Research on the 3D face recognition based on multi-class classifier with depth and point cloud data

Human face recognition technology usually takes advantages of two-dimensional or three-dimensional data. Rising from 1980s, three-dimensional face recognition technology soon become one of the headed topic because of its admirable resistance to interference and more information compared with two-dimensional face recognition technology. The new 3D face model standardization algorithm presented in this article provides a solution to transfer the obtained face model to standardized CANDIDE-3 face model. The article also provides a new Bayesian classification model based on multi-class classifier, which could overcome the difficulty that ono-verse-one classifier has a low recognition rate when facing more than two people. The article conduct the comparison experiment based on the provided algorithm. According to the experiment, it could raise the face recognition rate efficiently when applying the standardization algorithm and training model.