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

Personnel change detection method and system based on multi-level apparent characteristic comparison

A technology of appearance characteristics and detection methods, applied in the field of computer vision, can solve the problems of sneaking players, identity authentication difficulties, which have not been solved well, and achieve the effect of good real-time performance and high detection speed

Pending Publication Date: 2022-07-15
青岛根尖智能科技有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] During the game of ball sports, in order to change tactics or deal with other uncontrollable factors, player changes are relatively frequent. However, during player changes, illegal substitutions and stealing players often occur, which seriously affects the spirit of sports competition; based on Therefore, it is an important and challenging task to detect the identity of the players involved in entering or exiting the field when players change.
[0003] The inventors found that compared to the common pedestrian identification and face recognition scenarios, the identity comparison in the event is more challenging. In the case of such restrictions as highly dynamic, occluded or unavailable faces, the robust and accurate The identity authentication is very difficult, and the existing player change detection method or equipment has not been well solved

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
  • Personnel change detection method and system based on multi-level apparent characteristic comparison
  • Personnel change detection method and system based on multi-level apparent characteristic comparison

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] This embodiment provides a personnel change detection method based on multi-level apparent feature comparison, including:

[0034] Get image information of related players;

[0035] According to the image information, extract the player's face features, human skeleton proportion features and global identity features;

[0036] According to the extracted face features, human skeleton scale features and global identity features, as well as the preset attention weight model, the detection results of the changed players are obtained;

[0037] Among them, the attention weight model obtains fusion features after assigning corresponding weights to face features, human skeleton proportion features and global identity features; compares the fusion features with the preset apparent features of registered players to obtain a comparison result .

[0038] First, when acquiring the image information of the relevant players, it can be achieved by acquiring the player images in the ga...

Embodiment 2

[0051] In order to verify the method in Embodiment 1, this embodiment provides a personnel change detection method based on multi-level apparent feature comparison, taking a certain large-scale event image data set as an example, specifically:

[0052] Step S1, data set setting, take a large-scale event image data set as an example, in which the player images have been pre-cropped, a total of 6,000, that is, the player images only contain the body area of ​​the current player; the player image data set is divided into The registered player data set (4,000) and the player data set to be tested (2,000, including both registered and unregistered players) are marked as R and T respectively, of which the R data set contains a total of 20 players;

[0053] Step S2, data preprocessing. Normalize the player images in the registered dataset R in S1 to a uniform size (eg 128 × 64 pixels);

[0054] Step S3, face feature extraction. For the image in step S2, use a face detection method ...

Embodiment 3

[0064] This embodiment provides a personnel change detection system based on multi-level apparent feature comparison, including:

[0065] The data acquisition module is configured to: acquire the image information of the relevant players;

[0066] The feature extraction module is configured to: extract the player's face feature, human skeleton proportion feature and global identity feature according to the image information;

[0067]The detection module is configured to: obtain the detection result of the changed player according to the extracted face feature, the proportion feature of the human skeleton, the global identity feature, and the preset attention weight model;

[0068] Among them, the attention weight model obtains fusion features after assigning corresponding weights to face features, human skeleton proportion features and global identity features; the fusion features are compared with the preset apparent features of registered players to obtain a comparison resul...

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 computer vision, and provides a personnel change detection method and system based on multi-level apparent feature comparison, and the method comprises the steps: extracting the multi-level apparent features of the face feature, the human skeleton proportion feature and the global identity feature of a player during the detection of the change of the player, and carrying out the recognition of the change of the player through an attention weight model; after corresponding weights are distributed to the multi-level apparent features, fusion features are obtained; comparing the fusion feature with a preset apparent feature of a registered player to obtain a comparison result; the method solves the problem of personnel identity in the match, can be compatible with the extreme situation of missing some feature, and has high detection speed and good real-time performance.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a method and system for personnel change detection based on multi-level apparent feature comparison. Background technique [0002] During ball games, players are frequently changed in order to change tactics or deal with other uncontrollable factors. However, during the change of players, there are often illegal substitutions, stealing players and other behaviors, which seriously affect the spirit of sports competition; Therefore, it is an important and challenging task to detect the identity of the players involved in the on-court or off-court when the player changes. [0003] The inventors found that compared with common pedestrian identification and face recognition scenarios, the comparison of people's identities in events is more challenging. Under such restrictions of high dynamics, occlusion or unavailability of faces, it is robust and accurate. The id...

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): G06V40/10G06V40/16G06V40/20G06V20/40G06V10/26G06V10/32G06V10/774G06V10/74G06V10/764G06V10/80G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F18/22G06F18/253G06F18/214
Inventor 王海滨纪文峰李凯乔守良程伟鹏
Owner 青岛根尖智能科技有限公司
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