Face identification method and system

Abstract

The invention discloses a face identification method, which comprises the following steps of: obtaining a classifier model by carrying out random dimension reduction on testing samples and training samples, generating a training set and a testing set of the similarity learning, selecting a regular parameter and a Gaussian kernel function of a support vector machine and inputting the training set of the similarity learning into the regular parameter and the Gaussian kernel function; then inputting the testing set of the similarity learning into the classifier model to obtain classification results; obtaining a sum of the classification results, using the quotient of the sum and the sample number of a certain type of samples as a value of the similarity probability of the type, obtaining the maximum value and outputting the maximum value to obtain the value of the similarity probability so as to obtain the most accurate face identification result. By dimension reduction of the samples, the complexity of the samples is reduced, so that an algorithm of learning the similarity between face images on the basis of the SVM (Support Vector Machine) is rapid; and moreover, by an algorithm of carrying out processing on each type, the face identification rate is correspondingly improved.

Description

technical field [0001] The present invention relates to the technical field of biometric identification, and more specifically, relates to a face recognition method and system. Background technique [0002] In the past few decades, face recognition has developed into a very popular research topic in computer vision, and it is also one of the most successful applications in the field of image analysis. Today, the research on face recognition has great practical significance. Once the research is successful and put into application, it will produce huge social and economic benefits. In the research algorithm of face recognition, it is mainly divided into two categories. One is an image-based face recognition algorithm, and the other is an image-based face recognition algorithm. The image-based face recognition algorithm started earlier, and the technology is relatively mature; the image-based face recognition algorithm is more difficult than the image-based face recognition ...

AI Technical Summary

Problems solved by technology

[0005] In view of this, the present invention provides a method and system for face recognition to overcome the problem of increased computational complexity and difficulty in similarity learning due to the very high sample complexity of the difference space method in the prior art

Images