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

A method and device for extracting image features

An image feature and extraction method technology, applied in the field of image processing, can solve the problem of image feature value redundancy and so on

Inactive Publication Date: 2019-08-30
PEKING UNIV +2
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The present invention provides a method and equipment for extracting image features, which are used to solve the problem in the prior art that after multi-scale acquisition of the same image, the obtained image feature values ​​have more redundancy

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
  • A method and device for extracting image features
  • A method and device for extracting image features
  • A method and device for extracting image features

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0073] Such as figure 1 As shown, the method of the image feature extraction method of the embodiment of the present invention comprises:

[0074] Step 101: Determine the central pixel point of the image to be processed; take the central pixel point as the center, perform N times sampling on the image to be processed according to the set sampling scale sequence, and obtain N sampling sets, wherein the sampling scale of each sampling set is incomplete Similarly, N is an integer greater than 1.

[0075] Specifically, the image to be processed may be a complete image, or may be each block diagram after an image is divided into blocks. The gray value of each pixel of the image to be processed is known, centered on the central pixel, and the sampling scale is the radius to sample the area of ​​the central pixel, where the area can be a circular neighborhood or other shapes Neighborhoods are not limited here.

[0076] Further, in order to reduce the redundancy between features, t...

Embodiment 2

[0106] Hereinafter, the present invention will be further described in detail by taking the sampling number N equal to 3 as an example.

[0107] Step 1: Taking the central pixel as the center, sample the image to be processed three times according to the sequence of sampling scales 1, 2 and 3 to obtain 3 sampling sets.

[0108] Specifically, the neighborhood with a scale radius of 1 in the image to be processed is processed as follows: uniform sampling is performed on a circle with the center pixel as the center and a radius of 1, and the number of sampling points is 8, and the sampling set is obtained as

[0109] For the domains with scale radii of 2 and 3 in the image to be processed, the following processing is carried out: Uniform sampling is performed on the circles with the center pixel as the center and the radii of 2 and 3 respectively, and the initial sampling points are 24, and the initial sampling sets are respectively with Then take the average gray value of e...

Embodiment 3

[0129] Correspondingly, an embodiment of the present invention provides an image feature extraction device, such as Figure 7 shown, including:

[0130] Sampling module 701: used to determine the central pixel of the image to be processed; centering on the central pixel, the image to be processed is sampled N times according to the set sampling scale sequence to obtain N sampling sets, wherein the sampling of each sampling set The scales are not exactly the same, and N is an integer greater than 1;

[0131] The first determination module 702: used to determine the first gray value difference sequence of each sampling set in the N sampling sets, the first gray value difference sequence is composed of the gray value of each sampling point in the sampling set and the center The difference between the gray value of the pixel;

[0132] The second determination module 703: is used to perform discrete cosine transform on the first gray value difference sequence of N sampling sets acc...

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 a method and equipment for extracting image features, comprising: taking the center pixel as the center, sampling the image to be processed N times according to a set sampling scale sequence to obtain N sampling sets; determining The first gray value difference sequence of each sampling set in the N sampling sets; performing discrete cosine transform on the first gray value difference sequence of the N sampling sets to obtain the Jth sampling set The second gray value difference sequence of the second gray value difference sequence of the Jth sampling set; the second gray value difference value of each sampling point in the second gray value difference sequence of the Jth sampling set is symbolized to obtain the Jth The sign function value of each sampling point in the Jth sampling set, and determine the local binary pattern eigenvalue of the Jth sampling set. The image feature extraction method proposed by the invention is simple and has high feature recognition rate.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a method and device for extracting image features. Background technique [0002] In computer vision and digital image processing technology, the representation and acquisition of image features is an important basic work. In recent years, the feature extraction method of Local Binary Pattern (LBP) has achieved remarkable results in texture analysis and face recognition applications. The principle of LBP extraction is relatively simple, the computational complexity is low, and it has significant advantages such as rotation invariance and gray invariance, so it is widely used in texture classification, image retrieval, face recognition, target detection and tracking, biological and Medical image analysis, remote sensing image analysis and other fields. [0003] The multi-region histogram sequence of LBP feature spectrum at different scales is used to describe the image texture fea...

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): G06K9/46
CPCG06V10/50
Inventor 胡然李晓龙郭宗明
Owner PEKING UNIV
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