Large-angle profile recognition method and system based on face recognition

Abstract

The application belongs to the technical field of face recognition, and discloses a large-angle side face recognition method and system based on face recognition, wherein the method comprises the following steps: acquiring a face image, constructing a training data set, then representing pixel modules of the face image as a plurality of feature points, then analyzing coordinates of the feature points to determine the orientation and side angle of the face image, selecting corresponding face images from the training data set to perform perspective transformation, then performing data augmentation on the face images after the perspective transformation, and comparing the face images after the data augmentation with face images of the training data set. Through face image acquisition and feature point analysis, when the angle of the face image does not meet the recognition requirement, the large-angle face can be reconstructed, the face image can be converted into a recognizable type, and thus the face recognition precision is improved.

Images