Low-rank discriminant feature subspace learning method

Abstract

The invention discloses a low-rank discriminant feature subspace learning method, and belongs to the field of image classification. The technical problem of ignoring information of a low-dimensional subspace hidden in a sample is solved. The method comprises the steps of defining an objective function of a discriminant feature learning formula; reformulating The class label is adopted as supervision information, and an objective function; Applying an orthogonal constraint to the feature subspace in the objective function; Dividing one image data set into a test set and a training set; Solvingthe value of each variable when the objective function value is minimized through the training set; Obtaining a feature subspace after the objective function is solved; Obtaining all features of all types of images in a data set through the feature subspace projection test set, and finally obtaining the recognition rate of the data set through a classifier; According to the method, the low-rank representation coefficient is used as the constraint to construct the discrimination item for feature learning, the subspace structure similarity constraint can be introduced into the discrimination feature learning model suitable for image recognition and classification tasks, and model adaptability and robustness are promoted.

Description

technical field [0001] The invention belongs to the field of image classification, in particular to a low-rank discriminant feature subspace learning method. Background technique [0002] Eigensubspace learning plays an important role in pattern recognition, and many efforts have been made to produce more discriminative learned models. In recent years, many discriminative feature learning methods based on representation models have been proposed, which have not only attracted much attention, but also been successfully applied in practical work. However, the discriminative models built by these methods only rely on the category labels of training samples, ignoring the consideration of the basic subspace structure information hidden in them. Contents of the invention [0003] The present invention overcomes the deficiencies of the above-mentioned prior art, and provides a low-rank discriminant feature subspace learning method, which uses low-rank constraints to construct di...

AI Technical Summary

Problems solved by technology

However, the discriminative models constructed by these methods only rely on the category labels of training samples, ignoring the consideration of the basic subspace structure information hidden in them.

Images