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

Digital image marking method based on higher-order graph structure p-Laplacian sparse codes

A digital image and marking method technology, applied in the direction of character and pattern recognition, instruments, computer components, etc., can solve unfavorable image marking, does not take into account the high-order graph structure information of image sample distribution, and cannot represent image samples more accurately inner connection

Active Publication Date: 2016-01-20
CHINA UNIV OF PETROLEUM (EAST CHINA)
View PDF3 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Since the image representation method in the current image labeling technology does not take into account the high-order graph structure information of the image sample distribution, it cannot express the internal relationship of the image sample more accurately, which is not conducive to effective image labeling

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
  • Digital image marking method based on higher-order graph structure p-Laplacian sparse codes
  • Digital image marking method based on higher-order graph structure p-Laplacian sparse codes
  • Digital image marking method based on higher-order graph structure p-Laplacian sparse codes

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0045] Such as figure 1 As shown, the digital image storage device stores the digital images to be marked. In addition, there is a marked digital image library. The image library contains some marked digital images, and each digital image corresponds to a set of artificial images. Annotated concept tags, users can also add unmarked digital images that do not exist in the image library. The classic method of digital image processing can be used to generate appropriate image features, such as SIFT features, etc., whereby each image can be represented by a feature vector. After the feature representation of the image is obtained, the preset classification method is used to train the corresponding prediction model, and the image to be labeled in the image storage device and the new image input by the user are labeled based on the prediction model.

[0046] The methods involved in the present invention are as figure 2 Shown. Step 10 is the initial action. Step 11: Generate image f...

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 relates to a digital image marking method based on higher-order graph structure p-Laplacian sparse codes. The method comprises steps that, (1), characteristics of images including marked images, non-marked images and user input images are extracted; (2), higher-order graph structure p-Laplacian information of image samples including the marked images and the non-marked images in an image database are calculated; (3), sparse codes of the images are calculated based on p-Laplacian; (4), a prediction model is acquired based on learning the sparse codes of the images; (5), the non-marked images and the user input images are marked based on the prediction model; and (6), the marking process ends. The method is characterized in that, the image characteristics are expressed mainly, image sample distribution higher-order graph structure information is referred, inherent relationship of the image samples is more precisely expressed, and the images are more effectively marked.

Description

Technical field [0001] The present invention designs a digital image marking method, and particularly relates to a digital image marking method based on p-Laplacian, which is the structure information of a higher-order graph. Background technique [0002] With the continuous improvement of computer data processing capabilities and the popularization of portable smart devices (such as smart phones and digital cameras), it has become easier to acquire digital images with large amounts of data. Digital image marking technology is to use the existing marked image information to mark the unmarked images in the image database. An effective image labeling strategy is to treat the image labeling process as a learning process, use existing training images as samples required for learning, and use machine learning technology to learn the representation method of the image, and then obtain a predictive model to achieve the unlabeled The image tag. [0003] In the current image labeling tech...

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): G06K9/62
CPCG06F18/2411
Inventor 刘伟锋
Owner CHINA UNIV OF PETROLEUM (EAST CHINA)