An intelligent identification method, device, equipment and medium based on construction site safety model

Abstract

The present invention relates to an intelligent identification method, device, equipment and medium based on a construction site safety model, and belongs to the field of construction safety. The method includes face recognition, personnel positioning and risk identification through an intelligent safety helmet, the intelligent safety helmet is embedded with a face recognition module, a real-time positioning module, a risk identification module and a sensor group, the face recognition module obtains face recognition results through a face detection algorithm, the real-time positioning module obtains personnel location information through GPS positioning, and the risk identification module includes a self-risk identification module and an external risk identification module to obtain self-risk information and external risk information; a construction site safety model is constructed based on face recognition results, personnel location information, self-risk information and external risk information. The present invention realizes the construction of a construction site safety model through an intelligent safety helmet, identifies various risks of a construction site, and ensures the safety of workers in construction.

Description

Technical Field

[0001] The present invention belongs to the field of construction safety technology, and specifically relates to an intelligent identification method, device, equipment and medium based on a construction site safety model. Background Art

[0002] Safety management of construction sites is extremely important. It not only affects the life safety of construction workers, but also affects the smooth progress and overall quality of the project. There are many dangerous factors in the construction process of construction sites, such as falling from heights, fire, collapse, etc. Without effective construction safety management, these dangerous factors may lead to serious safety accidents and even cause irreparable casualties. Therefore, scientific and strict construction safety management is the premise for ensuring the life safety of construction workers.

[0003] Existing smart safety helmets have realized functions such as real-time calls, real-time positioning, and image capture, and are used for intelligent identification of risks on construction sites. However, current technology still has shortcomings in terms of whether there are risks in the user's surrounding environment, whether the user himself is at risk of falling from a height, and whether people other than the user are wearing the smart safety helmet correctly, and these technologies urgently need to be improved. Summary of the invention

[0004] In order to solve the above problems existing in the prior art, the present invention provides an intelligent identification method, device, equipment and medium based on a construction site safety model.

[0005] The purpose of the present invention can be achieved by the following technical solutions:

[0006] An intelligent identification method based on a construction site safety model, the implementation of the intelligent identification method comprises the following steps:

[0007] The smart helmet is used to perform face recognition, personnel positioning and risk identification. The smart helmet is embedded with a face recognition module, a real-time positioning module, a risk identification module and a sensor group. The face recognition module obtains face recognition results through a face detection algorithm. The face recognition results include face verification passing and face verification not matching. The real-time positioning module obtains personnel location information through GPS positioning. The risk identification module includes an own risk identification module and an external risk identification module to obtain own risk information and external risk information. The own risk information includes own high risk information, own medium risk information and own low risk information. The external risk information includes external high risk information and external low risk information.

[0008] A construction site safety model is constructed based on the face recognition results, the personnel location information, the own risk information and the external risk information.

[0009] Preferably, the face detection algorithm is specifically:

[0010] Capturing an initial face image and obtaining the size of the initial face image, predefining a set of fixed reference frames according to the size of the initial face image, the number of the fixed reference frames being m×n×k, wherein m×n is the size of the initial face image, k is the number of fixed reference frames corresponding to each feature point, and the number of the feature points is m×n;

[0011] The offset between the fixed reference frame and the real frame is obtained through network regression, and the real frame coordinates are calculated according to the fixed reference frame coordinates and the offset. The calculation formula is: , where (x z ,y z ,w z ,h z ) is the real frame coordinate, (x c ,y c ,w c ,h c ) is the fixed reference frame coordinate, (△x,△y,△w,△h) is the offset;

[0012] In the real frame, the left eye key point, the right eye key point, the nose tip key point, the left corner of the mouth key point and the right corner of the mouth key point are preset and the key point coordinates (x i ,y i ), i=1,2,…,5, by aligning the matrix and the alignment vector (f, g) to transform the keypoint coordinates into the standard keypoint coordinates (x i ',y i '), i=1,2,…,5, the elements in the alignment matrix and the alignment vector are obtained by minimizing the error function, and the error function is , then the expression of the standard key point coordinates is (x i ',y i ')=E(x i ,y i )+(f,g);

[0013] Face matching is performed using pre-entered face information to obtain the face recognition result.

[0014] Preferably, the obtaining of the own risk information and the external risk information specifically includes:

[0015] The self-risk identification module is used to obtain the self-risk value based on the user sensor data, and output the self-risk information through the self-risk value;

[0016] The external risk identification module is used to extract personnel anchor frames and smart helmet anchor frames through an improved feature extraction model and output the external risk information. The improved feature extraction model includes an input end, a backbone feature extraction network, a feature fusion network, a prediction network, and an output end.

[0017] Preferably, the obtaining of the user's own risk value based on the user's sensor data includes:

[0018] The user sensing data is acquired through the sensor group, the sensor group includes a speed sensor, a temperature sensor, a particle sensor and a height sensor, the user sensing data includes a user moving speed, an ambient temperature, an ambient particle concentration, a user height difference, and a user's current height, the user height difference being a height change value of the user in a vertical direction within a preset time;

[0019] Calculating the own risk value based on the user sensor data;

[0020] When the own risk value is 1, the own high risk information is output, when the own risk value is 0.5, the own medium risk information is output, and when the own risk value is 0, the own low risk information is output.

[0021] Preferably, the calculation formula of the self-risk value is: , where α is the risk value, T is the ambient temperature, PM is the ambient particulate matter concentration, and v is the user's moving speed. is the height difference of the user, h is the current height of the user, θ is the angle between the user's moving direction and the vertical direction, T1 is the preset high temperature threshold, T0 is the preset low temperature threshold, PM1 is the preset high concentration threshold of particulate matter, PM0 is the preset low concentration threshold of particulate matter, is the preset high drop threshold, is the preset low drop threshold, and g is the acceleration due to gravity.

[0022] Preferably, the improved feature extraction model specifically includes:

[0023] The real-time construction site image is collected by the smart helmet and input into the input end, and the input end performs image preprocessing;

[0024] The backbone feature extraction network includes a ShuffleNetv2 module and an SPPF module, wherein the ShuffleNetv2 module includes a 1×1 element-by-element group convolution, a Channel Shuffle channel shuffle mechanism, a 3×3 depth convolution, and a 1×1 element-by-element group convolution in sequence, and the SPPF includes three 5×5 maximum pooling layers connected in series;

[0025] The feature fusion network includes an FPN feature pyramid structure and a PAN structure;

[0026] The preset data set is used to train the improved feature extraction model, and the data set includes a training set, a test set and a validation set, which are a construction site portrait image set and a smart helmet image set;

[0027] The output end outputs the personnel anchor frame and the smart helmet anchor frame, matches the personnel anchor frame and the smart helmet anchor frame that are closest to each other and obtains anchor frame data, wherein the anchor frame data includes the personnel anchor frame center coordinates (X, Y), the smart helmet anchor frame center coordinates (x, y), the personnel anchor frame height T, the smart helmet anchor frame height t, and the personnel anchor frame width R;

[0028] The external risk value is calculated based on the anchor frame data, and the calculation formula is: , where β is the external risk value;

[0029] When the external risk value is 1, the external high risk information is output, and when the external risk value is 0, the external low risk information is output.

[0030] Preferably, the step of constructing a construction site safety model specifically includes:

[0031] When the output of the face verification is inconsistent, issuing a verification personnel instruction to the terminal;

[0032] The working range of the construction site is preset, and when the position information of the personnel exceeds the working range of the construction site, the personnel leaving the post verification instruction is fed back to the terminal;

[0033] When the self-high risk information or the self-medium risk information is output, a construction site safety confirmation instruction is fed back to the terminal;

[0034] When the external high-risk information is output, the smart helmet wearing verification instruction is fed back to the terminal.

[0035] An intelligent recognition device based on a construction site safety model, used to execute the intelligent recognition method described above, comprising an intelligent recognition module and a model building module;

[0036] The intelligent recognition module is used to perform face recognition, personnel positioning and risk identification through a smart helmet. The smart helmet is embedded with a face recognition module, a real-time positioning module, a risk identification module and a sensor group. The face recognition module obtains a face recognition result through a face detection algorithm. The face recognition result includes a face verification pass and a face verification failure. The real-time positioning module obtains personnel location information through GPS positioning. The risk identification module includes an own risk identification module and an external risk identification module to obtain own risk information and external risk information. The own risk information includes own high risk information, own medium risk information and own low risk information. The external risk information includes external high risk information and external low risk information.

[0037] The model building module is used to build a construction site safety model based on the face recognition results, the personnel location information, the own risk information and the external risk information.

[0038] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned intelligent recognition method when executing the program.

[0039] A storage medium containing computer executable instructions, which are used to execute the above-mentioned intelligent recognition method when executed by a computer processor.

[0040] The beneficial effects of the present invention are:

[0041] (1) Through the intelligent safety helmet, intelligent recognition such as face recognition, personnel positioning, self-risk recognition and external risk recognition can be realized, and a construction site safety model can be built to eliminate risks such as users leaving the workplace without permission, users falling from heights, site fires or collapses, and workers not wearing safety helmets correctly, so as to ensure the safety of construction sites;

[0042] (2) Face matching is performed through the face detection algorithm, which is a lightweight network structure that improves the efficiency of face recognition. The correction of key point coordinates by the alignment matrix and alignment vector can improve the accuracy and robustness of subsequent face recognition.

[0043] (3) The user's own risk is comprehensively judged based on the user's movement speed, ambient temperature, ambient particle concentration, user altitude difference, and user's current altitude. The judgment of the user's status is relatively accurate and comprehensive;

[0044] (4) Obtain anchor frame data and calculate external risk values ​​based on the personnel anchor frame and the smart helmet anchor frame to accurately determine whether the worker is wearing a smart helmet, whether the helmet is centered and worn correctly, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0046] Figure 1 The present invention is a flowchart of the steps of an intelligent identification method based on a construction site safety model. DETAILED DESCRIPTION

[0047] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0048] The working principle and use process of the present invention:

[0049] See also Figure 1 , an intelligent identification method based on a construction site safety model, comprising:

[0050] S1: Face recognition, personnel positioning and risk identification are performed through a smart helmet. The smart helmet is embedded with a face recognition module, a real-time positioning module, a risk identification module and a sensor group. The face recognition module obtains face recognition results through a face detection algorithm. The face recognition results include face verification passing and face verification not matching. The real-time positioning module obtains personnel location information through GPS positioning. The risk identification module includes an own risk identification module and an external risk identification module to obtain own risk information and external risk information. The own risk information includes own high risk information, own medium risk information and own low risk information. The external risk information includes external high risk information and external low risk information. The external risk identification module is used to warn the user whether the surrounding people are wearing the smart helmet correctly.

[0051] S2: Constructing a construction site safety model based on the face recognition result, the personnel location information, the own risk information and the external risk information.

[0052] In this embodiment, the face recognition result is obtained by using the face detection algorithm, which can be implemented by the following steps:

[0053] The face detection algorithm is specifically:

[0054] Capturing an initial face image and obtaining the size of the initial face image, predefining a set of fixed reference frames according to the size of the initial face image, the number of the fixed reference frames being m×n×k, wherein m×n is the size of the initial face image, k is the number of fixed reference frames corresponding to each feature point, and the number of the feature points is m×n;

[0055] The offset between the fixed reference frame and the real frame is obtained through network regression, and the real frame coordinates are calculated according to the fixed reference frame coordinates and the offset. The calculation formula is: , where (x z ,y z ,w z ,h z ) is the real frame coordinate, (x c ,y c ,w c ,h c ) is the fixed reference frame coordinate, (△x,△y,△w,△h) is the offset;

[0056] In the real frame, the left eye key point, the right eye key point, the nose tip key point, the left corner of the mouth key point and the right corner of the mouth key point are preset and the key point coordinates (x i ,y i ), i=1,2,…,5, by aligning the matrix and the alignment vector (f, g) to transform the keypoint coordinates into the standard keypoint coordinates (x i ',y i '), i=1,2,…,5, the elements in the alignment matrix and the alignment vector are obtained by minimizing the error function, and the error function is , then the expression of the standard key point coordinates is (x i ',y i ')=E(x i ,y i )+(f,g);

[0057] Face matching is performed using pre-entered face information to obtain the face recognition result.

[0058] In this embodiment, the risk identification module includes an internal risk identification module and an external risk identification module, and obtains internal risk information and external risk information, which can be implemented by the following steps:

[0059] S101: The self-risk identification module is used to obtain a self-risk value based on user sensor data, and output the self-risk information through the self-risk value;

[0060] S102: The external risk identification module is used to extract the personnel anchor frame and the smart helmet anchor frame through an improved feature extraction model, and output the external risk information, wherein the improved feature extraction model includes an input end, a backbone feature extraction network, a feature fusion network, a prediction network, and an output end;

[0061] In this embodiment, the self-risk identification module is used to obtain the self-risk value based on the user sensor data, which can be implemented by the following steps:

[0062] S101-1: Acquire the user sensor data through the sensor group, the sensor group includes a speed sensor, a temperature sensor, a particle sensor and a height sensor, the user sensor data includes the user's moving speed, the ambient temperature, the ambient particle concentration, the user's height difference, and the user's current height, the user's height difference is the height change value of the user in the vertical direction within a preset time, and the user sensor data can be used to indirectly determine whether a fire, collapse, high-speed fall of personnel, etc. may occur at the construction site;

[0063] S101-2: Calculate the user's own risk value based on the user's sensor data. The calculation formula is: , where α is the risk value, T is the ambient temperature, PM is the ambient particulate matter concentration, and v is the user's moving speed. is the height difference of the user, h is the current height of the user, θ is the angle between the user's moving direction and the vertical direction, T1 is the preset high temperature threshold, T0 is the preset low temperature threshold, PM1 is the preset high concentration threshold of particulate matter, PM0 is the preset low concentration threshold of particulate matter, is the preset high drop threshold, is the preset low drop threshold, g is the gravitational acceleration;

[0064] S101-3: When the own risk value is 1, the own high risk information is output; when the own risk value is 0.5, the own medium risk information is output; when the own risk value is 0, the own low risk information is output.

[0065] In this embodiment, the external risk identification module is used to extract the personnel anchor frame and the smart helmet anchor frame by improving the feature extraction model, which can be implemented by the following steps:

[0066] S102-1: The real-time construction site image is collected by the smart helmet and input into the input end, and the input end performs image preprocessing, and the image preprocessing includes but is not limited to random flipping, random cropping, contrast change, and scaling;

[0067] S102-2: The backbone feature extraction network includes a ShuffleNetv2 module and an SPPF module, wherein the ShuffleNetv2 module includes a 1×1 element-by-element group convolution, a Channel Shuffle channel shuffle mechanism, a 3×3 depth convolution, and a 1×1 element-by-element group convolution in sequence, and the SPPF includes three 5×5 maximum pooling layers connected in series;

[0068] S102-3: The feature fusion network includes an FPN feature pyramid structure and a PAN structure;

[0069] S102-4: A preset data set is used to train the improved feature extraction model, wherein the data set includes a training set, a test set, and a validation set, which are a construction site portrait image set and a smart helmet image set;

[0070] S102-5: The output end outputs the personnel anchor frame and the smart helmet anchor frame, matches the personnel anchor frame and the smart helmet anchor frame that are closest to each other and obtains anchor frame data, wherein the anchor frame data includes the personnel anchor frame center coordinates (X, Y), the smart helmet anchor frame center coordinates (x, y), the personnel anchor frame height T, the smart helmet anchor frame height t, and the personnel anchor frame width R;

[0071] S102-6: Calculate the external risk value based on the anchor frame data, and the calculation formula is: , where β is the external risk value;

[0072] S102-7: When the external risk value is 1, output the external high risk information; when the external risk value is 0, output the external low risk information.

[0073] In this embodiment, a construction site safety model is constructed based on the face recognition result, the personnel location information, the own risk information and the external risk information, which can be implemented by the following steps:

[0074] When the output of the face verification is inconsistent, a personnel verification instruction is issued to the terminal to verify the identity of the person present;

[0075] The working scope of the construction site is preset. When the position information of the personnel exceeds the working scope of the construction site, the personnel leaving the post verification instruction is fed back to the terminal to verify whether there are personnel who leave the construction site without permission during working hours;

[0076] When the high-risk information or medium-risk information is output, a construction site safety confirmation instruction is fed back to the terminal to check whether there is a fall of personnel, and whether there is a fire, collapse or other accidents at the construction site;

[0077] When the external high-risk information is output, the smart safety helmet wearing verification instruction is fed back to the terminal and it is checked whether there are people in the working range of the construction site who are not wearing the smart safety helmet correctly.

[0078] An intelligent recognition device based on a construction site safety model, comprising an intelligent recognition module and a model building module;

[0079] The intelligent recognition module is used to perform face recognition, personnel positioning and risk identification through a smart helmet. The smart helmet is embedded with a face recognition module, a real-time positioning module, a risk identification module and a sensor group. The face recognition module obtains face recognition results through a face detection algorithm. The face recognition results include face verification passing and face verification not matching. The real-time positioning module obtains personnel location information through GPS positioning. The risk identification module includes an own risk identification module and an external risk identification module to obtain own risk information and external risk information. The own risk information includes own high risk information, own medium risk information and own low risk information. The external risk information includes external high risk information and external low risk information. The external risk identification module is used to warn the user whether the surrounding people are wearing the smart helmet correctly.

[0080] The model building module is used to build a construction site safety model based on the face recognition results, the personnel location information, the own risk information and the external risk information.

[0081] The computer storage medium of the embodiment of the present invention may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device.

[0082] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0083] The program code included on the computer readable medium can be transmitted with any appropriate medium, including but not limited to wireless, electric wire, optical cable, RF, etc., or any suitable combination of the above. The computer program code for performing the operation of the present invention can be written in one or more programming languages ​​or their combinations, and the programming language includes object-oriented programming languages-such as Java, Smalltalk, C++, and also includes conventional procedural programming languages-such as "C" language or similar programming languages. The program code can be executed completely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on the remote computer, or completely on the remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0084] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. An intelligent identification method based on a construction site safety model, characterized in that: The implementation of the intelligent identification method comprises the following steps: The smart helmet is used to perform face recognition, personnel positioning and risk identification. The smart helmet is embedded with a face recognition module, a real-time positioning module, a risk identification module and a sensor group. The face recognition module obtains face recognition results through a face detection algorithm. The face recognition results include face verification passed and face verification not met. The real-time positioning module obtains personnel location information through GPS positioning. The risk identification module includes an own risk identification module and an external risk identification module to obtain own risk information and external risk information. The own risk information includes own high risk information, own medium risk information, and own low risk information. The external risk information includes external high risk information and external low risk information. The own risk identification module is used to obtain the own risk value based on user sensor data, and output the own risk information through the own risk value. The calculation formula of the own risk value is: , where α is the risk value, T is the ambient temperature, PM is the ambient particulate matter concentration, and v is the user's moving speed. is the height difference of the user, h is the current height of the user, θ is the angle between the user's moving direction and the vertical direction, T1 is the preset high temperature threshold, T0 is the preset low temperature threshold, PM1 is the preset high concentration threshold of particulate matter, PM0 is the preset low concentration threshold of particulate matter, is the preset high drop threshold, is the preset low drop threshold, g is the gravitational acceleration; A construction site safety model is constructed based on the face recognition results, the personnel location information, the own risk information and the external risk information.

2. The intelligent identification method according to claim 1, characterized in that: The face detection algorithm is specifically: Capturing an initial face image and obtaining the size of the initial face image, predefining a set of fixed reference frames according to the size of the initial face image, the number of the fixed reference frames being m×n×k, wherein m×n is the size of the initial face image, k is the number of fixed reference frames corresponding to each feature point, and the number of the feature points is m×n; The offset between the fixed reference frame and the real frame is obtained through network regression, and the real frame coordinates are calculated according to the fixed reference frame coordinates and the offset. The calculation formula is: , where (x z ,y z ,w z ,h z ) is the real frame coordinate, (x c ,y c ,w c ,h c ) is the fixed reference frame coordinate, (△x,△y,△w,△h) is the offset; In the real frame, the left eye key point, the right eye key point, the nose tip key point, the left corner of the mouth key point and the right corner of the mouth key point are preset and the key point coordinates (x i ,y i ), i=1,2,…,5, by aligning the matrix and the alignment vector (f, g) to transform the keypoint coordinates into the standard keypoint coordinates (x i ',y i '), i=1,2,…,5, the elements in the alignment matrix and the alignment vector are obtained by minimizing the error function, and the error function is , then the expression of the standard key point coordinates is (x i ',y i ')=E(x i ,y i )+(f,g); Face matching is performed using pre-entered face information to obtain the face recognition result.

3. The intelligent identification method according to claim 1, characterized in that: The external risk identification module is used to extract personnel anchor frames and smart helmet anchor frames through an improved feature extraction model and output the external risk information. The improved feature extraction model includes an input end, a backbone feature extraction network, a feature fusion network, a prediction network, and an output end.

4. The intelligent identification method according to claim 3, characterized in that: The obtaining of the user's own risk value based on the user's sensor data includes: The user sensing data is acquired through the sensor group, the sensor group includes a speed sensor, a temperature sensor, a particle sensor and a height sensor, the user sensing data includes a user moving speed, an ambient temperature, an ambient particle concentration, a user height difference, and a user's current height, the user height difference being a height change value of the user in a vertical direction within a preset time; Calculating the own risk value based on the user sensor data; When the own risk value is 1, the own high risk information is output, when the own risk value is 0.5, the own medium risk information is output, and when the own risk value is 0, the own low risk information is output.

5. The intelligent identification method according to claim 3, characterized in that: The improved feature extraction model specifically includes: The real-time construction site image is collected by the smart helmet and input into the input end, and the input end performs image preprocessing; The backbone feature extraction network includes a ShuffleNetv2 module and an SPPF module, wherein the ShuffleNetv2 module includes a 1×1 element-by-element group convolution, a Channel Shuffle channel shuffle mechanism, a 3×3 depth convolution, and a 1×1 element-by-element group convolution in sequence, and the SPPF includes three 5×5 maximum pooling layers connected in series; The feature fusion network includes an FPN feature pyramid structure and a PAN structure; The preset data set is used to train the improved feature extraction model, and the data set includes a training set, a test set and a validation set, which are a construction site portrait image set and a smart helmet image set; The output end outputs the personnel anchor frame and the smart helmet anchor frame, matches the personnel anchor frame and the smart helmet anchor frame that are closest to each other and obtains anchor frame data, wherein the anchor frame data includes the personnel anchor frame center coordinates (X, Y), the smart helmet anchor frame center coordinates (x, y), the personnel anchor frame height T, the smart helmet anchor frame height t, and the personnel anchor frame width R; The external risk value is calculated based on the anchor frame data, and the calculation formula is: , where β is the external risk value; When the external risk value is 1, the external high risk information is output, and when the external risk value is 0, the external low risk information is output.

6. The intelligent identification method according to claim 1, characterized in that: The construction site safety model is constructed as follows: When the output of the face verification is inconsistent, issuing a verification personnel instruction to the terminal; The working range of the construction site is preset, and when the position information of the personnel exceeds the working range of the construction site, the personnel leaving the post verification instruction is fed back to the terminal; When the self-high risk information or the self-medium risk information is output, a construction site safety confirmation instruction is fed back to the terminal; When the external high-risk information is output, the smart helmet wearing verification instruction is fed back to the terminal.

7. An intelligent identification device based on a construction site safety model, characterized in that: The device is applied to the intelligent recognition method as described in any one of claims 1 to 6, including an intelligent recognition module and a model building module; The intelligent recognition module is used to perform face recognition, personnel positioning and risk identification through a smart helmet. The smart helmet is embedded with a face recognition module, a real-time positioning module, a risk identification module and a sensor group. The face recognition module obtains a face recognition result through a face detection algorithm. The face recognition result includes a face verification pass and a face verification failure. The real-time positioning module obtains personnel location information through GPS positioning. The risk identification module includes an own risk identification module and an external risk identification module to obtain own risk information and external risk information. The own risk information includes own high risk information, own medium risk information and own low risk information. The external risk information includes external high risk information and external low risk information. The model building module is used to build a construction site safety model based on the face recognition results, the personnel location information, the own risk information and the external risk information.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the intelligent recognition method as described in any one of claims 1-6 is implemented.

9. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to perform the intelligent identification method as described in any one of claims 1 to 6 when executed by a computer processor.

Images