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

Safety helmet real-time detection method based on convolutional neural network

A convolutional neural network, real-time detection technology, applied in the fields of deep learning, image processing, and computer vision

Active Publication Date: 2019-08-13
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF6 Cites 35 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The present invention creatively combines the respective advantages of YOLOv3 and FaceNet, and solves the problem of accurate end-to-end detection of helmet wearing in real time in video streams

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
  • Safety helmet real-time detection method based on convolutional neural network
  • Safety helmet real-time detection method based on convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0054] The present invention is mainly verified by experiments and actual measurements, and all steps and conclusions are verified correctly on tensorflow1.12.0. The specific implementation steps are as follows:

[0055] Step 1. The video image to be detected is initialized and preprocessed:

[0056] Video image preprocessing, including: initializing the video image to be detected, denoted as X, and the dimension of X denoted as N 0 ×N 0 ′=1920×1080, the number of X is recorded as K=18800; the safety helmet and face in the video image X to be detected are manually marked, and the target position is recorded as P k j =(x k j ,y k j ,w k j , h k j ), k=1,2,…,18800, j=1,2,…,112800, where (x k j ,y k j ) are respectively recorded as the center coordinates of the jth target in the kth image, (w k j , h k j ) are respectively recorded as the width and height of the jth target in the kth image, and N is the number of targets in the kth image; using standard data en...

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 safety helmet real-time detection method based on a convolutional neural network. The method comprises the following steps: decoding a video in a multi-thread I frame extraction mode; realizing end-to-end safety helmet and face real-time detection through a YOLOv3 convolutional neural network algorithm. Provided is a method for judging whether to wear the safety helmet ornot based on the safety helmet and human face real-time detection result. For a person who does not wear the safety helmet, face recognition is realized through a FaceNet algorithm so as to carry outintelligent voice reminding on the person. And a channel pruning and quantifying method is adopted to compress the model, so that the integration of the model on the SoC system is more facilitated. According to the method, the respective advantages of YOLOv3 and FaceNet are combined, the problem of accurate end-to-end detection of the safety helmet worn in real time in the video stream is solved,real-time detection of wearing of the safety helmet in the video stream is realized, and whether a worker wears the safety helmet according to regulations or not can be monitored in real time to reduce the safety risk.

Description

technical field [0001] The invention belongs to the field of deep learning, and relates to the fields of computer vision and image processing. Background technique [0002] In production and business units, safety accidents caused by unsafe behaviors of staff occur from time to time. Unsafe behaviors refer to behaviors that violate the objective laws of production safety and may lead to accidents by employees of production and business units during production operations. They are the direct cause of a large number of accidents. [0003] Through the analysis of national safety production accidents in previous years, more than 90% of the safety accidents were caused by unsafe behaviors or violations of the staff, such as illegal operations, misoperations, fatigue work, and improper wearing of labor protection products. Detecting whether the staff wear safety helmets in accordance with the regulations can effectively reduce the injuries caused by safety accidents such as falli...

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
IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V40/161G06V40/168G06V40/172G06V20/41G06V20/46G06N3/045G06F18/23213G06F18/241
Inventor 韦顺军苏浩周泽南闫敏王琛张晓玲师君
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA
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