A HYBRID METHOD BASED ON DYNAMIC COMPENSATORY FUZZY NEURAL NETWORK ALGORITHM FOR FACE RECOGNITION

摘要

The present invention relates to a method for recognizing a face. According to the present invention, provided is a face recognition algorithm in which a dynamic compensatory fuzzy neutral network (DCFNN) and an eigenface-latent dirichlet allocation (LDA) are combined, the face recognition algorithm comprising: a size reducing step of projecting a high dimensional image to a low dimensional image to reduce the size of the high dimensional image; a feature extracting step of extracting features of an object face by using the eigenface algorithm and the LDA algorithm to build a database; an algorithm structuring step of structuring the DCFNN algorithm by using the database extracted in the feature extracting step; an inputting step of inputting the algorithm structure formed in the algorithm structuring step into a face extractor; a processing step of processing the data inputted in the inputting step through the face recognition algorithm; and a realizing step of realizing face recognition through a recognition process of the data extracted from the data processing step.