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

Unsupervised image retrieval method based on hash coding

A technology of image retrieval and hash coding, which is applied in the direction of still image data retrieval, special data processing applications, instruments, etc., and can solve problems such as ineffectiveness and complexity

Inactive Publication Date: 2017-12-08
SUN YAT SEN UNIV +1
View PDF8 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Therefore, using semantic tags as retrieval criteria cannot achieve ideal results.
In addition, as the scale of the database expands, it becomes more and more complicated to set semantic tags for all pictures. People need a new solution to free workers from the tedious work of setting tags.

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
  • Unsupervised image retrieval method based on hash coding
  • Unsupervised image retrieval method based on hash coding
  • Unsupervised image retrieval method based on hash coding

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0031] Such as figure 1 As shown, the present invention is an unsupervised image retrieval method based on hash coding. Since it is learned through video, the video must first be obtained. Use categories (categories are the types of pictures contained in the public library, which types of pictures are known in each library) plus some adjectives as keywords to search on youtube, such as mini bus, Asian female, green plant, etc., for each The category downloads 100 videos, and then captures a frame of pictures every 5 seconds in the video, until each category contains about 900 pictures, and finally gathers the pictures of each category together to form a picture library. Note that each picture in the image library has two labels i and j, representing that it is taken from the j-th video in the i-th category. The types of videos and the number of captured pictures are shown in Table 1:

[0032] Table 1. The number of pictures of each category extracted from the video

[0033]...

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 provides an unsupervised image retrieval method based on hash coding. The method comprises the steps of processing videos to obtain image library data; then training a classifier on an image library and selecting out first half of good-quality images by use of the classifier on a training set; retraining a new classifier by use of the selected-out images and performing screening on the image library by use of the new classifier; training the classifier by use of the images selected out from the image library again; and carrying out operations in the same manner, thereby obtaining a retrieval model by alternate iterative training of the classifier.

Description

technical field [0001] The invention relates to the field of digital image processing, and more specifically, to an unsupervised image retrieval method based on hash coding. Background technique [0002] In recent years, with the development of computer network, multimedia technology and digital image equipment, digital image technology has penetrated into all aspects of society, such as mass media, military aerospace, family life, etc. In this context, image databases are becoming more and more large-scale, and how to quickly retrieve images in huge databases has become an urgent and critical issue. The commonly used technology for early image retrieval is semantic labeling, but there is no clear standard for semantic labeling, that is, different people have different understandings of the same picture and choose different semantic labels. Therefore, using semantic tags as retrieval criteria cannot achieve ideal results. In addition, as the scale of the database expands, ...

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
IPC IPC(8): G06F17/30G06K9/62
CPCG06F16/50G06F18/214
Inventor 张熙杨伟伟赖韩江印鉴高静
Owner SUN YAT SEN UNIV
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com