Construction site personnel risk identification method and system based on image processing

Abstract

The invention provides a construction site personnel risk identification method and system based on image processing, and relates to the technical field of risk identification, and the method comprises the steps: collecting the real-time, working condition and personnel image data of a construction site, dividing a region into a plurality of risk regions, and working condition data such as site materials, equipment operation and fire-fighting supplies are combined to calculate a risk assessment value according to a preset proportion, and a risk assessment area is determined through scoring. The personnel image data are preprocessed to obtain personnel behavior data, behavior prediction probability data are obtained through a behavior prediction model, a danger judgment value is calculated according to the behavior prediction probability data and a risk assessment area, early warning is triggered if the danger judgment value exceeds a threshold value, and an optimization feedback mechanism is further arranged to record an early warning event to optimize an early warning mechanism. According to the method, the risk judgment values of the behavior risk data and the behavior probability data in the plurality of risk assessment areas are calculated, and early warning is performed according to the risk judgment values, so that the timeliness and accuracy of risk early warning are enhanced, and the safety of building workers can be further guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of risk identification, and in particular to a method and system for identifying risks of construction site personnel based on image processing. Background Art

[0002] Construction sites are complex and dangerous working environments with a large number of personnel, complex operation processes, and a variety of equipment and materials, which can easily lead to various safety accidents. Traditional safety management methods mainly rely on manual inspections and experience judgments, which have problems such as low efficiency, insufficient accuracy, and difficulty in achieving real-time monitoring. With the rapid development of computer technology and image processing technology, applying them to risk identification of personnel on construction sites has become an innovative and effective solution. By processing and analyzing personnel images on construction sites in real time, abnormal behaviors and potential dangers of personnel can be discovered in a timely manner, providing a scientific basis and decision-making support for safety management. Deploy high-definition cameras in key areas of construction sites to ensure that they can cover areas where personnel activities are frequent, such as construction areas, passages, and material storage areas. The installation position and angle of the camera need to be carefully designed to ensure that clear and complete personnel image data is obtained. At the same time, considering the environmental complexity of the construction site, such as light changes, dust, occlusion, etc., select camera equipment with a wide dynamic range, good low-light performance, and certain dust and water resistance functions. The collected personnel image data is transmitted to the back-end processing system in the form of video streams or image sequences, providing a basic data source for subsequent image processing and analysis.

[0003] Through the comprehensive application of a series of technical means such as personnel image data collection, preprocessing, feature extraction, behavior recognition and risk assessment, real-time and accurate monitoring and risk assessment of construction site personnel behavior can be achieved. This not only helps to improve the efficiency and accuracy of construction site safety management, timely discover and prevent safety accidents, but also provides new ideas and methods for intelligent safety management in the construction industry. With the continuous development and innovation of image processing technology and computer vision algorithms, it is believed that its application in the field of risk identification of construction site personnel will become more mature and perfect, making greater contributions to building a safer and more efficient construction site environment.

[0004] In the existing technology, the actual construction site environment is complex and it is difficult to accurately quantify the degree of danger of personnel in different risk areas. At the same time, construction workers are fully focused on their work and are prone to ignore the dangers that may result from their subsequent actions. There is a lack of early warning mechanisms to provide advance warnings to avoid accidents. Summary of the invention

[0005] The present invention provides a method and system for identifying personnel risks at construction sites based on image processing, so as to solve the defects in the prior art that it is difficult to accurately quantify the degree of danger of personnel in different risk areas and it is impossible to predict the dangerous behavior of personnel in advance.

[0006] In one aspect, the present invention provides a method for identifying risks of personnel at a construction site based on image processing, comprising: S1: Collect real-time data, working condition data and personnel image data in the construction site, divide the construction site into regions according to the preset safety standards, obtain multiple risk areas, calculate the working condition data according to the preset proportion, obtain the risk assessment value, and then score the multiple risk areas according to the risk assessment value to obtain the corresponding risk assessment area. The working condition data includes site material data, equipment operation data and firefighting equipment data. The steps to obtain the preset proportion include: The key factors affecting construction site safety are obtained by analyzing the durability and safety of materials, the reliability and failure rate of equipment, and the effectiveness and coverage of fire-fighting facilities.

[0007] The key factors are divided into target layer and criterion layer, and the weight vectors of the key factors are obtained by calculating the target layer and the criterion layer through the eigenvalue method.

[0008] Determine whether the sum of the weight vectors is equal to one, if so, obtain the preset weight, otherwise recalculate the target layer and criterion layer.

[0009] S2: Preprocess the personnel image data to obtain personnel behavior data, establish a behavior prediction model, input the personnel behavior data into the prediction model, and obtain behavior prediction probability data.

[0010] S3: The danger judgment value is calculated based on the behavior prediction probability data combined with the corresponding risk assessment area to determine whether the danger judgment value exceeds the preset threshold. If so, the person is considered to be in a dangerous state and the early warning mechanism is triggered. Otherwise, the person is considered to be in a safe state and no alarm is required.

[0011] S4: Establish an optimized feedback mechanism, record the events that trigger the early warning mechanism, obtain early warning events, and optimize the early warning mechanism based on the early warning events.

[0012] According to a construction site personnel risk identification method based on image processing provided by the present invention, in step S1, the step of obtaining a risk assessment value includes: S11: Based on industry standards, actual site conditions and safety management requirements, a weight setting algorithm is used to determine the preset weights of site material data, equipment operating status and firefighting supplies data.

[0013] S12: Calculate the risk assessment value in the risk assessment formula according to the preset weight.

[0014] According to a construction site personnel risk identification method based on image processing provided by the present invention, in step S12, the risk assessment formula is expressed as:

[0015] In the formula, is the weight of the site material data, is the weight of the device operating status, is the weight of the firefighting supplies data, is the site material data, Is the device operating status, It is the data of firefighting supplies. is the weight of each category in the site material data, is the weight of each construction equipment operating parameter in the equipment operating status, It is the weight of each category in the fire-fighting equipment data.

[0016] According to a method for identifying personnel risks at a construction site based on image processing provided by the present invention, in step S2, the step of obtaining personnel behavior data includes: According to the clarity standard, the noise of the personnel image data is removed to obtain the effective personnel activity image.

[0017] Extract posture features, motion trajectory features and behavior features from valid personnel activity data.

[0018] Posture features, motion trajectory features and behavior features are used to identify and classify personnel behavior to obtain personnel behavior sequences and behavior statistics.

[0019] The personnel behavior sequence and behavior statistics are cleaned, and feature data is extracted from the personnel behavior sequence. The feature data and behavior statistics are aggregated using the summation method to obtain the personnel behavior data.

[0020] According to a construction site personnel risk identification method based on image processing provided by the present invention, in step S2, the step of obtaining behavior prediction probability data includes: A behavior prediction model is constructed based on a recurrent neural network, and behavioral statistical data is used to train the behavior prediction model.

[0021] Use the grid search method to optimize the behavior prediction model, and evaluate whether the performance of the behavior prediction model reaches an excellent value through accuracy, precision, and recall. If so, stop optimizing the behavior prediction model and input the personnel behavior data into the behavior prediction model to obtain the behavior prediction probability data. Otherwise, continue to optimize the behavior prediction model.

[0022] According to a construction site personnel risk identification method based on image processing provided by the present invention, in step S3, the multiple risk areas include low-risk areas, medium-risk areas and high-risk areas. The step of obtaining the risk judgment value includes: S31: Based on the safety regulations of the construction industry and the statistical probabilities of different behaviors using historical data, the behavior prediction probability data is classified into behavior risk data and behavior probability data.

[0023] S32: Quantify the behavior risk data and the behavior probability data to obtain a risk coefficient.

[0024] S32: Calculate the risk judgment values ​​of the behavior risk data and the behavior probability data in multiple risk assessment areas.

[0025] According to a construction site personnel risk identification method based on image processing provided by the present invention, in step S33, the formula for the danger judgment value is expressed as:

[0026] In the formula, is the number of types of behavioral risk data, It is The probability of a behavior occurring, It is The risk factor corresponding to the behavior, It is The risk assessment value of the risk area where the behavior occurs.

[0027] According to a construction site personnel risk identification method based on image processing provided by the present invention, in step S4, the step of optimizing the early warning mechanism includes: S41: Clean the real-time data to obtain personnel activity data, and use the personnel activity data to update the behavior prediction model according to the feedback frequency, and re-extract various features in the effective personnel activity data to obtain the target features.

[0028] S42: Use the back propagation algorithm to adjust the target features.

[0029] S43: Record the event that triggers the early warning mechanism to obtain the early warning event, and conduct an in-depth analysis of the cause of the early warning event to obtain the analysis result.

[0030] S44: Based on the analysis results, the warning rules and preset thresholds within the warning mechanism are adjusted, and the priority and notification method of the warning mechanism are adjusted according to the importance and urgency.

[0031] According to a construction site personnel risk identification method based on image processing provided by the present invention, in step S42, the step of adjusting the target feature includes: S421: Combine target features into an input vector, input it into the behavior detection model, and output the target behavior.

[0032] S422: Select the cross entropy loss function for the classification task, and calculate the gradient of weights and biases according to the chain rule.

[0033] S423: Update weights and biases using a gradient descent algorithm based on the gradient, thereby adjusting the target features.

[0034] On the other hand, the present invention further provides a construction site personnel risk identification system based on image processing, which adopts any of the above-mentioned construction site personnel risk identification methods based on image processing, and the risk identification system comprises: Data collection and regional risk assessment module: collects real-time data, working condition data and personnel image data in the construction site, divides the construction site into regions according to the preset safety standards, obtains multiple risk areas, calculates the working condition data according to the preset proportion, obtains the risk assessment value, and then scores the multiple risk areas according to the risk assessment value to obtain the corresponding risk assessment area; Personnel image data processing and behavior prediction module: pre-process personnel image data to obtain personnel behavior data, establish a behavior prediction model, input personnel behavior data into the prediction model, and obtain behavior prediction probability data; Danger judgment and warning trigger module: The danger judgment value is calculated based on the behavior prediction probability data combined with the corresponding risk assessment area, and it is determined whether the danger judgment value exceeds the preset threshold. If so, the person is considered to be in a dangerous state and the warning mechanism is triggered. Otherwise, the person is considered to be in a safe state and no alarm is required.

[0035] Optimization feedback mechanism module: Establish an optimization feedback mechanism, record the events that trigger the early warning mechanism, obtain early warning events, and optimize the early warning mechanism based on the early warning events.

[0036] The present invention provides a method and system for identifying risks to construction site personnel based on image processing. Risk assessment areas are obtained by dividing risk areas according to risk assessment values, which solves the problem that personnel safety risks on construction sites are difficult to effectively identify and accurately control. High-risk areas can be accurately located in advance, so that the construction party can strengthen safety protection measures in a targeted manner, optimize resource allocation, and improve the overall safety level of the construction site.

[0037] The present invention provides a method and system for identifying personnel risks on construction sites based on image processing. By establishing a behavior prediction model and inputting personnel behavior data to obtain behavior prediction probability data, the problem of difficulty in predicting personnel dangerous behaviors in advance and lack of accurate basis for personnel risk warning is solved, and the beneficial effects of enhancing the timeliness and accuracy of risk warning and optimizing safety management measures are achieved.

[0038] The present invention provides a method and system for identifying risks to construction site personnel based on image processing. By calculating the danger judgment values ​​in multiple risk assessment areas based on behavior risk data and behavior probability data, and issuing warnings based on the danger judgment values, it solves the problems in the past that it is difficult to accurately quantify the danger levels of personnel in different risk areas at construction sites, and that the warning mechanisms are often broad and cannot promptly issue alarms for dangerous conditions of specific personnel in specific areas. Early warnings can be issued in advance when construction workers are working, which can further ensure the safety of construction workers. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 It is a flow chart of a method for identifying risks of construction site personnel based on image processing provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] Combine the following Figure 1 The present invention describes a method and system for identifying risks of construction site personnel based on image processing.

[0043] like Figure 1 As shown, an embodiment of the present invention provides a method and system for identifying risks of construction site personnel based on image processing, and the method mainly includes the following steps: S1: Collect real-time data, working condition data and personnel image data in the construction site, divide the construction site into regions according to the preset safety standards, obtain multiple risk areas, calculate the working condition data according to the preset proportion, obtain the risk assessment value, and then score the multiple risk areas according to the risk assessment value to obtain the corresponding risk assessment area. The working condition data includes site material data, equipment operation data and firefighting equipment data. The steps to obtain the preset proportion include: The key factors affecting construction site safety are obtained by analyzing the durability and safety of materials, the reliability and failure rate of equipment, and the effectiveness and coverage of fire-fighting facilities.

[0044] The key factors are divided into the target layer and the criterion layer, and the weight vector of the key factors is obtained by calculating the target layer and the criterion layer through the eigenvalue method. The formula is expressed as:

[0045] In the formula, is the first factors, The key factor is factors, The number of factors representing key factors, is the index variable to be summed.

[0046] Determine whether the sum of the weight vectors is equal to one, if so, obtain the preset proportion, otherwise recalculate the weight vector.

[0047] In step S1, the step of obtaining a risk assessment value includes: S11: Based on industry standards, actual site conditions and safety management requirements, use a weight setting algorithm to determine the preset weights of site material data, equipment operating status and firefighting equipment data. Define site material data categories as follows: ,include( ),in is the material size, The material shape, is the material density, is the material hardness, defining the equipment operating status classification as ,include( ),in is a performance parameter, is the running mode, is operational stability, It is an indication of the equipment status, and defines the classification of firefighting equipment data as ,include( ),in It is basic information data. is the performance parameter data, is to use and manipulate data, It is certification and compliance data.

[0048] S12: Calculate the risk assessment value in the risk assessment formula according to the preset weight.

[0049] The risk assessment formula is expressed as:

[0050] In the formula, is the weight of the site material data, is the weight of the device operating status, is the weight of the firefighting supplies data, is the site material data, Is the device operating status, It is the data of firefighting supplies. is the weight of each category in the site material data, is the weight of each construction equipment operating parameter in the equipment operating status, It is the weight of each category in the fire-fighting equipment data.

[0051] S2: Preprocess the personnel image data to obtain personnel behavior data, establish a behavior prediction model, input the personnel behavior data into the prediction model, and obtain behavior prediction probability data.

[0052] In step S2, the step of obtaining personnel behavior data includes: According to the clarity standard, the noise of the personnel image data is removed to obtain an effective personnel activity image. By accurately identifying the noise components in the image and eliminating them with appropriate filtering technology, a clear and accurate effective personnel activity image is successfully obtained, laying a solid foundation for subsequent in-depth analysis.

[0053] Extract posture features, motion trajectory features and behavior features from valid personnel activity data. Among them, posture features cover the relative position and angle information of various parts of the personnel's body, motion trajectory features record in detail the movement path and speed changes of personnel within a certain time range, and behavior features comprehensively reflect the behavior patterns and purpose tendencies of personnel.

[0054] The posture features, motion trajectory features and behavior features are used to identify and classify personnel behaviors to obtain personnel behavior sequences and behavior statistics. These data can intuitively reflect the key information such as the behavior patterns, activity rules and occurrence frequency of various behaviors of personnel on the construction site, which is of great value for in-depth analysis of personnel behavior risks and formulation of targeted safety management strategies.

[0055] The personnel behavior sequence and behavior statistics are cleaned, and feature data is extracted from the personnel behavior sequence. The feature data and behavior statistics are aggregated using the summation method to obtain the personnel behavior data.

[0056] The steps to obtain behavior prediction probability data include: A behavior prediction model is constructed based on a recurrent neural network, and behavioral statistical data is used to train the behavior prediction model.

[0057] Use the grid search method to optimize the behavior prediction model, and evaluate whether the performance of the behavior prediction model reaches an excellent value through accuracy, precision, and recall. If so, stop optimizing the behavior prediction model and input the personnel behavior data into the behavior prediction model to obtain the behavior prediction probability data. Otherwise, continue to optimize the behavior prediction model.

[0058] S3: The danger judgment value is calculated based on the behavior prediction probability data combined with the corresponding risk assessment area to determine whether the danger judgment value exceeds the preset threshold. If so, the person is considered to be in a dangerous state and the early warning mechanism is triggered. Otherwise, the person is considered to be in a safe state and no alarm is required.

[0059] In step S3, the multiple risk areas include low risk areas, medium risk areas and high risk areas. The steps of obtaining the risk judgment value include: S31: Based on the safety regulations of the construction industry and the statistical probabilities of different behaviors using historical data, the behavior prediction probability data is classified into behavior risk data and behavior probability data.

[0060] S32: Quantify the behavior risk data and behavior probability data to obtain the risk coefficient. The formula is:

[0061] In the formula, is an identification variable ( =1 represents behavioral risk data, =2 represents behavioral probability data), is the relevant indicator count, is the probability of the behavior occurring, is the hazard degree coefficient, is the frequency quantization value, is the quantity conversion factor.

[0062] For behavioral risk data, corresponding weight coefficients are set according to the different levels of safety hazards, and then weighted calculations are performed based on the probability of occurrence, so as to obtain a risk coefficient that can intuitively reflect the severity of the risk. For behavioral probability data, quantitative conversion is mainly performed based on the frequency of occurrence in different construction stages and different regional environments, and risk coefficients with clear numerical meanings are also generated. Through this quantitative method, the originally abstract behavioral data can be converted into measurable and comparable risk coefficient indicators, providing an accurate data basis for subsequent hazard judgments.

[0063] S33: Calculate the risk judgment value of behavior risk data and behavior probability data in multiple risk assessment areas. The risk judgment value can accurately reflect the overall risk level of personnel behavior in a specific risk assessment area, and provide a key basis for judging whether to trigger the early warning mechanism in the future, so as to prevent and respond to possible dangerous situations in a timely and effective manner.

[0064] The formula for the risk judgment value is expressed as:

[0065] In the formula, is the number of types of behavioral risk data, It is The probability of a behavior occurring, It is The risk factor corresponding to the behavior, It is The risk assessment value of the risk area where the behavior occurs.

[0066] S4: Establish an optimized feedback mechanism, record the events that trigger the early warning mechanism, obtain early warning events, and optimize the early warning mechanism based on the early warning events.

[0067] In step S4, the steps of optimizing the early warning mechanism include: S41: Clean the real-time data to obtain personnel activity data, and use the personnel activity data to update the behavior prediction model according to the feedback frequency, and re-extract various features in the effective personnel activity data to obtain the target features.

[0068] S42: Use the back propagation algorithm to adjust the target features.

[0069] S43: Record the event that triggers the early warning mechanism to obtain the early warning event, and conduct an in-depth analysis of the cause of the early warning event to obtain the analysis result.

[0070] S44: Based on the analysis results, the warning rules and preset thresholds within the warning mechanism are adjusted, and the priority and notification method of the warning mechanism are adjusted according to the importance and urgency.

[0071] In step S42, the step of adjusting the target feature includes: S421: Carefully select representative and critical target features from the many features obtained after preprocessing and feature extraction. Then, strictly follow the specific combination rules and order to cleverly combine these target features into a complete input vector. The construction of this input vector needs to fully consider the correlation between the various features and their importance in the overall behavior representation. Then, this carefully constructed input vector is accurately input into the pre-trained behavior detection model designed and developed specifically for construction site personnel behavior detection. After a complex calculation and analysis process within the model, the final output is the target behavior corresponding to the input vector, which is the specific behavior category or action pattern of the construction site personnel determined by the model based on the input features.

[0072] S422: In the training and optimization process of the behavior detection model, for the classification task, after careful consideration and comparison, the cross entropy loss function was selected as the key indicator to measure the degree of difference between the model prediction results and the true label. This is because the cross entropy loss function can effectively reflect the degree of deviation between the probability distribution of the model prediction and the probability distribution of the true label when dealing with classification problems, thereby providing a clear direction for the optimization of the model. Subsequently, according to the chain rule, a widely used and extremely important derivation rule in deep learning, the loss function is derived with respect to the weights and biases in the model. Through a meticulous and rigorous calculation process, the gradients of the weights and biases are obtained. These gradient values ​​accurately represent the sensitivity of the loss function to changes in weights and biases under the current model parameters. They will serve as the key basis for subsequent updates of the model parameters, guiding the model to gradually optimize in the direction of reducing the loss function value.

[0073] S423: After obtaining the gradients of weights and biases, the weights and biases in the model are updated according to the classic and effective optimization algorithm, the gradient descent algorithm. The core idea of ​​the gradient descent algorithm is to gradually adjust the model parameters in the opposite direction of the gradient so that the loss function value can be continuously reduced. Specifically, according to the pre-set learning rate, the weights and biases are slightly adjusted in the opposite direction of their respective gradients. The choice of learning rate is crucial, as it controls the step size of each parameter update. If the learning rate is too large, the model may skip the optimal solution during training or even fail to converge; if the learning rate is too small, the model training speed will be too slow, requiring a lot of time and computing resources. Through this weight and bias update method based on the gradient descent algorithm, the effect of the target feature in the model can be effectively adjusted, so that the model can more accurately detect and classify the behavior of construction site personnel, continuously improve the performance and accuracy of the model, and thus better adapt to the complex and changeable construction site environment and personnel behavior patterns.

[0074] The present embodiment provides a method for identifying personnel risks at construction sites based on image processing. The risk assessment area is obtained by dividing the risk area by the risk assessment value, which solves the problem that the personnel safety risks at the construction site are difficult to effectively identify and accurately control, and achieves the beneficial effect of being able to accurately locate the high-risk area in advance, so that the construction party can strengthen the safety protection measures in a targeted manner, optimize the resource allocation, and improve the overall safety level of the construction site. In addition, by establishing a behavior prediction model and inputting the personnel behavior data to obtain the behavior prediction probability data, the problem of being difficult to predict the dangerous behavior of personnel in advance and lacking the basis for accurate personnel risk warning is solved, and the timeliness and accuracy of risk warning and the beneficial effect of optimizing safety management measures are achieved. In addition, by calculating the danger judgment value of the behavior risk data and the behavior probability data in multiple risk assessment areas, and issuing an early warning according to the danger judgment value, the problem that it is difficult to accurately quantify the danger degree of personnel in different risk areas in the past construction sites and the early warning mechanism is often broad, and the problem that the danger situation of specific personnel in a specific area cannot be timely issued is solved. When the construction workers are working, early warning can be issued, which can further ensure the safety of the construction workers.

[0075] Based on the same general inventive concept, the present invention also protects a construction site personnel risk identification system based on image processing, which can adopt the above-mentioned construction site personnel risk identification method based on image processing. The risk identification system includes: Data collection and regional risk assessment module: collects real-time data, working condition data and personnel image data in the construction site, divides the construction site into regions according to the preset safety standards, obtains multiple risk areas, calculates the working condition data according to the preset proportion, obtains the risk assessment value, and then scores the multiple risk areas according to the risk assessment value to obtain the corresponding risk assessment area; Personnel image data processing and behavior prediction module: pre-process personnel image data to obtain personnel behavior data, establish a behavior prediction model, input personnel behavior data into the prediction model, and obtain behavior prediction probability data; Danger judgment and warning trigger module: The danger judgment value is calculated based on the behavior prediction probability data combined with the corresponding risk assessment area, and it is determined whether the danger judgment value exceeds the preset threshold. If so, the person is considered to be in a dangerous state and the warning mechanism is triggered. Otherwise, the person is considered to be in a safe state and no alarm is required.

[0076] Optimization feedback mechanism module: Establish an optimization feedback mechanism, record the events that trigger the early warning mechanism, obtain early warning events, and optimize the early warning mechanism based on the early warning events.

[0077] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0078] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A construction site personnel risk identification method based on image processing, characterized in that: include: S1: Collecting real-time data, working condition data and personnel image data in the construction site, dividing the construction site into regions according to preset safety standards to obtain multiple risk areas, calculating the working condition data according to preset weights to obtain risk assessment values, and then scoring the multiple risk areas according to the risk assessment values ​​to obtain corresponding risk assessment areas; The working condition data includes site material data, equipment operation data and firefighting supplies data; The step of obtaining the preset specific gravity includes: The key factors affecting construction site safety are obtained by analyzing the durability and safety of materials, the reliability and failure rate of equipment, and the effectiveness and coverage of fire-fighting facilities; Dividing the key factors into a target layer and a criterion layer, and calculating the target layer and the criterion layer by an eigenvalue method to obtain a weight vector of the key factors; Determine whether the sum of the weight vectors is equal to one, if so, obtain the preset proportion, otherwise recalculate the weight vector; S2: preprocessing the personnel image data to obtain personnel behavior data, establishing a behavior prediction model, inputting the personnel behavior data into the prediction model, and obtaining behavior prediction probability data; S3: Calculate a danger judgment value based on the behavior prediction probability data in combination with the corresponding risk assessment area, and determine whether the danger judgment value exceeds a preset threshold. If so, the person is considered to be in a dangerous state and a warning mechanism is triggered; otherwise, the person is considered to be in a safe state and no alarm is required; S4: Establish an optimization feedback mechanism, record the events that trigger the early warning mechanism, obtain early warning events, and optimize the early warning mechanism according to the early warning events.

2. The method for identifying construction site personnel risks based on image processing according to claim 1, characterized in that: In step S1, the step of obtaining the risk assessment value includes: S11: Determine the preset weights of the site material data, the equipment operation status and the firefighting equipment data using a weight setting algorithm according to industry standards, actual site conditions and safety management requirements; S12: Calculating the risk assessment value in a risk assessment formula according to the preset weight.

3. The method for identifying construction site personnel risks based on image processing according to claim 2 is characterized in that: In step S12, the risk assessment formula is expressed as: In the formula, is the weight of the site material data, is the weight of the device operating status, is the weight of the firefighting supplies data, is the site material data, Is the device operating status, It is the data of firefighting supplies. is the weight of each category in the site material data, is the weight of each construction equipment operating parameter in the equipment operating status, It is the weight of each category in the fire-fighting equipment data.

4. The method for identifying construction site personnel risks based on image processing according to claim 1, characterized in that: In step S2, the step of obtaining the personnel behavior data includes: Denoising the personnel image data according to a clarity standard to obtain a valid personnel activity image; Extracting posture features, motion trajectory features and behavior features from the valid personnel activity image; The posture features, the motion trajectory features and the behavior features are used to perform personnel behavior recognition and behavior classification to obtain personnel behavior sequences and behavior statistical data; The personnel behavior sequence and the behavior statistical data are cleaned, and feature data is extracted from the personnel behavior sequence. The feature data and the behavior statistical data are aggregated using a summation method to obtain the personnel behavior data.

5. The method for identifying construction site personnel risks based on image processing according to claim 1, characterized in that: In step S2, the step of obtaining the behavior prediction probability data includes: Building a behavior prediction model based on a recurrent neural network, and using the behavior statistical data to train the behavior prediction model; The behavior prediction model is optimized using the grid search method, and the accuracy, precision and recall are used to evaluate whether the performance of the behavior prediction model reaches an excellent value. If so, the optimization of the behavior prediction model is stopped, and the personnel behavior data is input into the behavior prediction model to obtain the behavior prediction probability data; otherwise, the behavior prediction model is continued to be optimized.

6. The method for identifying construction site personnel risks based on image processing according to claim 1, characterized in that: In step S3, the plurality of risk areas include a low risk area, a medium risk area and a high risk area; The step of obtaining the risk judgment value includes: S31: classifying the behavior prediction probability data into behavior risk data and behavior probability data according to the safety regulations of the construction industry and the probability of occurrence of different behaviors using historical data; S32: quantifying the behavior risk data and the behavior probability data to obtain a risk coefficient; S32: Calculate the risk judgment values ​​of the behavior risk data and the behavior probability data in the plurality of risk assessment areas.

7. The method for identifying construction site personnel risks based on image processing according to claim 1, characterized in that: In step S33, the formula for the risk judgment value is expressed as: In the formula, is the number of types of behavioral risk data, It is The probability of a behavior occurring, It is The risk factor corresponding to the behavior, It is The risk assessment value of the risk area where the behavior occurs.

8. The method for identifying construction site personnel risks based on image processing according to claim 4 is characterized in that: In step S4, the step of optimizing the early warning mechanism includes: S41: cleaning the real-time data to obtain personnel activity data, and using the personnel activity data to update the behavior prediction model according to the feedback frequency, and re-extracting various features in the personnel activity image to obtain target features; S42: using a back propagation algorithm to adjust the target feature; S43: recording the event that triggers the early warning mechanism to obtain an early warning event, and deeply analyzing the cause of the early warning event to obtain an analysis result; S44: According to the analysis result, the warning rules in the warning mechanism and the preset threshold are adjusted, and according to the importance and urgency, the priority and notification method of the warning mechanism are adjusted.

9. The method for identifying construction site personnel risks based on image processing according to claim 8, characterized in that: In step S42, the step of adjusting the target feature includes: S421: Combining the target features into an input vector, inputting the vector into the behavior detection model, and outputting the target behavior; S422: Select the classification task and the cross entropy loss function, and calculate the gradient of the weight and bias according to the chain rule; S423: Update the weights and biases using a gradient descent algorithm according to the gradient, thereby adjusting the target feature.

10. A construction site personnel risk identification system based on image processing, which adopts a construction site personnel risk identification method based on image processing as claimed in any one of claims 1 to 9, characterized in that: The risk identification system comprises: Data collection and regional risk assessment module: collects real-time data, working condition data and personnel image data in the construction site, divides the construction site into regions according to the preset safety standards, obtains multiple risk areas, calculates the working condition data according to the preset proportion, obtains the risk assessment value, and then scores the multiple risk areas according to the risk assessment value to obtain the corresponding risk assessment area; Personnel image data processing and behavior prediction module: pre-process personnel image data to obtain personnel behavior data, establish a behavior prediction model, input personnel behavior data into the prediction model, and obtain behavior prediction probability data; Danger judgment and warning trigger module: The danger judgment value is calculated based on the behavior prediction probability data combined with the corresponding risk assessment area, and it is determined whether the danger judgment value exceeds the preset threshold. If so, it is considered that the person is in a dangerous state and the warning mechanism is triggered. Otherwise, it is considered that the person is in a safe state and no alarm is required; Optimization feedback mechanism module: establish an optimization feedback mechanism, record the events that trigger the early warning mechanism, obtain early warning events, and optimize the early warning mechanism according to the early warning events.

Images