This paper presents an efficient face recognition system based on Hidden Markov
Model (HMM) and the simplest type “Haar” of the Discrete Wavelet Transform
(DWT). The one dimensional ergodic HMM with Gaussian outputs, which represent
the simplest and robust type of HMM, is used in the proposed work. A novel method
is introduced for selecting the training images implemented by choosing the images
that have the odd identifying numbers from the database. Some of these images are
replaced according to the trial-and-error results. The proposed work achieves the
maximum recognition rate (100%), where the experiments are carried out on the ORL
face database.