Learner Pose Recognition Method

A recognition method, a learner's technology, applied in character and pattern recognition, biological models, instruments, etc., which can solve the problems of information redundancy and inability to fully extract information.

Active Publication Date: 2020-07-28
SHAANXI NORMAL UNIV
View PDF5 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The above pose recognition model uses depth image or image difference method to extract the human target in the target extraction stage; in the feature extraction stage, the Hu moment is extracted as the feature vector, and the Hu moment cannot completely extract the information in the image, and they are non-normal handed over, with information 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
  • Learner Pose Recognition Method
  • Learner Pose Recognition Method
  • Learner Pose Recognition Method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0092] Taking the postures of the learner raising his hand, sitting upright, and bowing his head as examples, the steps of the learner posture recognition method are as follows:

[0093] (1) Separation of portrait and background

[0094] The bee colony algorithm that introduces Levi's flight and dynamic weights optimizes the normalized cut, segments the learner's pose image, obtains the binary image of the original image, and realizes the separation of the portrait from the background. The steps are as follows:

[0095] 1) Perform fuzzy C-means clustering preprocessing on the red, green, and blue color channels of the color image respectively, and divide the image into n blocks of maximum similarity areas. In this embodiment, n is 14, and each area is selected on the three color channels The average value of the gray value represents the pixel value of this area, and the undirected weighted graph G=(V, E) is constructed with the pixel values ​​of all areas, V is a vertex in th...

Embodiment 2

[0161] Taking the postures of the learner raising his hand, sitting upright, and bowing his head as examples, the steps of the learner posture recognition method are as follows:

[0162] (1) Separation of portrait and background

[0163] The bee colony algorithm that introduces Levi's flight and dynamic weights optimizes the normalized cut, segments the learner's pose image, obtains the binary image of the original image, and realizes the separation of the portrait from the background. The steps are as follows:

[0164] 1) Perform fuzzy C-means clustering preprocessing on the three color channels of red, green, and blue respectively, divide the image into n blocks of maximum similarity areas, n is 6, and take the gray value of each area on the three color channels The average value of represents the pixel value of this area, and the undirected weighted graph G=(V,E) is constructed with the pixel values ​​of all areas, V is a vertex in the graph, and E is an edge connecting two...

Embodiment 3

[0193] Taking the postures of the learner raising his hand, sitting upright, and bowing his head as examples, the steps of the learner posture recognition method are as follows:

[0194] (2) Separation of portrait and background

[0195] The bee colony algorithm that introduces Levi's flight and dynamic weights optimizes the normalized cut, segments the learner's pose image, obtains the binary image of the original image, and realizes the separation of the portrait from the background. The steps are as follows:

[0196] 1) Perform fuzzy C-means clustering preprocessing on the red, green, and blue color channels of the color image respectively, divide the image into n blocks of maximum similarity areas, n is 20, and take the gray value of each area on the three color channels The average value of represents the pixel value of this area, and the undirected weighted graph G=(V,E) is constructed with the pixel values ​​of all areas, V is a vertex in the graph, and E is an edge con...

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

A learner gesture recognition method, which consists of separating the portrait from the background, extracting the outline image of the learner by using mathematical morphological operations on the binarized image, using Zernike moments for feature extraction, and using support vector machines for feature vectors. Training and recognition of the pose composition of the learner. The present invention introduces Levi's flight mechanism into the bee colony algorithm, and adopts different search methods according to different flight steps, which can enrich the diversity of the population and avoid premature convergence and fall into local optimum; and adopt a dynamic weight mechanism in the bee colony algorithm , according to the evolution rate of the population, the individual bee search mode is adjusted, and the global search and local search capabilities are dynamically balanced. It has the advantages of good segmentation effect and high recognition rate, and can be used for learner gesture recognition and other image recognition and classification.

Description

technical field [0001] The invention relates to the technical fields of image processing and machine vision, in particular to a learner gesture recognition method based on bee colony algorithm optimized normalized cut. Background technique [0002] With the development of network technology and the advent of the digital age, online learning, as a convenient and novel way of learning, has become a dominant form of distance education, which is increasingly affecting our study and work. The learner's body posture partly reflects the learner's learning state. Learner gesture recognition can effectively evaluate the learning status of learners in the online learning process, so that teachers can get more feedback information, and it plays an important role in analyzing the learning status of learners and improving the teaching process. [0003] Zhang Hongyu and others proposed a multi-learner pose recognition method based on depth images. Firstly, the infrared sensor of Kinect i...

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): G06K9/62G06N3/00
CPCG06N3/006G06F18/2411G06F18/214
Inventor 郭敏邝毓茜马苗陈昱莅郭宗华
Owner SHAANXI NORMAL 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