Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Image classification method based on high-constraint high-dispersion principal component analysis network

A principal component analysis and classification method technology, applied in the field of image classification based on highly constrained and highly dispersed principal component analysis network, can solve problems such as computational complexity and time complexity

Active Publication Date: 2017-03-08
EAST CHINA NORMAL UNIV
View PDF4 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Although the deep neural network framework has been successfully applied to some sub-problems, we still need to face some unavoidable problems: computational complexity and time complexity

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
  • Image classification method based on high-constraint high-dispersion principal component analysis network
  • Image classification method based on high-constraint high-dispersion principal component analysis network
  • Image classification method based on high-constraint high-dispersion principal component analysis network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0038] The present invention will be further described in detail in conjunction with the following specific embodiments and accompanying drawings. The process, conditions, experimental methods, etc. for implementing the present invention, except for the content specifically mentioned below, are common knowledge and common knowledge in this field, and the present invention has no special limitation content.

[0039] Such as figure 1 As shown, the image classification method based on the highly constrained and highly dispersed principal component analysis network of the present invention, the input image passes through at least one set of convolutional layers and nonlinear transformation layers, and a feature pooling layer, which specifically includes the following steps:

[0040] Convolution and nonlinear change step: in the convolution layer, a plurality of convolution kernels for each stage feature extraction are learned from the training set by PCA; in the nonlinear transfor...

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 invention discloses an image classification method based on a high-constraint high-dispersion principal component analysis network. The image classification method comprises the following steps of a convolution and nonlinear variation step: in a convolution layer, obtaining a plurality of convolution kernels used for extracting the characteristics of each stage, and in a nonlinear transformation layer, utilizing the convolution kernels to carry out nonlinear transformation on an input image to obtain a characteristic pattern; a characteristic pooling step: in a characteristic pooling layer, importing a multiscale characteristic analysis formula, and outputting characteristics after a value meets a condition that a high-dispersion distribution formula and the optimal value of the scale zooming factor [Sigma] of the high-dispersion distribution formula are optimal is deduced; an integration step: unfolding the characteristics into vectors, and utilizing the vectors to form a characteristic matrix; and an image classification step: inputting the characteristics into a linear support vector machine to finish an image classification task. The image classification method is simple and efficient and exhibits self-adaption and expansibility, and only the structural parameter of the network needs to be input.

Description

technical field [0001] The invention relates to the technical field of pattern recognition for computational image processing, belongs to the category of deep learning in machine learning, and in particular relates to an image classification method based on a highly constrained and highly dispersed principal component analysis network. Background technique [0002] In the field of computer vision and pattern recognition, finding suitable features to represent images is very critical in solving classification problems. For example, the most famous local or global feature description operators (Scale Invariant Feature Transformation SIFT and Histogram of Oriented Gradients HOG) have made great progress in object recognition and matching. Interestingly, many successful feature representations are similar, and can actually be thought of as computing histograms of edge gradients or adding some convolution operations. Although these descriptor operators have good results in extra...

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 Applications(China)
IPC IPC(8): G06K9/62
CPCG06F18/2411G06F18/214
Inventor 孟丹曹桂涛陈伟婷
Owner EAST CHINA NORMAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com