Employee work clothes detection method and device

Abstract

The invention relates to an employee work clothes detection method and apparatus. The method comprises the steps of obtaining region video data of a target region shot by a camera; identifying the regional video data through a safety helmet identification sub-algorithm to obtain safety helmet wearing results of each employee in the regional video data; identifying the regional video data through a work clothes identification sub-algorithm to obtain a work clothes wearing result of each employee in the regional video data; and if the safety helmet wearing result is a wearing abnormal result or the work clothes wearing result is a clothes abnormal result, generating a work clothes abnormal alarm. According to the invention, when the employee in the target area does not wear the clothes meeting the requirements of the work clothes, the work clothes abnormity alarm is generated, and the manpower resources required by the detection of the work clothes of the employee are saved.

Description

Technical Field

[0001] The present application relates to the technical field of clothing detection, and in particular, to a method and device for detecting the work clothing of employees. Background Art

[0002] Work clothing is the clothing customized for employees by many companies or factories. Among them, work clothing can not only help to establish the overall image of the enterprise, enhance the sense of belonging and collective honor of employees, and improve the cohesion of the enterprise, but also help to identify the positions and departments of employees. For factories, the work clothing of employees also has the effect of safety protection. For example, safety helmets that can protect the head, work clothes with functions such as anti-static, flame retardant, and reflective, etc. Therefore, for many factories, it is required that employees wear work clothing well when working. However, in the prior art, personnel are arranged to check one by one whether employees wear work clothing well in the work area or on the only way to enter the work area. This requires at least one person to be arranged in each workshop of the factory to be specifically responsible for checking whether work clothing is worn well, which is very wasteful of human resources. Summary of the Invention

[0003] Based on this, the purpose of the present application is to provide a method and device for detecting the work clothing of employees, which can overcome the deficiencies of the prior art.

[0004] To achieve the above purpose, the technical solution adopted in the present application is as follows:

[0005] A method for detecting the work clothing of employees includes:

[0006] Obtaining the regional video data of the target area captured by the camera;

[0007] Identifying the regional video data through the safety helmet recognition sub-algorithm to obtain the safety helmet wearing results of each employee in the regional video data;

[0008] Identifying the regional video data through the work clothing recognition sub-algorithm to obtain the work clothing wearing results of each employee in the regional video data;

[0009] If the safety helmet wearing result is an abnormal wearing result, or the work clothing wearing result is an abnormal clothing result, a work clothing abnormality alarm is generated.

[0010] As an implementation manner, the step of identifying the regional video data through the safety helmet recognition sub-algorithm to obtain the safety helmet wearing results of each employee in the regional video data includes:

[0011] Identifying the regional video data through the safety helmet recognition sub-algorithm to obtain the number of employees and the number of safety helmets in the regional video data;

[0012] If the number of the employees is greater than the number of the safety helmets, determine that the safety helmet wearing result is an abnormal wearing result.

[0013] As an implementation manner, after the step of identifying the regional video data through the safety helmet recognition sub-algorithm to obtain the number of employees and the number of safety helmets in the regional video data, it includes:

[0014] If the number of the employees is less than or equal to the number of the safety helmets, obtain the first person-helmet distance data between each employee and each safety helmet;

[0015] According to the first person-helmet distance data, determine the corresponding relationship between each employee and each safety helmet, and obtain the second person-helmet distance data corresponding to several employees and safety helmets;

[0016] Compare the second person-helmet distance data with a preset person-helmet distance data threshold. If all the second person-helmet distance data are less than or equal to the person-helmet distance data threshold, determine that the safety helmet wearing result is a normal wearing result; if there is second person-helmet distance data greater than the person-helmet distance data threshold, determine that the safety helmet wearing result is an abnormal wearing result.

[0017] As an implementation manner, the step of identifying the regional video data through the work clothes recognition sub-algorithm to obtain the work clothes wearing results of each employee in the regional video data includes:

[0018] Identify the regional video data through the work clothes recognition sub-algorithm to obtain the number of employees and the number of work clothes in the regional video data;

[0019] If the number of the employees is greater than the number of the work clothes, determine that the work clothes wearing result is an abnormal wearing result.

[0020] As an implementation manner, after the step of identifying the regional video data through the work clothes recognition sub-algorithm to obtain the number of employees and the number of work clothes in the regional video data, it includes:

[0021] If the number of the employees is less than or equal to the number of the work clothes, obtain the first person-clothes distance data between each employee and each work clothes;

[0022] According to the first person-clothes distance data, determine the corresponding relationship between each employee and each work clothes, and obtain the second person-clothes distance data corresponding to several employees and work clothes;

[0023] Compare the second person-clothing distance data with a preset person-clothing distance data threshold. If all the second person-clothing distance data are less than or equal to the person-clothing distance data threshold, determine that the work clothing wearing result is a normal wearing result; if there is second person-clothing distance data greater than the person-clothing distance data threshold, determine that the work clothing wearing result is an abnormal wearing result.

[0024] As an implementation manner, the work clothing is a reflective clothing, or a black jacket work clothing, or a blue long-sleeved shirt work clothing, or a blue short-sleeved shirt work clothing.

[0025] As an implementation manner, it further includes:

[0026] Identify the regional video data through a short-sleeved and short-pants recognition sub-algorithm to obtain the short-sleeved and short-pants wearing results of each employee in the regional video data;

[0027] If the short-sleeved and short-pants wearing result indicates that there is an employee wearing a short-sleeved top or shorts, generate a work clothing anomaly alarm.

[0028] As an implementation manner, it is applied to an edge server, and the edge server is connected to a plurality of cameras; the edge server receives a safety helmet recognition sub-algorithm issued by a superior server;

[0029] The steps for the edge server to identify the safety helmet wearing results of each employee in the regional video data through the safety helmet recognition sub-algorithm include:

[0030] The edge server identifies the regional video data through the safety helmet recognition sub-algorithm to obtain the safety helmet wearing results of each employee and the corresponding wearing result credibility;

[0031] The edge server determines whether the corresponding safety helmet wearing result has training effectiveness according to the wearing result credibility;

[0032] The edge server trains the safety helmet recognition sub-algorithm according to multiple safety helmet wearing results with training effectiveness and a preset number of training times to obtain the algorithm parameter difference before and after training;

[0033] The edge server uploads the training algorithm difference to the superior server, so that the superior server trains the safety helmet recognition sub-algorithm stored in the superior server according to the algorithm parameter differences uploaded by multiple edge servers connected to different cameras to update the safety helmet recognition sub-algorithm stored in the superior server;

[0034] The edge server receives the updated hard hat recognition sub-algorithm sent by the superior server, and identifies the regional video data captured by the camera according to the updated hard hat recognition sub-algorithm, so as to obtain the hard hat wearing results of each employee and the corresponding wearing result credibility.

[0035] As an implementation, the step that the edge server determines whether the corresponding hard hat wearing result has training effectiveness according to the wearing result credibility includes:

[0036] If the wearing result credibility is greater than a preset credibility threshold, the edge server determines that the corresponding hard hat wearing result has training effectiveness and stores the hard hat wearing result;

[0037] If the wearing result credibility is less than or equal to the credibility threshold, the edge server determines that the corresponding hard hat wearing result lacks training effectiveness, and the edge server uploads the hard hat wearing result lacking training effectiveness to the management terminal;

[0038] The edge server receives and stores the hard hat wearing result modified to have training effectiveness sent by the management terminal.

[0039] The second embodiment of the present application provides an employee work clothing detection device, including:

[0040] A video data acquisition module, configured to acquire regional video data of a target area captured by a camera;

[0041] A hard hat wearing result acquisition module, configured to identify the regional video data through a hard hat recognition sub-algorithm to obtain the hard hat wearing results of each employee in the regional video data;

[0042] A work clothing wearing result acquisition module, configured to identify the regional video data through a work clothing recognition sub-algorithm to obtain the work clothing wearing results of each employee in the regional video data;

[0043] A work clothing abnormality alarm module, configured to generate a work clothing abnormality alarm if the hard hat wearing result is a wearing abnormality result or the work clothing wearing result is a clothing abnormality result.

[0044] Compared with the traditional technology, the beneficial effects of the present application are:

[0045] This application can identify the regional video data through the safety helmet recognition sub-algorithm and the work clothing recognition sub-algorithm respectively, and obtain the safety helmet wearing results and work clothing wearing results of each employee in the regional video data. When the safety helmet wearing result is an abnormal wearing result or the work clothing wearing result is an abnormal clothing result, a work clothing abnormality alarm can be generated. This saves the human resources required for the detection of employees' work clothing. Moreover, this application can also identify the regional video data through the short-sleeved shirt and shorts recognition sub-algorithm, and when it detects that there are employees wearing short-sleeved shirts or shorts, a work clothing abnormality alarm is generated to prevent employees from entering the work area wearing short-sleeved shirts or shorts with low safety. Further, the employee work clothing detection method of this application is applied to the edge server, and the safety helmet recognition sub-algorithm can be trained according to multiple safety helmet wearing results with training effectiveness and a preset number of training times to obtain the difference in algorithm parameters before and after training, and then the training algorithm difference is uploaded to the superior server, so that the superior server trains the safety helmet recognition sub-algorithm stored in the superior server according to the algorithm parameter differences uploaded by multiple edge servers connected to different cameras, so as to update the safety helmet recognition sub-algorithm stored in the superior server, enabling the edge server and the superior server to share the training burden together, and allowing the edge server and the superior server to train the safety helmet recognition sub-algorithm in a state of low burden, improving the training efficiency of the safety helmet recognition sub-algorithm.

[0046] For better understanding and implementation, the present application will be described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a flowchart of the employee work clothing detection method according to an embodiment of the present application;

[0048] Figure 2 It is a first schematic diagram of the regional video data of the employee work clothing detection method according to an embodiment of the present application;

[0049] Figure 3 It is a second schematic diagram of the regional video data of the employee work clothing detection method according to an embodiment of the present application;

[0050] Figure 4 It is a flowchart of the work of the edge server of the employee work clothing detection method according to an embodiment of the present application;

[0051] Figure 5 It is a schematic diagram of the module connection of the employee work clothing detection device according to an embodiment of the present application.

[0052] 1. Video data acquisition module; 2. Safety helmet wearing result acquisition module; 3. Work clothing wearing result acquisition module; 4. Work clothing abnormality alarm module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] To make the objectives, technical solutions, and advantages of this application clearer, the following will further describe in detail the embodiments of this application in conjunction with the accompanying drawings.

[0054] It should be clear that the described embodiments are only a part of the embodiments of this application, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without creative efforts belong to the scope protected by the embodiments of this application.

[0055] When the following description involves the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. In the description of this application, it should be understood that the terms "first", "second", "third", etc. are only used to distinguish similar objects, and do not have to be used to describe a specific order or sequence, nor can they be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances. The singular forms "a", "the", and "said" used in this application and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. The words "if" / "when" used herein can be interpreted as "when...", "when...", or "in response to a determination".

[0056] In addition, in the description of this application, unless otherwise specified, "a plurality of" means two or more. "And / or" describes the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects before and after.

[0057] Please refer to Figure 1 , which is a flowchart of the employee work clothing detection method according to the first embodiment of this application, including:

[0058] S1: Obtain the regional video data of the target area captured by the camera;

[0059] S2: Identify the regional video data through the safety helmet recognition sub-algorithm to obtain the safety helmet wearing results of each employee in the regional video data.

[0060] The safety helmet recognition sub-algorithm is a deep learning algorithm and can be trained through multiple safety helmet algorithm training samples. Among them, the safety helmet algorithm training samples are marked with employees and safety helmets, as well as the wearing relationship between employees and safety helmets.

[0061] S3: Identify the regional video data through the work clothing recognition sub-algorithm to obtain the work clothing wearing results of each employee in the regional video data.

[0062] The work uniform recognition sub - algorithm is a deep - learning algorithm, which can be obtained by training with multiple work uniform algorithm training samples. Among them, the work uniform algorithm training samples are marked with employees, work uniforms, and the wearing relationship between employees and work uniforms. The work uniform is a reflective vest, or a black jacket work uniform, or a blue long - sleeve shirt work uniform, or a blue short - sleeve shirt work uniform.

[0063] S4: If the safety helmet wearing result is an abnormal wearing result, or the work uniform wearing result is an abnormal clothing result, generate a work clothing abnormality alarm.

[0064] Among them, the work clothing abnormality alarm can be an alarm by a speaker, a light alarm, or an alarm method such as sending an alarm message to the electronic device of the plant manager, etc.

[0065] In a feasible embodiment, the step S2: identifying the regional video data through the safety helmet recognition sub - algorithm to obtain the safety helmet wearing results of each employee in the regional video data includes:

[0066] S21: Identify the regional video data through the safety helmet recognition sub - algorithm to obtain the number of employees and the number of safety helmets in the regional video data.

[0067] S22: If the number of employees is greater than the number of safety helmets, determine that the safety helmet wearing result is an abnormal wearing result.

[0068] In a feasible embodiment, after the step S21: identifying the regional video data through the safety helmet recognition sub - algorithm to obtain the number of employees and the number of safety helmets in the regional video data, it includes:

[0069] S23: If the number of employees is less than or equal to the number of safety helmets, obtain the first - person hat - distance data between each employee and each safety helmet.

[0070] S24: According to the first - person hat - distance data, determine the corresponding relationship between each employee and each safety helmet to obtain the second - person hat - distance data corresponding to several employees and safety helmets.

[0071] Among them, from the first - person hat - distance data between an employee and each safety helmet, the safety helmet corresponding to the employee can be determined according to the first - person hat - distance data with the smallest value; or from the first - person hat - distance data between a safety helmet and each employee, the employee corresponding to the safety helmet can be determined according to the first - person hat - distance data with the smallest value. When multiple employees correspond to the same safety helmet, or multiple safety helmets correspond to the same employee, determine that the safety helmet wearing result is an abnormal wearing result.

[0072] S25: Compare the second person-helmet distance data with a preset person-helmet distance data threshold. If all of the second person-helmet distance data is less than or equal to the person-helmet distance data threshold, determine that the safety helmet wearing result is a normal wearing result; if there is second person-helmet distance data greater than the person-helmet distance data threshold, determine that the safety helmet wearing result is an abnormal wearing result.

[0073] Among them, the person-helmet distance data threshold can be set by users such as the plant administrator and the plant safety officer.

[0074] In a feasible embodiment, the step S3 of identifying the regional video data through the work uniform recognition sub-algorithm to obtain the work uniform wearing results of each employee in the regional video data includes:

[0075] S31: Identify the regional video data through the work uniform recognition sub-algorithm to obtain the number of employees and the number of work uniforms in the regional video data.

[0076] S32: If the number of employees is greater than the number of work uniforms, determine that the work uniform wearing result is an abnormal wearing result.

[0077] In a feasible embodiment, after the step S31 of identifying the regional video data through the work uniform recognition sub-algorithm to obtain the number of employees and the number of work uniforms in the regional video data, it includes:

[0078] S33: If the number of employees is less than or equal to the number of work uniforms, obtain the first person-clothing distance data between each employee and each work uniform.

[0079] S34: According to the first person-clothing distance data, determine the corresponding relationship between each employee and each work uniform to obtain the second person-clothing distance data corresponding to several employees and work uniforms.

[0080] Among them, from the first person-clothing distance data between an employee and each work uniform, the work uniform corresponding to the employee can be determined according to the first person-clothing distance data with the smallest value; or from the first person-clothing distance data between a work uniform and each employee, the employee corresponding to the work uniform can be determined according to the first person-clothing distance data with the smallest value. When multiple employees correspond to the same work uniform, or multiple work uniforms correspond to the same employee, determine that the work uniform wearing result is an abnormal wearing result.

[0081] S35: Compare the second person-clothing distance data with a preset person-clothing distance data threshold. If all the second person-clothing distance data are less than or equal to the person-clothing distance data threshold, determine that the work clothing wearing result is a normal wearing result; if there is second person-clothing distance data greater than the person-clothing distance data threshold, determine that the work clothing wearing result is an abnormal wearing result.

[0082] Among them, the person-clothing distance data threshold can be set by users such as the plant administrator and the plant safety officer.

[0083] In a feasible embodiment, it further includes:

[0084] Identify the regional video data through the short-sleeved shirt and shorts recognition sub-algorithm to obtain the short-sleeved shirt and shorts wearing results of each employee in the regional video data;

[0085] If the short-sleeved shirt and shorts wearing result indicates that there is an employee wearing a short-sleeved top or shorts, generate a work clothing anomaly alarm.

[0086] This application can identify the regional video data through the safety helmet recognition sub-algorithm and the work clothing recognition sub-algorithm respectively to obtain the safety helmet wearing results and work clothing wearing results of each employee in the regional video data. When the safety helmet wearing result is an abnormal wearing result, or the work clothing wearing result is an abnormal clothing result, a work clothing anomaly alarm can be generated. This saves the human resources required for the detection of employees' work clothing. Moreover, this application can also identify the regional video data through the short-sleeved shirt and shorts recognition sub-algorithm, and then generate a work clothing anomaly alarm when it detects that there is an employee wearing a short-sleeved top or shorts, preventing employees from entering the work area wearing short-sleeved tops or shorts with low safety.

[0087] Next, a specific implementation manner of the safety helmet recognition sub-algorithm will be exemplified in conjunction with Figure 2 For example, in the situation shown in Figure 2 , three people P 1 , P 2 , P 3 and three safety helmets H 1 , H 2 , H 3 are detected. First, match the people and the safety helmets, and calculate the distances between them respectively. The distance between the first person P 1 and the first safety helmet H 1 is, and the distance between the first person P 1 and the second safety helmet H 2 is H 12 And so on to obtain each H ij . Finally, a distance matrix is formed:

[0088]

[0089] Specifically, the data of the distance matrix is as follows:

[0090]

[0091] First, find the element with the smallest value in each row, and then subtract this minimum value from all elements in that row. Then the distance matrix becomes:

[0092]

[0093] Then, for each column, find the minimum value in that column, and then subtract this minimum value from all numbers in that column. Then the distance matrix becomes:

[0094]

[0095] Next, use the fewest horizontal lines and the fewest vertical lines to cover all the 0 elements in the matrix. Then, among these 0 elements, there is an optimal solution, and the algorithm ends. It can be seen that in this example, the lines covering the zero elements are as follows:

[0096]

[0097] Finally, judge that the matching of the person and the safety helmet is P 1 →H 1 , P 2 →H 2 , P 3 →H 3 , and the corresponding distance is H 11 , H 22 , H 33 , if H 22 in these distances is greater than the preset threshold of the person-helmet distance data, an alarm is generated.

[0098] Next, an implementation manner of the work uniform recognition sub-algorithm will be illustrated by combining Figure 3 :

[0099] As shown in the figure, three people P 1 , P 2 , P 3 and three reflective vests V 1 , V 2 , V 3 are detected. First, match the people and the reflective vests, and calculate the distances between them respectively. The distance between the first person P 1 and the first reflective vest V 1 is H 11 , and the distance between the first person P 1 and the second reflective vest V 2 is H 12And so on to obtain each H ij Finally, a distance matrix is formed:

[0100]

[0101] Specifically, the data of the distance matrix is as follows:

[0102]

[0103] First, find the element with the smallest value in each row, and then subtract this minimum value from all elements in that row. Then the distance matrix becomes:

[0104]

[0105] Then, for each column, find the minimum value in that column, and then subtract this minimum value from the numbers in that column. Then the distance matrix becomes:

[0106]

[0107] Next, use the fewest horizontal lines and the fewest vertical lines to cover all the 0 elements of the matrix. Then, among these 0 elements, there is an optimal solution, and the algorithm ends. It can be seen that in this example, the lines covering the zero elements are as follows:

[0108]

[0109] Finally, judge that the matching between the person and the safety helmet is P 1 →V 1 ,P 2 →V 2 ,P 3 →V 3 ,and the corresponding distances are H 11 ,H 22 ,H 33 If any of these distances is greater than the preset threshold of the person-clothing distance data, an alarm is generated.

[0110] Please refer to Figure 4 In a feasible embodiment, the employee work clothing detection method of the present application is applied to an edge server, and the edge server is connected to a plurality of cameras; the edge server receives the safety helmet recognition sub-algorithm issued by the superior server;

[0111] The step of the edge server identifying the regional video data through the safety helmet recognition sub-algorithm to obtain the safety helmet wearing results of each employee in the regional video data includes:

[0112] S100: The edge server identifies the regional video data through the safety helmet recognition sub-algorithm, obtains the safety helmet wearing results of each employee, and the corresponding wearing result credibility.

[0113] S200: The edge server determines whether the corresponding safety helmet wearing result has training effectiveness according to the wearing result credibility.

[0114] S300: The edge server trains the safety helmet recognition sub-algorithm according to multiple safety helmet wearing results with training effectiveness and a preset number of training times, and obtains the difference in algorithm parameters before and after training.

[0115] The difference in the first model parameters can be obtained through the following formula:

[0116] p m =w 0 -w

[0117] where p m is the difference in algorithm parameters, w 0 is the safety helmet recognition sub-algorithm of the edge server before training, and w is the safety helmet recognition sub-algorithm of the edge server after training.

[0118] S400: The edge server uploads the training algorithm difference to the superior server, so that the superior server trains the safety helmet recognition sub-algorithm stored in the superior server according to the algorithm parameter differences uploaded by multiple edge servers connected to different cameras, so as to update the safety helmet recognition sub-algorithm stored in the superior server.

[0119] S500: The edge server receives the updated safety helmet recognition sub-algorithm sent by the superior server, and identifies the regional video data captured by the camera according to the updated safety helmet recognition sub-algorithm, obtains the safety helmet wearing results of each employee, and the corresponding wearing result credibility.

[0120] In a feasible embodiment, the step of S200: the edge server determines whether the corresponding safety helmet wearing result has training effectiveness according to the wearing result credibility includes:

[0121] S210: If the wearing result credibility is greater than a preset credibility threshold, the edge server determines that the corresponding safety helmet wearing result has training effectiveness and stores the safety helmet wearing result.

[0122] where the credibility threshold can be set by users such as the factory building administrator and the factory building safety officer.

[0123] S220: If the credibility of the wearing result is less than or equal to the credibility threshold, the edge server determines that the corresponding safety helmet wearing result lacks training effectiveness, and the edge server uploads the safety helmet wearing result lacking training effectiveness to the management terminal.

[0124] Among them, the safety helmet wearing result at least includes an image marked with the detection object and the attribute of lacking training effectiveness. The management terminal can be an electronic device used by the factory building administrator or the factory building safety officer, such as a smart phone, a computer, etc. The management terminal can display the image corresponding to the safety helmet wearing result and the attribute of lacking training effectiveness. The user can also modify the attribute of lacking training effectiveness to having training effectiveness through the management terminal and then send it to the edge server.

[0125] S230: The edge server receives and stores the safety helmet wearing result modified to have training effectiveness sent by the management terminal.

[0126] Further, the employee work clothing detection method of the present application is applied to the edge server. It can train the safety helmet recognition sub-algorithm according to multiple safety helmet wearing results with training effectiveness and a preset number of training times, obtain the difference in algorithm parameters before and after training, and then upload the training algorithm difference to the upper-level server, so that the upper-level server trains the safety helmet recognition sub-algorithm stored in the upper-level server according to the algorithm parameter differences uploaded by multiple edge servers connected to different cameras, so as to update the safety helmet recognition sub-algorithm stored in the upper-level server, enabling the edge server and the upper-level server to share the training burden together, and allowing the edge server and the upper-level server to train the safety helmet recognition sub-algorithm in a state with a small burden, improving the training efficiency of the safety helmet recognition sub-algorithm.

[0127] Please refer to Figure 5 , the second embodiment of the present application provides an employee work clothing detection device, including:

[0128] A video data acquisition module 1, configured to acquire regional video data of a target area captured by a camera;

[0129] A safety helmet wearing result acquisition module 2, configured to identify the regional video data through a safety helmet recognition sub-algorithm to obtain the safety helmet wearing results of each employee in the regional video data;

[0130] A work clothing wearing result acquisition module 3, configured to identify the regional video data through a work clothing recognition sub-algorithm to obtain the work clothing wearing results of each employee in the regional video data;

[0131] The work clothing abnormality alarm module 4 is used to generate a work clothing abnormality alarm if the helmet wearing result is an abnormal wearing result or the work clothing wearing result is a clothing abnormality result.

[0132] It should be noted that when the employee work clothing detection device provided in the second embodiment of the present application executes the employee work clothing detection method, only the above-mentioned division of each functional module is used as an example for illustration. In actual application, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the employee work clothing detection device provided in the second embodiment of the present application and the employee work clothing detection method of the first embodiment of the present application belong to the same concept, and the implementation process is detailed in the method embodiment and will not be repeated here.

[0133] The third aspect of the embodiments of the present application provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the employee work clothing detection method described above are implemented.

[0134] The fourth aspect of the embodiments of the present application provides a computer device, including a storage, a processor, and a computer program stored in the storage and executable by the processor. When the processor executes the computer program, the steps of the employee work clothing detection method described above are implemented.

[0135] The device embodiments described above are merely illustrative. The components described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the present application. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0136] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0137] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing device produce means for implementing the selected functions in the process Figure 1 one process or multiple processes and / or blocks Figure 1 These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufacture including instruction means that implement the selected functions in the process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks

[0138] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the selected functions in the process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks

[0139] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and a memory.

[0140] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of computer-readable media.

[0141] A computer-readable medium includes both permanent and non-permanent, removable and non-removable media that can implement information storage by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic tape disk storage or other magnetic storage devices, or any other non-transitory medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer-readable medium does not include transitory computer-readable media such as modulated data signals and carrier waves.

[0142] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or apparatus comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or apparatus. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising the element.

[0143] The above are only embodiments of the present application and are not used to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. A method for detecting employee work clothes, characterized in that: include: Get the regional video data of the target area captured by the camera; Identify the regional video data by using a helmet identification sub-algorithm to obtain helmet wearing results of each employee in the regional video data; Identify the regional video data by using a work clothes identification sub-algorithm to obtain the work clothes wearing results of each employee in the regional video data; If the helmet wearing result is an abnormal wearing result, or the work clothes wearing result is an abnormal clothing result, an abnormal work clothing alarm is generated.

2. The employee work clothing detection method according to claim 1, characterized in that: The step of identifying the regional video data by using the helmet recognition sub-algorithm to obtain the helmet wearing results of each employee in the regional video data includes: Identify the regional video data by using the helmet identification sub-algorithm to obtain the number of employees and the number of helmets in the regional video data; If the number of employees is greater than the number of helmets, it is determined that the helmet wearing result is an abnormal wearing result.

3. The employee work clothing detection method according to claim 2, characterized in that: After the step of identifying the regional video data by the helmet identification sub-algorithm and obtaining the number of employees and the number of helmets in the regional video data, the method includes: If the number of employees is less than or equal to the number of helmets, obtaining first person-hat distance data between each employee and each helmet; According to the first person-hat distance data, the corresponding relationship between each employee and each safety helmet is determined to obtain the second person-hat distance data corresponding to a plurality of employees and the safety helmets; The second person-hat distance data is compared with a preset person-hat distance data threshold. If all of the second person-hat distance data are less than or equal to the person-hat distance data threshold, it is determined that the safety helmet wearing result is a normal wearing result; if there is any second person-hat distance data greater than the person-hat distance data threshold, it is determined that the safety helmet wearing result is an abnormal wearing result.

4. The employee work clothing detection method according to claim 1, characterized in that: The step of identifying the regional video data by using the work clothes identification sub-algorithm to obtain the work clothes wearing results of each employee in the regional video data includes: Identify the regional video data by using the work clothes identification sub-algorithm to obtain the number of employees and the number of work clothes in the regional video data; If the number of employees is greater than the number of work clothes, it is determined that the work clothes wearing result is an abnormal wearing result.

5. The employee work clothing detection method according to claim 2, characterized in that: After the step of identifying the regional video data by the work clothes identification sub-algorithm and obtaining the number of employees and the number of work clothes in the regional video data, the method includes: If the number of employees is less than or equal to the number of work clothes, first person-clothes distance data between each employee and each work clothes is obtained; According to the first person-clothing distance data, the corresponding relationship between each employee and each work clothes is determined to obtain the second person-clothing distance data corresponding to a plurality of employees and work clothes; The second person-clothing distance data is compared with a preset person-clothing distance data threshold. If all of the second person-clothing distance data are less than or equal to the person-clothing distance data threshold, it is determined that the work clothes wearing result is a normal wearing result; if there is any second person-clothing distance data greater than the person-clothing distance data threshold, it is determined that the work clothes wearing result is an abnormal wearing result.

6. The employee work clothing detection method according to claim 5, characterized in that: The work clothes are reflective clothing, or black jacket work clothes, or blue long-sleeved shirt work clothes, or blue short-sleeved shirt work clothes.

7. The employee work clothing detection method according to claim 1, characterized in that: Also includes: Identify the regional video data by using a short-sleeved shirt and shorts identification sub-algorithm to obtain the short-sleeved shirt and shorts wearing results of each employee in the regional video data; If the short-sleeved shirt or shorts wearing result indicates that an employee is wearing a short-sleeved shirt or shorts, an abnormal work clothing alarm is generated.

8. The employee work clothing detection method according to claim 1, characterized in that: Applied to an edge server, the edge server is connected to a plurality of cameras; the edge server receives a helmet recognition sub-algorithm sent by an upper-level server; The step of the edge server identifying the regional video data by using a helmet identification sub-algorithm to obtain the helmet wearing result of each employee in the regional video data includes: The edge server identifies the regional video data through a helmet identification sub-algorithm to obtain the helmet wearing results of each employee and the corresponding wearing result credibility; The edge server determines whether the corresponding helmet wearing result has training validity according to the wearing result credibility; The edge server trains the helmet recognition sub-algorithm according to the plurality of helmet wearing results with training validity and a preset number of training times to obtain a difference in algorithm parameters before and after training; The edge server uploads the training algorithm difference to the upper-level server, so that the upper-level server trains the helmet recognition sub-algorithm stored in the upper-level server according to the algorithm parameter difference uploaded by multiple edge servers connected to different cameras, so as to update the helmet recognition sub-algorithm stored in the upper-level server; The edge server receives the updated helmet recognition sub-algorithm sent by the upper server, and identifies the regional video data captured by the camera according to the updated helmet recognition sub-algorithm to obtain the helmet wearing results of each employee and the corresponding wearing result credibility.

9. The employee work clothing detection method according to claim 8, characterized in that: The step of the edge server determining whether the corresponding helmet wearing result has training validity according to the wearing result credibility includes: If the credibility of the wearing result is greater than a preset credibility threshold, the edge server determines that the corresponding helmet wearing result has training validity and stores the helmet wearing result; If the wearing result credibility is less than or equal to the credibility threshold, the edge server determines that the corresponding helmet wearing result lacks training effectiveness, and the edge server uploads the helmet wearing result lacking training effectiveness to the management terminal; The edge server receives and stores the helmet wearing result sent by the management terminal and modified to have training validity.

10. A device for detecting employee work clothes, characterized in that: include: A video data acquisition module is used to acquire regional video data of a target area captured by a camera; A helmet wearing result acquisition module is used to identify the regional video data through a helmet recognition sub-algorithm to obtain the helmet wearing result of each employee in the regional video data; A work clothes wearing result acquisition module, used to identify the regional video data through a work clothes recognition sub-algorithm, and obtain the work clothes wearing result of each employee in the regional video data; The work clothing abnormality alarm module is used to generate a work clothing abnormality alarm if the helmet wearing result is a wearing abnormality result, or the work clothing wearing result is a clothing abnormality result.

Images