Construction site accident prediction method and device based on image recognition

An accident prediction and image recognition technology, applied in the field of image processing, can solve problems such as failure to investigate the root cause of accidents and failure to predict construction site accidents, and achieve the effect of improving the accuracy rate

Active Publication Date: 2021-01-01
GUANGDONG UNIV OF TECH
View PDF8 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Benefits of technology

The technology described in this patented describes an approach for accurately identifying accidents caused due to various factors such as environmental conditions or workers working under stress during construction work hours. This helps improve efficiency and reduce costs associated therewith.

Problems solved by technology

The technical problem addressed in this patents relates to monitoring and investigating building sites that require frequent maintenance due to their potential hazards or incurrences caused by factors like welders being used at different stages of the project's life cycle. Current methods involve manual checks with human inspectors which can only provide limited data points about specific issues such as poor visibility conditions or lack of knowledge regarding what happened next.

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
  • Construction site accident prediction method and device based on image recognition
  • Construction site accident prediction method and device based on image recognition
  • Construction site accident prediction method and device based on image recognition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0059] Such as figure 1 As shown, this embodiment provides a method for predicting construction site accidents based on image recognition, including the following steps:

[0060] S1: Obtain the video streaming data of the construction site.

[0061] In the embodiment of the present invention, the video stream data can be collected through the cameras on each construction site.

[0062] S2: Use the preset convolutional neural network to identify security equipment on video stream data in real time to obtain security identification results.

[0063] Among them, the safety identification results include recorded information that workers did not wear safety equipment correctly.

[0064] Optionally, the safety equipment includes safety helmets and safety belts; in step S2, identifying the safety equipment on the video stream data may include: identifying whether each worker included in the video stream data is wearing the safety equipment correctly, if any If the worker does not...

Embodiment 2

[0124] Such as figure 2 As shown, this embodiment provides a construction site accident prediction device based on image recognition, including a data acquisition unit 201, a safety identification unit 202, a face analysis unit 203, and a prediction unit 204, wherein:

[0125] A data acquisition unit 201, configured to acquire video stream data of a construction site;

[0126] The safety identification unit 202 is used to use the preset convolutional neural network to perform safety equipment identification on the video stream data in real time to obtain a safety identification result; the safety identification result includes record information that workers do not wear safety equipment correctly;

[0127] The face analysis unit 203 is used to perform face recognition and analysis on the video stream data in real time to obtain face analysis results; the face analysis results include foreign personnel record information, local personnel's working hours record information and ...

Embodiment 3

[0146] This embodiment provides a construction site accident prediction system based on image recognition, which executes Figure 5 The image recognition-based construction site accident prediction method shown. Such as Figure 5 As shown, the construction site accident prediction system includes a safety helmet, a safety belt recognition unit, a face recognition unit, a construction site risk assessment unit, a worker’s working hours judgment unit, and an environmental data collection unit; Figure 5 As shown, construction site accident prediction methods include:

[0147] Firstly, the video stream data is obtained through the cameras of each construction site, and the video stream data are respectively transmitted to the helmet and seat belt identification unit, wherein the helmet and seat belt identification unit receives the video stream data and performs the identification of the helmet and seat belt , and then judge whether the worker is wearing the safety equipment. I...

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 construction site accident prediction method and device based on image recognition, and the method comprises the steps: carrying out the safety equipment recognition of obtained video stream data in real time through a preset convolution neural network, so as to obtain a safety recognition result; performing face recognition and analysis on the video stream data in real time to obtain a face analysis result; inputting the safety identification result, the face analysis result and the climate geographic information of the construction site into a multiple regression equation to obtain an accident occurrence prediction value of the construction site, and identifying a characteristic variable with a relatively large weight in the multiple regression equation as an accident occurrence reason; therefore, the multivariate regression equation can be continuously updated according to the historical accident occurrence numerical value and the historical accident occurrence prediction numerical value, the accident occurrence reason can be identified according to the latest multivariate regression equation, the accident source can be checked, and construction site accident prediction can be carried out from the source.

Description

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

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
Owner GUANGDONG UNIV OF TECH
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