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

Image Classification Method Based on Supervised Shared Component Topic Model

A topic model and classification method technology, applied in the field of image processing, can solve the problems of ignoring topic relevance and increasing the time complexity of the classification method, so as to overcome the poor representation effect, improve the classification accuracy rate, and achieve the effect of good representation effect.

Active Publication Date: 2017-03-01
XIDIAN UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although the method disclosed in this patent application increases the number of dictionaries, improves the ability of the dictionary to represent images, and improves the accuracy of classification, it still has the disadvantage that the multi-scale dictionary increases the time complexity of the classification method. and ignores the correlation between topics

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 Supervised Shared Component Topic Model
  • Image Classification Method Based on Supervised Shared Component Topic Model
  • Image Classification Method Based on Supervised Shared Component Topic Model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0038] The present invention will be further described below in conjunction with the drawings.

[0039] Attached figure 1 , The specific steps for implementing the present invention are described as follows:

[0040] Step 1. Establish natural image training set and natural image test set.

[0041] The present invention uses an international standard natural image category library, which contains 13 image categories. From each image category, 100 images are randomly selected, and the selected images are formed into a natural image training set. The images remaining after the natural image training set is selected from the international standard natural image category library form the natural image test set. In the embodiment of the present invention, the images of each category in the international standard natural image category library are as figure 2 Shown. figure 2 in figure 2 (a) is an image of a suburban villa, figure 2 (b) is the coast image, figure 2 (c) is the forest i...

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 supervised shared component topic model, which mainly solves the problems in the prior art that the number of parameters is large, the correlation between topics and latent semantic features are ignored, and the image representation effect is poor. The implementation steps are: establish a natural image training set and a natural image test set; generate a visual dictionary; generate an image sparse representation vector; generate a subject distribution vector; construct a natural image classification model; and perform natural image classification. The present invention adopts the Gibbs sampling method and the topic unbalanced prior probability method, reduces the number of parameters, increases the correlation between the topics, and the topic distribution vector of the image has a better representation effect on the image, and improves the accuracy of image classification. Rate.

Description

Technical field [0001] The present invention belongs to the field of image processing technology, and further relates to an image classification method based on a supervised shared component topic model in the field of image classification technology. The invention can be used for target recognition and detection, vehicle navigation, and diagnosis of medical diseases. Background technique [0002] At present, natural image classification has become a very important research subject in the field of image processing technology. Natural image classification has a wide range of applications, such as target recognition and detection, vehicle navigation, and medical disease diagnosis. Due to different lighting conditions, shooting angles and other conditions, there will be certain differences in natural image categories, and due to the lack of image feature extraction methods, there will be a certain degree of consistency between natural image categories. Big challenge. [0003] Recen...

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): G06F17/30G06K9/62
CPCG06V30/194G06F18/2411G06F18/24
Inventor 王爽焦李成陈阳平霍丽娜侯彪马文萍马晶晶张雪
Owner XIDIAN 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