Unlock instant, AI-driven research and patent intelligence for your innovation.

A Classification and Recognition Method Based on Subspace Analysis

A classification recognition and subspace technology, applied in the field of classification recognition, can solve the problems that samples are easy to fall into local minimum and long training time, etc., and achieve the effect of fast recognition speed, low hardware requirements and stable performance

Active Publication Date: 2019-08-16
ZHEJIANG UNIVERSITY OF SCIENCE AND TECHNOLOGY
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Classifiers based on supervised learning include support vector machines and neural network classifiers, which require a long training time and are prone to fall into local minimum when the samples are insufficient

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A Classification and Recognition Method Based on Subspace Analysis
  • A Classification and Recognition Method Based on Subspace Analysis
  • A Classification and Recognition Method Based on Subspace Analysis

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0037] The present invention will be further described below in conjunction with specific examples, but the protection scope of the present invention is not limited thereto.

[0038] A new type of classification recognition algorithm based on subspace analysis. For face sample signals, a face classifier is designed:

[0039] make to L j ≤N is determined by the following formula

[0040]

[0041]

[0042] To solve the above problem, define

[0043]

[0044] Notice

[0045]

[0046] therefore

[0047]

[0048] This shows that Ω(l,:) T Belonging to the symmetric vector R j The corresponding set of N normalized feature vectors {±V j (:,n)}, here

[0049] here and It is constrained by rank(Ω j ) = L j , the solution of formula (1) is as follows:

[0050] Ω j (l,:) T =±V j (:, l), l=1, 2, ..., L j ;

[0051] can be designed like this

[0052]

[0053] Here L j Satisfy γL j ≥0 and γL j+1 <0;

[0054] The classification result is determine...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The present invention relates to a novel classification recognition algorithm based on subspace analysis, which refers to the idea of ​​sparsity analysis model and support vector machine, designs J-type classifier combinations, and for each classifier group, projects the matched test samples near the zero point, Unmatched samples are projected as far away from the zero point as possible; and then classified according to the nearest neighbor principle. The present invention uses the subspace analysis classifier for face recognition. In the experiment on the ORL face recognition database, the subspace analysis classifier has shown a classification performance superior to that of the traditional classifier; in the small sample data experiment, it has the ability to identify It has the characteristics of fast speed, high precision, and does not require long-term training; the experimental results fully confirm the feasibility and effectiveness of the subspace analysis classifier.

Description

technical field [0001] The invention relates to a classification recognition method, in particular to a classification recognition method based on subspace analysis. Background technique [0002] With the continuous development of pattern recognition, machine learning, artificial intelligence and other fields and the continuous emergence of new technologies, classification algorithms have been greatly developed. As the last link of pattern recognition, classification recognition is crucial. The work at this stage is to design a classifier, match the image features to be recognized with the training sample image features, and obtain the classification results. In the classifier design of unsupervised learning, the nearest neighbor (NN, NearestNeighbor) idea is most widely used. The main principle is that if most of the most adjacent samples of a sample in the feature space belong to a certain category, then the sample also belongs to this category and has the characteristic...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/62G06K9/00
CPCG06V40/172G06F18/285G06F18/22G06F18/2411
Inventor 于爱华侯北平李刚冯晞张震宇孙勇智
Owner ZHEJIANG UNIVERSITY OF SCIENCE AND TECHNOLOGY