Face comparison method based on high-dimensional LBP (Local Binary Patterns) and convolutional neural network feature fusion

Abstract

The invention provides a face comparison method based on high-dimensional LBP (Local Binary Patterns) and convolutional neural network feature fusion. The method comprises the following steps: firstly, two types of face images are input, preprocessing is independently carried out, then, each image independently extracts the high-dimensional LBP features and the CNN (Convolutional Neural Network) features of the image, the two features are combined and are subjected to dimensionality reduction via PCA (Principal Component Analysis), and finally, a Joint Bayesian method is used for obtaining a similarity of the two images. In a feature extraction process, since the high-dimensional LBP extracts local information and the CNN extracts global information, two types of information are fused, and the information extracted by the features is complete. Compared with a method which separately uses the high-dimensional LBP or the CNN, the method is higher in accuracy and better in robustness and achieves a real-time face comparison rate.

Description

technical field [0001] The invention provides a face comparison method based on the fusion of LBP features and convolutional neural network features. Background technique [0002] With the rapid development of society and the rapid progress of science and technology, such as access control, video security monitoring, human-computer interaction and other technologies continue to develop, people urgently need an accurate identification method. Traditional identification methods include: keys, identity documents, access cards, etc. However, these identification verification methods are easy to be stolen, forged, or embezzled. These traditional identity verification methods are increasingly unable to meet the needs of society. Face recognition technology just makes up for this gap. At present, face recognition technology has made great progress, the recognition rate is very high, and the speed is also very fast. However, there are still many unsolved problems in this technolo...

AI Technical Summary

Problems solved by technology

[0009] The technical problem solved by the present invention is to overcome the lack of ability to obtain global information based on high-dimensional LBP feature method and the method based on CNN feature to obtain local information, and provide a feature fusion based on high-dimensional LBP and convolutional neural network face comparison method

Images