a hidden markov model and neural network face recognition method using the

摘要

this invention is a face recognition system in a hidden markov model and neural network, using their own two t's face on the way to identify, through the discrimination of the party, so i have to take the j's face image from each of the special t beijing vector (Ov) extraction, and then the characteristics of each object in the binary vector to use hidden markov models to the training a small cluster with hidden markov model parameters (u03bb t) is generated and the generation of the receiver; markov model parameter (u03bb i) and the other half of the feature vector (Ov) to use the feature vector (Ov) is the conditional probability of the data to find the conditional probability of the data; the use of the neural network learning.the learning of the neural network through the conditional probability of the other half of the data from each person in the first stage, including the characteristics of.;therefore, the hidden markov model to identify the object in the use of the conditional probability of the feature vector, the first to identify the neural network implementation, the use of the first type are a result of the use of secondary data to identify the implementation. with more accurate face recognition is possible.