Structured image description method

An image description and structuring technology, applied in special data processing applications, instruments, electrical and digital data processing, etc., can solve the problem of not being able to match the user's retrieval intention well, and achieve the elimination of semantic gaps, better retrieval results, and improved differentiation. degree of effect

Inactive Publication Date: 2014-01-22
TIANJIN UNIV
View PDF3 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

On the other hand, in the field of image retrieval, since the computer uses low-level features to represent an image, the retrieval results given by the computer cannot match the user's retrieval intention very well. The "semantic gap" also requires us to use richer and more precise semantic information to describe an image

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
  • Structured image description method
  • Structured image description method
  • Structured image description method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0026] Here, two images are selected as the images to be described, which are image 3 , Figure 4 For the image on the left in the center, use the method described in the present invention to predict and output a 3-layer tree structure semantic unit.

[0027] First, it is necessary to train the model parameters of the conditional random field (CRF): the specific steps are as follows:

[0028] 1. The steps to construct the training set are as follows:

[0029] (1) Write a crawler program to download the images in the retrieval results of Google Image Search to form an image collection where N d is the total number of images in the collection IMG;

[0030] (2) Use the image segmentation algorithm to segment the objects contained in each image in the set IMG to form an object set where N m is the total number of objects in the collection OBJ, because there may be multiple objects in an image, so N m ≥N d ;

[0031] (3) Use the Amazons Mechanical Turk tool to label eac...

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 belongs to the technical field of image retrieval, in particular to a structured image description method. The structured image description method comprises the steps that an image for training is obtained, and three-layer tree-shaped structure label is established for each object in the image, so that a training set is formed; the bottom-layer characteristics of each object of the image in the training set are extracted, all candidate classes, subclasses and classifiers with corresponding attributes are obtained through training, and therefore intermediate data required for modeling of the next step are formed; a conditional random field model is established and model parameters are obtained through training; image segmentation is firstly conducted, objects contained in an image to be described are segmented, and the bottom-layer characteristics of each object of the image to be described are further extracted; tree-shaped structure label of each object of the image to be described is predicated through the established CRF model and the model parameters obtained through training and according to the maximum product belief propagation algorithm. According to the structured image description method, the distinction degree between images can be improved and a good retrieval result is generated.

Description

Technical field [0001] The invention belongs to the technical field of image retrieval, in particular to a structured image description method. Background technique [0002] Using richer semantic information to describe an image is extremely important for both understanding the image and retrieving the image from the Web. On the one hand, when faced with a new image, the first thing people want to know is which class the object in the image belongs to (such as an animal or a vehicle), and after obtaining its class information, go further One wants to know which subclass it belongs to (is it a bird, or is it a cat), and in addition, each object has its own unique attribute information, such as whether it has feathers, whether it can fly, whether it eats meat, etc. Through this information, people can understand an image more accurately from multiple angles, and at the same time acquire more knowledge about the objects in the image. On the other hand, in the field of image r...

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): G06F17/30G06K9/62
CPCG06F16/5838
Inventor 韦星星韩亚洪操晓春
Owner TIANJIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap