Search

CN-122024121-A - Intelligent identification method and system for safety risk of construction site based on image identification

CN122024121ACN 122024121 ACN122024121 ACN 122024121ACN-122024121-A

Abstract

The invention relates to the technical field of construction site risk identification, in particular to an intelligent construction site safety risk identification method and system based on image identification. According to the method, firstly, the distribution positions and construction types of a plurality of construction areas are identified according to a construction plan of a building site, then a monitoring camera is configured for each construction area, the wearing state of protective equipment of operators in the current area is monitored in real time, and in the construction area of the building site, the construction area which is directly communicated with other plurality of construction areas is marked as a comprehensive monitoring area. By taking a site image, on the one hand, the position data and construction type of the site area are identified, and on the other hand, the work information and the pre-preparation personnel are determined according to the route where the worker is located. The identification mode of the invention is mainly carried out through the comprehensive monitoring area, so that the prior identification of the operator can be finished when the operator does not reach the target construction area yet.

Inventors

  • JIANG ZHAOFEI
  • YAN LING

Assignees

  • 青岛恒星科技学院

Dates

Publication Date
20260512
Application Date
20251218

Claims (9)

  1. 1. The intelligent identification method for the safety risk of the construction site based on the image identification is characterized by comprising the following steps: s1, identifying the distribution positions and construction types of a plurality of construction areas according to a construction plan of a building site; S2, configuring a monitoring camera for each construction area, and monitoring the wearing state of protective equipment of operators in the current area in real time; s3, marking a construction area which is directly communicated with other construction areas as a comprehensive monitoring area in the construction area of the construction site; S4, according to the information recorded by the working image, carrying out target area tendency analysis on the operators passing through the comprehensive monitoring area, and matching one or more pre-preparation staff as auxiliary verification, wherein the content of the target area tendency analysis is as follows: Primarily predicting the forward construction area of the operator according to the movement direction, the speed change and the track vector of the operator; Analyzing the corresponding states of the construction type of the forward construction area and the objects on the worker, and marking the primarily predicted forward construction area as a target area of the worker when the construction type of the forward construction area and the corresponding states of the objects on the worker correspond to each other; S5, identifying a target construction area of the operator according to the position information of the early-stage preparation personnel, and taking the preset construction type of the target construction area as a basis for identifying the safety risk of the operator.
  2. 2. The intelligent identification method for the safety risk of the construction site based on the image identification of claim 1 is characterized in that the method for identifying the distribution positions of a plurality of construction areas in S1 is as follows: s1.1, acquiring a construction plan of a building site, and analyzing the position and construction type of each construction area; s1.2, performing size normalization and color correction processing on an image of the work station; S1.3, identifying the position of an operator in a construction site image and an object on the body by using a target detection model, and extracting skeleton key points of similar operators by matching with a human body posture estimation model to obtain a human body posture; S1.4, carrying out similarity recognition on the human body gestures, marking the region where the operators are located as a verification region when the number of operators making similar gestures is larger than a preset threshold value, and sending a prompt to the operators when the construction region is inconsistent with the verification region.
  3. 3. The intelligent identification method for the safety risk of the construction site based on the image identification of claim 1 is characterized in that the method for analyzing the construction type of each construction area in S1.1 is as follows: s1.1.1, analyzing application scenes of objects on an operator through industrial big data, and determining construction types of construction areas according to the application scenes; s1.1.2, analyzing the construction type obtained in the S1.1.1 and the construction type in the S1.1, and when the construction type and the construction type are inconsistent, reminding the staff.
  4. 4. The intelligent identification method for the safety risk of the construction site based on the image identification of claim 1, wherein the step S2 comprises the following steps: the wearing state of the protective equipment of the operator is identified through the image information of the construction area, and when the protective equipment of the operator is missing or inconsistent with the protective equipment required by the construction type, the current state of the operator is identified as risk, otherwise, the current state of the operator is identified as safety.
  5. 5. The intelligent identification method for the safety risk of the construction site based on the image identification of claim 1, wherein the method for marking the comprehensive monitoring area in the S3 is characterized by comprising the following steps: s3.1, identifying the position data and the route of each construction area from the construction site image through a computer vision technology ; S3.2, obtaining the number of independent routes of one construction area directly going to at least two other construction areas When the number of routes And (3) marking the construction area as a comprehensive monitoring area when the construction area is more than or equal to 2.
  6. 6. The intelligent recognition method for the safety risk of the construction site based on the image recognition of claim 5, wherein the method is characterized in that the route is recognized from the image of the construction site The method of (2) is as follows: s3.1.1, based on the continuous pictures of the monitoring camera, identifying continuous position points of operators by adopting a multi-target tracking technology; s3.1.2 connecting two adjacent position points to generate a route 。
  7. 7. The intelligent identification method for the safety risk of the construction site based on the image identification of claim 1, wherein the method for matching the pre-preparation personnel in the step S4 is as follows: S4.1, identifying the same route An operator in the house; s4.2, according to the route Generating work type information of operators according to the construction type of the construction area corresponding to the terminal point; S4.3, obtaining distance route The operator whose end point is the nearest and is marked as a pre-readiness person by personnel identity data or a specific readiness tool carried by the operator.
  8. 8. The intelligent identification method for the safety risk of the construction site based on the image identification of claim 7, wherein the step S5 comprises the following steps: s5.1, acquiring a monitoring picture in the comprehensive monitoring area, and drawing a construction area and a safety area in the monitoring picture; S5.2, immediately acquiring a target area tendency analysis result of the personnel passing through the safety area, performing risk preliminary judgment, and then continuously acquiring movement information of the early-stage preparation personnel and the operation personnel; when the preparation personnel in the current period confirms to enter a certain target construction area: taking the construction type of the target construction area as a final confirmation basis, and performing final security risk assessment on operators still on the way; If the target construction area which is entered by the early-stage preparation personnel is inconsistent with the tool or work information carried by the target construction area, an alarm is sent out for reminding; if the pre-preparer is not identified or does not first confirm the target area: And adopting a target area tendency analysis result of the operator as a basis for risk assessment.
  9. 9. A system for the intelligent recognition method for the safety risk of the construction site based on the image recognition according to any one of claims 1 to 8, which is characterized by comprising an image acquisition module, a risk recognition module, a region management module, a personnel matching module and a comprehensive decision recognition module, wherein: the image acquisition module is used for identifying the distribution positions and construction types of a plurality of construction areas according to the construction plan of the building site; the risk identification module monitors the wearing state of protective equipment of operators in a construction area in real time; The area management module is used for marking a construction area which is directly communicated with other multiple construction areas as a comprehensive monitoring area; the personnel matching module matches one or more early preparation personnel for the operators passing through the comprehensive monitoring area according to the information recorded by the working image; the comprehensive decision-making identification module identifies a target construction area of the operator according to the position information of the early-stage preparation personnel, and takes the preset construction type of the target construction area as a basis for identifying the safety risk of the operator.

Description

Intelligent identification method and system for safety risk of construction site based on image identification Technical Field The invention relates to the technical field of construction site risk identification, in particular to an intelligent construction site safety risk identification method and system based on image identification. Background With the transition of the building industry to intelligent, intelligent construction sites have become an important direction for promoting construction safety modernization. Under the background, the intelligent safety risk identification is used as a core link thereof, and aims to realize early detection and early warning of potential safety hazards through real-time monitoring and intelligent analysis of various risk factors on a construction site, thereby effectively restraining accidents. The traditional safety supervision means mainly depend on modes such as manual inspection, video monitoring and sensor monitoring, however, the methods have obvious shortboards that the manual inspection is limited by subjective experience, the coverage range is limited, the response is delayed, the video monitoring is mainly used for retrospective backtracking, the active recognition and real-time analysis capability is lacking, the sensor monitoring is often limited to single parameter sensing, the deployment and maintenance cost is high, and the cross-system linkage is difficult to realize. The method is influenced by factors such as information island, environmental dynamic change, insufficient intelligent level, artificial omission and the like, and the traditional method is difficult to realize full coverage and accurate identification of safety risks. Although the current single-frame image recognition technology based on a fixed camera can detect a part of explicit risks, the technology still has serious defects in the aspect of risk prejudgment. On one hand, each camera usually only covers a construction area, a monitoring blind area is formed by a moving path (more temporary paths of a building site) among the areas, a system cannot effectively analyze a moving track of an operator and a target area, on the other hand, the construction area can constantly adjust a construction type, but the construction type is uncertainty in adjustment due to the influence of the construction progress, a constructor and monitoring are not usually in a system, and the difficulty of data intercommunication between the constructor and the monitoring is high. Thus, the construction type and the moving path are uncertain, so that the target construction area of the operator cannot be determined, and at the moment, whether the protection of the operator meets the requirement of the construction type is difficult to identify. Disclosure of Invention The invention aims to provide an intelligent identification method and system for safety risks of a construction site based on image identification, which are used for analyzing a target area and a construction type of an operator when the operator passes through a construction area, so that the problems in the background technology, namely the problem that the single-frame image identification technology based on a fixed camera lacks in risk pre-judgment, are solved. In order to achieve the above object, one of the objects of the present invention is to provide an intelligent identification method for safety risk of construction sites based on image identification, comprising the following steps: s1, identifying the distribution positions and construction types of a plurality of construction areas according to a construction plan of a building site; S2, configuring a monitoring camera for each construction area, and monitoring the wearing state of protective equipment of operators in the current area in real time; s3, marking a construction area which is directly communicated with other construction areas as a comprehensive monitoring area in the construction area of the construction site; S4, according to the information recorded by the working image, carrying out target area tendency analysis on the operators passing through the comprehensive monitoring area, and matching one or more pre-preparation staff as auxiliary verification, wherein the content of the target area tendency analysis is as follows: Primarily predicting the forward construction area of the operator according to the movement direction, the speed change and the track vector of the operator; Analyzing the corresponding states of the construction type of the forward construction area and the objects on the worker, and marking the primarily predicted forward construction area as a target area of the worker when the construction type of the forward construction area and the corresponding states of the objects on the worker correspond to each other; S5, identifying a target construction area of the operator according to the position information of the early-stage preparatio