Human face identifying method under small sample condition

Abstract

The invention relates to a face recognition method that suits for the small sample, which uses two-dimensional matrix, to establish sample data matrix, similar member matrix, getting typical correlation matrix and finally generalized inverse matrix. It solves the problem that the queer covariance matrix is added artificial disturbance to process correction, which brings disturbance data introducing more random and uncertainty, making the result of small sample is not satisfied, leading to a significant increase in calculation, discarding the image null space ahead due to the method of pre-treating samples by dimensionality reduction, losing some important distinguishing information. The invention replaces number 1 with a matrix, that is the same size as sample image, to make the similar member matrix not only reflect the affiliation on each sample and any kind, but retain the space information of rows and columns in the sample images.

Description

technical field [0001] The invention relates to a face recognition method, in particular to a face recognition method suitable for small sample conditions. Background technique [0002] Canonical Correlation Analysis (CCA) is a statistical analysis method for analyzing the correlation between two groups of random variables. Its purpose is to find the linear combination of the two groups of random variables so that the two random variables The correlation of the variables is maximized. At present, the CCA method has been widely used in many fields, such as signal processing, medical research and pattern recognition. [0003] Before the present invention, in applications such as image recognition, when the number of image samples is smaller than the sample dimension, the CCA method will face the small sample problem (Small Sample Size, SSS for short) caused by the singularity of the covariance matrix. In response to this problem, researchers have proposed many methods to try...

AI Technical Summary

Problems solved by technology

This type of method essentially uses statistical methods such as PCA to discard the null space of the image first. The main problem and defect is that the above null space is approximated by the eigenvector corresponding to the small eigenvalue, and discarding it may lose some important discriminant information

[0006] Therefore, none of the above methods for solving the small-sample problem is ideal

Images