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

Image semantic complete labeling method based on convolutional neural network and concept lattices

A convolutional neural network and concept lattice technology, applied in the field of image processing, can solve the problems of lack of semantic correlation of tags, cumbersome underlying combination features of images, etc., and achieve the effect of rich tags

Active Publication Date: 2020-04-28
TAIYUAN UNIVERSITY OF SCIENCE AND TECHNOLOGY
View PDF4 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] In view of the above problems, the present invention proposes a method for complete image semantics based on convolutional neural network and concept lattice, so as to solve the existing image labeling problems such as cumbersome bottom layer combination features and lack of label semantic correlation.

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 semantic complete labeling method based on convolutional neural network and concept lattices
  • Image semantic complete labeling method based on convolutional neural network and concept lattices
  • Image semantic complete labeling method based on convolutional neural network and concept lattices

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0066] The present invention will be further described in detail below in conjunction with specific embodiments, but the protection scope of the present invention is not limited to these embodiments, and all changes or equivalent substitutions that do not depart from the concept of the present invention are included within the protection scope of the present invention.

[0067] The present invention is based on the convolutional neural network and the concept lattice image semantic complete labeling method, including using the VGG19 model to carry out the general model pre-training method of the convolutional neural network; extracting the initial labeling words and depth features of the image to be labeled; the concept lattice improves the initial labeling results ; Use the candidate label set to predict four parts of the label, as follows:

[0068] The present invention selects the VGG19 network structure as the pre-training model for the initial labeling of the model. First...

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 semantic complete labeling method based on a convolutional neural network and concept lattices, and the method comprises the following steps: firstly constructing an adaptive CNN network, segmenting a to-be- labeled image, and extracting the features of the to-be- labeled image, so as to obtain a neighbor image set and a series of corresponding label sets; then, using the concept lattice for carrying out potential semantic analysis on the label, so that the labeling effect is effectively improved, and the completeness of semantic labeling is guaranteed; and finally, obtaining an optimal semantic tag by utilizing a voting mode. The reference data set Corel5k is adopted for an experiment, it is verified that the method can effectively enrich image label semantics, the label recall rate is increased, and the image semantic retrieval efficiency is improved.

Description

technical field [0001] This paper invents an image semantic complete annotation method based on convolutional neural network and concept lattice, which belongs to the field of image processing. Background technique [0002] The explosive growth of network image data and the subjectivity and arbitrariness of image labeling have resulted in missing labels and semantic noise for a large number of images, which cannot describe image content well. Moreover, these massive image data usually contain rich semantic content, but the incompleteness of tags has brought great challenges to text-based image retrieval and affected the development of other related industries. In order to enrich the content of image tags and improve the accuracy of image retrieval, many researchers have carried out in-depth research on the complete method of image tags for automatic completion of missing tags, but there are also the following shortcomings: 1) It is necessary to select the underlying features...

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): G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/047G06N3/045G06F18/2415G06F18/241
Inventor 张素兰李雯莉胡立华张继福杨海峰
Owner TAIYUAN UNIVERSITY OF SCIENCE AND TECHNOLOGY