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

An Image Representation Method and Its Application in Image Matching and Recognition

An image representation and image technology, applied in the field of image processing, can solve the problems of large differences in description effects and poor feature robustness, and achieve the effect of improving the accuracy of quantity and matching, improving the accuracy of matching, and enriching image features

Active Publication Date: 2017-11-07
SHENZHEN INST OF ADVANCED TECH
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Most of the image feature extraction methods in the prior art revolve around the color, texture, shape and spatial relationship of the image. Although these methods can describe the features of the image, they have the disadvantage of poor robustness of the obtained features.
There is also an existing technology based on the invariant feature description method of the transform domain. Due to the inherent properties of the transform domain, the description effect of this prior art is quite different before and after the image is clipped.

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
  • An Image Representation Method and Its Application in Image Matching and Recognition
  • An Image Representation Method and Its Application in Image Matching and Recognition
  • An Image Representation Method and Its Application in Image Matching and Recognition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0052] Embodiment 1: Image representation method based on phase singularity

[0053] Such as figure 1 As shown, the flow of the image representation method of this application is as follows:

[0054] S1-1 Read image information

[0055] In this step, the two-dimensional image equation I (x, y) is obtained by reading the image information;

[0056] S1-2 Two-dimensional complex filter processing

[0057] Combine the two-dimensional image equation I(x, y) with the two-dimensional complex filter T(x, y)=T r (x,y)+iT l (x, y) is convolved to obtain a complex image:

[0058]

[0059] Among them, * is the convolution calculation symbol, (x, y) indicates the position of the image pixel.

[0060] The two-dimensional complex filter used in this embodiment is a Laguerre-Gaussian (LG) complex filter:

[0061]

[0062] Tr(x,y)=Re{LG(x,y,σ)}=G x (x, y, σ)

[0063]Ti(x,y)=Im{LG(x,y,σ)}=G y (x, y, σ)

[0064] In the formula, σ is the scale parameter factor. Two-dimensional co...

Embodiment 2

[0104] Embodiment 2: Image matching based on phase singularity

[0105] S2-1 read image I and image I' information

[0106] Read the two-dimensional image equations I(x, y) and I'(x, y) of image I and image I' respectively;

[0107] S2-2 Perform two-dimensional complex filter processing on image I and image I'

[0108] I (x, y) and I' (x, y) are carried out respectively according to the method of S1-2 in the embodiment 1 complex number filtering process, obtain:

[0109] The complex image equation of image I is

[0110]

[0111]

[0112] The complex image equation of image I' is

[0113] ;

[0114]

[0115] S2-3 Phase singular point detection of image I and image I'

[0116] According to the method for S1-3 phase singularity detection among the embodiment 1, image I and image I ' are carried out phase singularity detection, the collection of the phase singularity detected in image I is {P i}, the set of phase singularities detected in image I’ is {P i '};

...

Embodiment 3

[0135] Embodiment 3: Image recognition based on phase singularity

[0136] S3-1 According to the category attributes of the samples, use the SVM classifier to perform classifier training for different types of samples, and obtain classifiers for different types of images

[0137] Creation of S3-1-1 samples

[0138] The training samples are divided into positive samples and negative samples, where the positive sample is the target sample to be detected, such as training a car image classifier, the positive sample is a car; the negative sample refers to any other picture other than the positive sample picture .

[0139] S3-1-2 Training classifier

[0140] According to the method in embodiment 1, all the pictures in the sample are detected and positioned at the phase singularity;

[0141] According to the method in embodiment 1, descriptors are generated for all phase singularities extracted;

[0142] Selection of kernel function:

[0143] A kernel function is selected to de...

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 present invention provides an image representation method based on phase singularity, which includes: performing two-dimensional complex filter processing on image information to generate a filter response image; detecting image phase singularity; accurately locating phase singularity; generating a localized phase singularity Descriptors serve as image features. The present invention also provides a method for image matching and recognition using the above image representation. The image representation method of the present invention effectively utilizes the information of phase singularity, develops a new image representation method, enriches image features, improves the number of image matching points and the accuracy of matching, and develops a method based on phase singularity packet representation Image classification methods.

Description

technical field [0001] The invention relates to image processing technology, in particular to an image representation method, and a method for image matching and recognition using the image representation. Background technique [0002] The fast and accurate description of image information has always been an important and difficult point in the research of image matching and image recognition technology. Most of the image feature extraction methods in the prior art revolve around the color, texture, shape and spatial relationship of the image. Although these methods can describe the features of the image, they have the disadvantage of poor robustness of the obtained features. There is another existing technology based on an invariant feature description method in the transform domain. Due to the inherent properties of the transform domain, the description effect of this prior art is quite different before and after image clipping. [0003] Another existing technology uses t...

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): G06K9/64G06K9/66
Inventor 乔宇王星星李志锋汤晓鸥
Owner SHENZHEN INST OF ADVANCED TECH
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