA based generalized half band polynomial wavelet filter bank

Face recognition plays an important role in biometric based person authentication systems. Various algorithms have been proposed in last three decades. This paper presents an efficient face recognition method based on combination of linear discriminant analysis (LDA) and a 2-channel wavelet filter bank. The 2-channel wavelet filter bank is designed using factorization of generalized half band polynomial (GHBP). The proposed method is compared with PCA and LDA based existing wavelet transform methods. Results shows higher recognition rates compared to various existing wavelet filters when combined with LDA.