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

A method for identifying safety wearing conditions

A recognition method and safe technology, applied in the field of image recognition, can solve the problems of false detection and missed detection of image recognition technology, and achieve the effect of accurate judgment

Active Publication Date: 2021-10-26
GUANGZHOU HUAWEI TOMORROW SOFTWARE TECH
View PDF14 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, when the existing image recognition technology is applied in target detection, it is easy to have false detection and missed detection.

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 for identifying safety wearing conditions

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0045] A method for identifying a safe wearing situation, such as figure 1 shown, including steps:

[0046] Step 1. Collect multiple material images wearing safety clothing;

[0047] Step 2: Labeling the material images obtained in Step 1 with standard wearable areas respectively to obtain multiple standard wearable area frame information;

[0048] Step 3, re-clustering the multiple standard wearable area frame information obtained in step 2 to obtain re-clustering grouping data;

[0049] Step 4. Use the deep learning neural network yolov3 algorithm of the darknet framework to train the re-clustered grouping data obtained in step 3 and the standard wearable area frame information obtained in step 2 to obtain the optimal model;

[0050] Step 5. Analyze the video stream data in the collection area to obtain multi-frame images, respectively input the multi-frame images into the optimal model in step 4 to obtain the frame information and object frame information of the safety we...

Embodiment 2

[0068] A method for identifying a safe wearing situation, other features are the same as in Embodiment 1, and also have the following features: Step 5 includes:

[0069] Step 5.1, analyzing the video stream data of the object in the collection area to obtain multiple frames of images;

[0070] Step 5.2. Input multiple frames of images into the optimal model in step 4 to obtain the corresponding class, score and box of the corresponding safety clothing, helmet and object, where class is the category information, and score is the confidence level of the recognition target. Box is the frame information (x, y, w, h) of the recognition target, where x is the x-axis coordinate of the center point of the frame, y is the y-axis coordinate of the center point of the frame, w is the width of the frame, and h is the height of the frame;

[0071] Step 5.3, respectively compare the score with the confidence threshold θ, when the score<θ, it is judged as a false detection target, and delete...

Embodiment 3

[0081] A method for identifying a safe wearing situation, other features are the same as in Embodiment 1, the difference is that in this embodiment α is specifically 0.5, and T 1 for 10 seconds, T 2 for 5 seconds.

[0082] When the duration of the video stream data is less than 10 seconds, record the cumulative duration t of the safety wearing data in step 5 within the 5 second period, and when t≥2.5 seconds, the judgment result is that the object is wearing safety clothing; when t<2.5 The second judgment result is that the object is not wearing safety clothing and an alarm is issued. That is to say, record the situation of personnel wearing safety clothing and helmets within 5 seconds, and generate an alarm when personnel do not wear safety clothing and helmets for more than half of the accumulated time within 5 seconds.

[0083] When the video stream data has the duration of the object frame information for less than 10 seconds, record the cumulative duration t of the safe...

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 safety wearing situation recognition method, through six steps, can obtain the wearing judgment result of an object within a time period according to the collected video stream data. The present invention uses the long-short memory algorithm to activate the long-memory algorithm structure or the short-memory algorithm structure correspondingly according to the wearing situation of the object within a period of time, so as to avoid errors caused by target deformation, sudden movement, background clutter, occlusion, video frame loss, etc. Inspection, missed inspection of safety clothing and helmets. The number of alarms generated by false detection and missed detection of safety clothing and helmets is greatly reduced, and the alarm information is output reasonably, which has the advantage of accurate judgment.

Description

technical field [0001] The invention relates to the technical field of image recognition, in particular to a method for recognizing a safe wearing situation. Background technique [0002] In petrochemical work sites, construction sites, power plants, rail interiors, etc., due to the complex scene environment, there are various factors that threaten personal safety, so the staff in the above occasions are required to wear safety helmets or safety clothing. [0003] Intelligent detection of whether workers wear safety helmets and safety clothing is of great significance to the safety protection management and intelligent information management of construction sites. If it can effectively improve the on-site management efficiency of supervisors on the wearing of helmets and safety clothing, it can greatly reduce the labor cost of manual inspections, and at the same time provide safety guarantees for workers, reducing the occurrence of safety accidents to a certain extent. Howe...

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): G06K9/00G06K9/62G06N3/04G06N3/08
CPCG06N3/049G06N3/08G06V40/10G06V20/52G06N3/045G06F18/23G06F18/241
Inventor 李静王荣秋李朝辉
Owner GUANGZHOU HUAWEI TOMORROW SOFTWARE 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