Workshop supervision decision method, device and computer readable storage medium

Abstract

The present application discloses a workshop supervision decision method, device and computer-readable storage medium, which belongs to the field of production supervision technology, and discloses a workshop supervision decision method, including: when a target person who is not wearing safety protection equipment is identified based on a monitoring image, the personnel information of the target person is determined according to the personnel database and the facial information in the monitoring image; according to the time when the target person is first recorded not wearing the safety protection equipment and the current time, the violation duration of the target person not wearing the safety protection equipment is determined; according to the personnel information and the violation duration, the judgment decision result is determined and output. Through the above steps, it is possible to make up for the gap in previous supervision decisions that can only provide early warnings but cannot directly give judgment decision results.

Description

Technical Field

[0001] The present application relates to the technical field of production supervision, and in particular to a workshop supervision decision method, device and computer-readable storage medium. Background Art

[0002] Some production workshops usually have great safety risks due to factors such as the large variety of production equipment, heavy equipment, and complex environment. In response to this, manufacturers often use safety identification and decision-making processing to supervise the safety of production workshops. In the traditional safety identification and decision-making processing method, artificial intelligence technology is used to identify potential safety hazards or abnormal situations, and corresponding decisions can be made automatically. However, the decision-making means of this method can often only issue alarms or warnings, lack intelligent decision-making penalties, and therefore it is difficult to cover more complex safety supervision scenarios.

[0003] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0004] The main purpose of this application is to provide a workshop supervision decision-making method, device and computer-readable storage medium, aiming to solve the technical problem that previous decision-making methods can only issue alarms or warnings, lack intelligent decision-making penalties, and are therefore difficult to cover more complex safety supervision scenarios.

[0005] To achieve the above objectives, the present application proposes a workshop supervision decision-making method, wherein the method comprises:

[0006] When a target person who is not wearing safety protection equipment is identified based on the surveillance image, the personnel information of the target person is determined according to the personnel database and the facial information in the surveillance image;

[0007] Determine the violation duration of the target person not wearing the safety protection equipment according to the time when the target person is first recorded not wearing the safety protection equipment and the current time;

[0008] According to the personnel information and the duration of the violation, a determination and decision result is determined and output.

[0009] In one embodiment, before the step of determining the personnel information of the target person according to the personnel database and the facial information in the surveillance image, the step further includes:

[0010] Extracting a person image of a target person who is not wearing the safety protection equipment in the monitoring image based on the safety protection equipment recognition model;

[0011] Extracting image features of the person image based on a face recognition model;

[0012] According to the image features, corresponding face information is determined.

[0013] In one embodiment, the step of determining and outputting a determination decision result based on the personnel information and the violation duration includes:

[0014] Matching corresponding workshop work rules and regulations according to the personnel information and the keywords of the violation duration;

[0015] Determine the corresponding numerical indicator data according to the workshop work rules and regulations;

[0016] Based on the personnel information, the duration of the violation and the digital indicator data, the violation behavior of the target person is classified and the judgment decision result is determined.

[0017] In one embodiment, when a target person who is not wearing safety protection equipment is identified based on a surveillance image, before the step of determining the personnel information of the target person according to a personnel database and the facial information in the surveillance image, the step further includes:

[0018] Based on the employee facial information and safety protection equipment data set of the workshop, a corresponding training set and a verification set are divided and noise reduction processing is performed on the training set and the verification set;

[0019] Based on the processed training set and verification set of the safety protection equipment, the initial safety protection equipment recognition model is trained to obtain the safety protection equipment recognition model;

[0020] Based on the processed training set and verification set of the employee's facial information, the initial face recognition model is trained to obtain a face recognition model.

[0021] In one embodiment, the step of training the initial safety protection equipment identification model based on the processed training set and verification set of the safety protection equipment to obtain the safety protection equipment identification model includes:

[0022] Inputting the training set into the input end of the initial safety protection equipment recognition model in batches, and performing image data enhancement processing;

[0023] Extracting image features of the image data based on a neural network of the initial safety protection equipment recognition model;

[0024] After the initial safety protection equipment identification model outputs a predicted value, the predicted value and the label value are calculated based on a preset loss function, and a weight parameter of the initial safety protection equipment identification model is adjusted according to the calculation result;

[0025] The initial safety protection equipment identification model is verified based on the verification set, and when the test loss is minimum, the weight parameter is retained to obtain the safety protection equipment identification model.

[0026] In one embodiment, the step of training the initial face recognition model based on the processed training set and verification set of the employee face information to obtain the face recognition model includes:

[0027] Converting the image data of the training set into corresponding sequence data, and inputting the sequence data into the initial face recognition model;

[0028] Based on the encoder and decoder of the initial face recognition model, the sequence data is linearly transformed and an output sequence is generated;

[0029] Calculating an output score based on the output sequence, and adjusting parameters of the encoder and decoder according to the output score;

[0030] The adjusted encoder and decoder are verified based on the verification set to obtain the face recognition model.

[0031] In one embodiment, after the step of determining and outputting the determination decision result according to the personnel information and the violation duration, the method further includes:

[0032] Submitting the determination decision result to the confidence evaluation node;

[0033] Based on the confidence evaluation node, a confidence evaluation is performed on the personnel information and violation duration of the decision-making result;

[0034] When the evaluation score is higher than a preset threshold, the determination decision result is uploaded to the performance system.

[0035] In one embodiment, after the step of determining and outputting the determination decision result according to the personnel information and the violation duration, the method further includes:

[0036] Obtain the equipment type of the target person not wearing safety protection equipment;

[0037] generating a risk impact assessment result based on the device type, the personnel information and the violation data;

[0038] Formulate corresponding control measures according to the risk impact assessment results and the determination and decision results;

[0039] Feedback the control measures to the corresponding supervisory personnel.

[0040] In addition, to achieve the above-mentioned purpose, the present application also proposes a workshop supervision decision-making device, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the computer program is configured to implement the steps of the workshop supervision decision-making method as described above.

[0041] In addition, to achieve the above-mentioned purpose, the present application also proposes a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the workshop supervision decision-making method as described above are implemented.

[0042] The present application proposes a workshop supervision decision method. When a target person who is not wearing safety protection equipment is identified based on a monitoring image, the facial information in the monitoring image is matched with the personnel information stored in the personnel database to determine the personnel information of the target person. Furthermore, the violation duration of the target person not wearing safety protection equipment is determined based on the time when the target person is first recorded not wearing safety protection equipment and the current time. Finally, the judgment decision result is determined and output based on the personnel information and the violation duration. Through the above steps, the gap in previous supervision decisions that can only provide early warnings but cannot directly provide judgment decision results can be made can be made up. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0045] Figure 1 A flowchart of the first embodiment of the workshop supervision decision-making method for this application;

[0046] Figure 2 A flow chart of the second embodiment of the workshop supervision decision-making method for this application;

[0047] Figure 3 A flow chart of the third embodiment of the workshop supervision decision-making method of this application;

[0048] Figure 4 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the workshop supervision decision-making method in the embodiment of the present application.

[0049] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0050] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0051] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0052] The main solution of the embodiment of the present application is: when a target person who is not wearing safety protection equipment is identified based on a surveillance image, the personnel information of the target person is determined according to the personnel database and the facial information in the surveillance image; the violation duration of the target person not wearing the safety protection equipment is determined according to the time when the target person is first recorded as not wearing the safety protection equipment and the current time; and the judgment decision result is determined and output based on the personnel information and the violation duration.

[0053] Some production workshops usually have great safety risks due to factors such as the large variety of production equipment, heavy equipment, and complex environment. In response to this, manufacturers often use safety identification and decision-making processing to supervise the safety of production workshops. In the traditional safety identification and decision-making processing method, artificial intelligence technology is used to identify potential safety hazards or abnormal situations, and corresponding decisions can be made automatically. However, the decision-making means of this method can often only issue alarms or warnings, lack intelligent decision-making penalties, and therefore it is difficult to cover more complex safety supervision scenarios.

[0054] A workshop supervision decision method proposed in this application can make up for the gap in previous supervision decisions that can only provide early warnings but cannot directly give judgment and decision results.

[0055] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a workshop supervision decision device, etc. The following takes the workshop supervision decision device as an example to illustrate this embodiment and the following embodiments.

[0056] Based on this, the embodiment of the present application provides a workshop supervision decision method, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the workshop supervision decision-making method of the present application.

[0057] In this embodiment, the workshop supervision decision-making method includes steps S10 to S30:

[0058] Step S10: When a target person who is not wearing safety protection equipment is identified based on a surveillance image, the personnel information of the target person is determined according to a personnel database and the facial information in the surveillance image.

[0059] In this embodiment, the monitoring image is an image collected by a camera deployed in the production workshop, and the safety protection equipment refers to equipment and instruments used to protect the safety of employees, such as safety helmets, eye protection equipment, protective masks, etc. The personnel database stores the personnel information of all the staff members in the current workshop, including name, position, work number, department, employment, etc. The monitoring image is pre-trained with a safety protection equipment recognition model, which can identify the target personnel in the picture who are not wearing safety protection equipment, and then recognize the facial information of the person who is not wearing safety protection equipment in the image, and match it with the personnel information in the preset personnel database, so as to determine the personnel information of the target personnel who are not currently wearing safety protection equipment.

[0060] Further, in a feasible implementation, for the identification of personnel who are not wearing safety protection equipment, a safety protection equipment identification model pre-built based on the YOLOv5 (a visual recognition algorithm name) algorithm can be used. The YOLOv5 algorithm is an efficient target detection algorithm that can simplify the target detection task into regression and classification problems to achieve rapid detection. The safety protection equipment identification model is trained by pre-labeled safety protection equipment at different angles, light conditions and scenes, monitoring images of people who are not wearing safety protection equipment, and video clips. The trained safety protection equipment identification model can extract image features from images captured by the workshop monitoring camera, and can identify the target personnel who are not wearing safety helmets in the workshop based on the features. Further, the helmet recognition algorithm based on YOLOv5 is combined with the Transform (a neural network name) face recognition algorithm to train a face recognition model, extract face information for personnel who are not wearing safety protection equipment, and compare the face information with the face information pre-stored in the personnel database to determine the personnel information of the target personnel, including name, title, department and other information, so as to achieve the purpose of identifying the identity of personnel who are not wearing safety protection equipment in the workshop.

[0061] In this embodiment, by analyzing the monitoring images of the workshop, it is possible to identify the target personnel in the workshop who are not wearing safety protection equipment, so that in the subsequent process, corresponding judgment decisions can be made based on the personnel information of the target personnel.

[0062] Step S20: Determine the violation duration of the target person not wearing the safety protection equipment according to the time when the target person is first recorded not wearing the safety protection equipment and the current time.

[0063] Step S30: Determine and output a decision result based on the personnel information and the violation duration.

[0064] In this embodiment, whenever the personnel information of the target personnel who are not wearing safety protection equipment is identified, the identification result and identification time will be recorded through the background database or server. According to the violation time of the current target personnel recorded for the first time and the current time, the violation duration of the target personnel not wearing safety protection equipment can be known. According to the personnel information and the violation duration, the corresponding keywords are extracted, and the keywords are matched with the workshop's operating rules and regulations to determine the operating rules and regulations used to make decisions on the current violation and the digital indicator data corresponding to the system, and the personnel information and the violation duration are compared with the digital indicator data to generate and output the judgment decision result for the target personnel.

[0065] Further, in a feasible implementation, a database or data storage system can be created in advance to record the information of the target personnel and the wearing of their safety protection equipment. Every time a violation is found, the violation time and personnel information are recorded. For each target person, the system calculates the violation duration of not wearing safety protection equipment by the difference between the first recorded violation time and the current time. When making judgments and decisions based on the violation duration and the violators, the work rules and regulations of the workshop are relied upon. The work rules and regulations record the requirements for safety regulations, behavioral norms and disciplinary sanctions in the workshop, in order to facilitate accurate judgments and decisions on target personnel who do not wear safety protection equipment. This application converts the work rules and regulations into digital indicator data and stores them in association with the corresponding decision results. When generating a judgment decision for a target person who does not wear safety protection equipment, the personnel information and the violation duration are compared with the digital indicator data to obtain the judgment decision result for the target person.

[0066] Furthermore, the various clauses in the work rules and regulations are sorted out and defined to ensure that each provision can be converted into a quantifiable indicator or condition, such as the requirements for wearing safety protection equipment, specific specifications for the workplace, etc. According to the content of the rules and regulations, corresponding digital evaluation standards are formulated. For example, for the requirements for wearing safety protection equipment, a binary indicator is set to indicate whether it is worn (1 means worn, 0 means not worn). According to the rules and regulations and digital evaluation standards, a decision model is established. The model can classify and judge employees according to their behavior or situation, such as whether they meet the requirements of the rules and regulations, and store the various clauses of the rules and regulations, digital evaluation standards and decision models in association. Using a database or a similar data management system, the rules and regulations are converted into tables or entries in the database, and are stored in association with the corresponding decision model results. When a decision is required for a target person, the relevant information is input into the decision model. The model will output the corresponding decision result according to the digital indicator data and the requirements of the rules and regulations. The decision result includes the punishment result and other contents. Optionally, the decision results output by employees with different positions and different violation durations will also be different. For example, for ordinary jobs with a violation duration within a certain range, the decision result can be to notify the corresponding management personnel and give the target personnel a verbal warning or a written warning, while for serious violations, fines, suspensions, and other penalties can be imposed. Optionally, based on the target personnel's information and violation duration, the system determines the decision result based on the matching rules and regulations, and then uploads the result to the personnel performance appraisal system and outputs it to the relevant responsible person or the target personnel themselves. The relevant responsible person can take corresponding measures, such as warnings, fines, etc., to ensure the implementation of safety regulations.

[0067] Optionally, in order to ensure the compliance of the decision-making results, the duration of violations, frequency of violations, identity information of violators, etc. can also be collected in advance. Through data analysis, the severity and trend of violations can be understood to provide a basis for decision-making. According to the current rules and regulations, determine the punishment measures applicable to the incident. Consider the duration of the violation, the historical records of the violators, and the specific circumstances of the incident to determine the most appropriate punishment plan. In the subsequent process, supervise the implementation of the punishment measures and regularly follow up on the improvement of the violation situation. If it is found necessary to adjust the punishment measures or strengthen the supervision measures, make adjustments in a timely manner.

[0068] When the face recognition model completes the identification of the person, the identification result is submitted to the confidence judgment node. The confidence judgment node can be an independent model or algorithm used to evaluate the confidence of the face recognition result. After the confidence judgment is completed, the judgment score is compared with the preset threshold. When the judgment score is higher than the preset threshold, the judgment decision result is uploaded to the performance system. Furthermore, as a feasible implementation method, streaming processing technology can be used to receive and process face recognition results in real time, and the results are passed to the confidence judgment node. Establish a machine learning model of the confidence judgment node, such as a neural network-based model or a probabilistic graph model, to evaluate the credibility of the recognition results. Design a rule engine to judge personnel information, violation duration, etc. according to specific rules. Upload the judgment decision results to the performance system through the API to ensure that the judgment decision results can be recorded and processed in a timely manner.

[0069] In this embodiment, through the above steps, the gap in previous supervision decisions, which can only provide early warnings but cannot directly give judgment and decision results, can be made up. The level of intelligent decision-making for unsafe behaviors in the workshop is greatly improved, the safety of personnel is improved, and the safety production standards of enterprises are strengthened.

[0070] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 2 The second embodiment of the method for adjusting the level difficulty of the present application, before step S10, further includes the following steps:

[0071] Step S40: based on the employee facial information and safety protection equipment data set of the workshop, divide the corresponding training set and verification set and perform noise reduction processing on the training set and the verification set;

[0072] Step S50: training the initial safety protection equipment recognition model based on the processed training set and verification set of the safety protection equipment to obtain a safety protection equipment recognition model;

[0073] Step S60: training the initial face recognition model based on the processed training set and verification set of the employee face information to obtain a face recognition model.

[0074] In this embodiment, the safety protection equipment recognition model is used to identify personnel who are not wearing safety protection equipment in the monitoring image, and the face recognition model is used to perform face recognition on the target personnel who are not wearing safety protection equipment to determine the corresponding personnel information. In the process of building the safety protection equipment model, the face information of the employees in the workshop is collected, and the helmet wearing data set is used to divide the two parts of the data set into corresponding training sets and verification sets according to the proportion. The division ratio can be 70% training set and 30% verification set. At the same time, the two parts of the data are labeled to ensure that each image has a corresponding helmet position label. The two parts of the data set are image preprocessed, and the Gaussian filtering algorithm is used to reduce the noise of the image data to facilitate the training of the later image recognition algorithm and improve the image recognition accuracy. Use the preprocessed helmet dataset to train the YOLOv5 algorithm. Input the dataset in batches to the input of the initial safety protection equipment recognition model initially constructed with the YOLOv5 algorithm. Perform random scaling, random cropping, and random arrangement to enhance the image data. Then extract features from the image data through BackBone (a network structure name) and Neck (a network structure name). Finally, output the predicted value at the output end. Use CIOU_Loss (a loss function name) as the loss function of the Bounding box to calculate the predicted value and the label value. Use the gradient back propagation algorithm to update and optimize the weight parameters of yolov5. When the validation set loss is minimized, save the yolov5 helmet recognition algorithm model. During the training process, the model will continuously learn how to identify helmets and adjust parameters according to the loss function.

[0075] Furthermore, in the process of constructing the face recognition model, the preprocessed face information of the five-axis CNC machine tool workshop personnel is used to train and test the initial face recognition model based on the Transform neural network. The Transform neural network consists of an encoder and a decoder. The encoder part of the Transformer consists of a total of 6 identical encoder layers. The encoder is composed of multiple layers including multi-head self-attention layers and feedforward neural networks, which are responsible for processing the input sequence. The decoder part of the Transformer also consists of a total of 6 identical decoder layers. The decoder is composed of multiple layers including masked self-attention layers, encoder-decoder attention layers and feedforward neural networks, which are responsible for generating output sequences. Both optimize information flow through residual connections and layer normalization. The multi-head attention mechanism is to perform multiple linear transformations on the input at the same time. Each transformation has its own Q, K, and V matrices, and calculates its own attention score respectively. Finally, the output of all transformations is combined to obtain the final output. In the specific training process, the image data of the training set containing the face images of the workshop employees is converted into sequence data by serializing pixel values, feature sequences extracted by feature extractors, and other methods. The sequence data is input into the encoder of the initial face recognition model to obtain the encoded representation. The encoder can be a neural network structure suitable for sequence data, such as a recurrent neural network (RNN) or a long short-term memory network (LSTM). The sequence data is linearly transformed using the encoder and decoder to generate an output sequence. The output of the decoder can be a reconstructed sequence data for comparison with the original sequence, or another representation, such as a face feature vector. The score is calculated based on the output sequence, and the loss function can also be used to evaluate the difference between the output sequence and the target sequence. The backpropagation algorithm is then used to adjust the parameters of the encoder and decoder to minimize the loss function. The adjusted encoder and decoder are verified using the validation set. By evaluating the performance of the model on the validation set, the generalization ability and accuracy of the model can be determined. After multiple iterations of adjustment and verification, the final face recognition model is obtained, which can recognize everyone in the workshop, and finally the face recognition model is saved.

[0076] It is understandable that the combination of the YOLOv5-based safety protection equipment recognition model and the transform-based face recognition model can be used on images captured by the workshop surveillance camera to identify people in the workshop who are not wearing safety protection equipment. For example, the above two models are combined and applied in a five-axis CNC machine tool workshop to detect whether anyone is wearing a safety helmet in the surveillance images in the workshop, as well as the facial information of the person. The information and facial features of people who are not wearing safety helmets are stored in the MySQL database, including information such as employee name and work number.

[0077] In this embodiment, by combining the face recognition model with the safety protection equipment recognition model, facial features can be automatically recognized and extracted from images or videos without human intervention, thereby improving recognition efficiency and saving labor costs.

[0078] Based on the first embodiment and / or the second embodiment of the present application, in the third embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 3 Before step S30, the workshop supervision decision method further includes the steps of:

[0079] Step S70: Obtain the device type of the target person who is not wearing the safety function device.

[0080] Step S80: Generate a risk impact assessment result based on the device type, the personnel information and the violation data.

[0081] Step S90: formulate corresponding control measures according to the risk impact assessment result and the determination and decision result, and feed back the control measures to the corresponding supervisor.

[0082] In this embodiment, the equipment type of the safety protection equipment not worn by the target personnel is obtained, and the corresponding risk assessment result is generated according to the equipment type, personnel information and violation duration. The risk assessment result and the judgment decision result will be used to formulate the corresponding control measures, and the control measures will be fed back to the corresponding supervisors to improve the safety supervision of the workshop. In this process, it is necessary to first determine the equipment type of the safety protection equipment that the target personnel are not wearing, which can be. Such as safety helmets, goggles, protective gloves, etc. In a feasible implementation, it can be determined directly according to the recognition result of the safety equipment recognition model, but considering that there are many types of safety protection equipment that the staff in the workshop need to wear, relying solely on visual recognition, it is impossible to guarantee the comprehensiveness and accuracy of the recognition. Therefore, RFID (Radio Frequency Identification) tags can also be implanted in the safety protection equipment in advance, and RFID readers and writers are deployed in various places in the workshop to synchronously detect whether the target personnel are wearing all the safety protection equipment and whether the recognition result is the same as the safety protection equipment model. Optionally, in another feasible implementation, smart sensors can also be embedded in safety protection equipment. When it is identified that the target person is not wearing safety protection equipment, it is determined whether the target person is wearing these devices by acquiring monitoring sensor data of each device, such as detecting whether a safety helmet is worn on the head.

[0083] Further, after the personnel information of the target personnel, the type of safety protection equipment not worn, and the violation data, including the violation time and the number of violations, are determined. According to the equipment type, personnel information, and violation data of the safety protection equipment not worn, a risk impact assessment is performed. Different risk levels or weights can be assigned to different types of safety protection equipment, depending on their importance and impact in safety work. For example, a helmet may be more important than a glove because a head injury may cause more serious harm to a person, so a helmet may be assigned a higher risk weight. In addition, considering that there may be differences in work experience, skill level, and safety awareness among different personnel, these factors can be incorporated into the assessment. The work experience of the target personnel is determined based on the personnel information. Personnel with more work experience or more comprehensive safety training are identified as having a lower risk, while novices or untrained personnel may be identified as having a higher risk. Different thresholds or standards can be set for the duration of the violation, such as short-term violations, medium-term violations, and long-term violations. By combining the above factors, the risk impact assessment results are calculated. According to the risk impact assessment results, corresponding control measures are formulated. The formulated control measures are fed back to the supervisor. This can be done through reporting, data visualization or system notifications to ensure that regulators are aware of violations and corresponding treatment measures in a timely manner. Optionally, the entire process can be evaluated and monitored regularly to identify problems and make timely improvements, including evaluating the accuracy of risk assessment models, the effectiveness of control measures, and the timeliness of regulatory feedback mechanisms.

[0084] In this embodiment, a risk assessment model or algorithm is established to evaluate and troubleshoot the situation of not wearing safety protection equipment, and corresponding control measures are taken according to the assessment results to ensure the safety and health of the workplace. At the same time, the assessment model is regularly updated and optimized to adapt to changes in the working environment and personnel, and to improve the safety supervision capabilities of the workshop.

[0085] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the workshop supervision decision-making method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0086] The present application provides a workshop supervision decision-making device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the workshop supervision decision-making method in the above-mentioned embodiment one.

[0087] Reference below Figure 4, which shows a schematic diagram of the structure of a workshop supervision decision device suitable for implementing the embodiment of the present application. The workshop supervision decision device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 4 The workshop supervision decision-making device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0088] like Figure 4 As shown, the workshop supervision decision-making device may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the workshop supervision decision-making device are also stored. The processing device 1001, ROM1002 and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the workshop supervision decision device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows a workshop supervision decision device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0089] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0090] The workshop supervision decision-making device provided by the present application adopts the workshop supervision decision-making method in the above-mentioned embodiment, which can solve the technical problem that the decision-making means in the previous way can only issue alarms or warnings, lack intelligent decision-making penalties, and therefore it is difficult to cover more complex safety supervision scenarios. Compared with the prior art, the beneficial effects of the workshop supervision decision-making device provided by the present application are the same as the beneficial effects of the workshop supervision decision-making method provided by the above-mentioned embodiment, and the other technical features in the workshop supervision decision-making device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0091] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0092] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0093] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the workshop supervision decision-making method in the above-mentioned embodiment.

[0094] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM: Random Access Memory), a read-only memory (ROM: Read Only Memory), an erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency: Radio Frequency), etc., or any suitable combination of the above.

[0095] The above-mentioned computer-readable storage medium may be included in the workshop supervision decision-making device; or it may exist independently without being assembled into the workshop supervision decision-making device.

[0096] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the workshop supervision decision-making device, the workshop supervision decision-making device:

[0097] When a target person who is not wearing safety protection equipment is identified based on the surveillance image, the personnel information of the target person is determined according to the personnel database and the facial information in the surveillance image;

[0098] Determine the violation duration of the target person not wearing the safety protection equipment according to the time when the target person is first recorded not wearing the safety protection equipment and the current time;

[0099] According to the personnel information and the duration of the violation, a determination and decision result is determined and output.

[0100] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may 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 may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0101] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0102] The modules involved in the embodiments described in the present application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0103] The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned workshop supervision decision-making method, and can solve the technical problems of workshop supervision decision-making. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the workshop supervision decision-making method provided in the above-mentioned embodiment, and will not be repeated here.

[0104] The present application also provides a computer program product, including a computer program, which implements the steps of the workshop supervision decision-making method as described above when executed by a processor.

[0105] The computer program product provided by this application can solve the technical problem of workshop supervision decision-making. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the workshop supervision decision-making method provided by the above embodiment, which will not be repeated here.

[0106] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A workshop supervision decision-making method, characterized in that: The method includes: When a target person who is not wearing safety protection equipment is identified based on the surveillance image, the personnel information of the target person is determined according to the personnel database and the facial information in the surveillance image; Determine the violation duration of the target person not wearing the safety protection equipment according to the time when the target person is first recorded not wearing the safety protection equipment and the current time; Matching corresponding workshop work rules and regulations according to the personnel information and the keywords of the violation duration; According to the workshop work rules and regulations, corresponding digital index data are determined, and the index data is associated with the corresponding determination and decision results and stored; Based on the personnel information, the duration of the violation and the digital indicator data, the violation behavior of the target personnel is classified and the determination decision result is determined, and the determination decision result includes a penalty result, wherein the operating rules and regulations used for making decisions on the current violation situation and the digital indicator data corresponding to the operating rules and regulations are determined according to the keywords, the personnel information and the duration of the violation are compared with the digital indicator data, and the determination decision result for the target personnel is generated and output; Obtain the equipment type of the target person not wearing safety protection equipment; generating a risk impact assessment result based on the device type, the personnel information and the violation data; Formulate corresponding control measures according to the risk impact assessment results and the determination and decision results; Feedback the control measures to the corresponding supervisory personnel.

2. The method according to claim 1, characterized in that Before the step of determining the personnel information of the target person according to the personnel database and the facial information in the surveillance image, the method further includes: Extracting a person image of a target person who is not wearing the safety protection equipment in the monitoring image based on the safety protection equipment recognition model; Extracting image features of the person image based on a face recognition model; According to the image features, corresponding face information is determined.

3. The method according to claim 1, characterized in that Before the step of determining the personnel information of the target person according to the personnel database and the facial information in the surveillance image, the method further includes: Based on the employee facial information and safety protection equipment data set of the workshop, a corresponding training set and a verification set are divided and noise reduction processing is performed on the training set and the verification set; Based on the processed training set and verification set of the safety protection equipment, the initial safety protection equipment recognition model is trained to obtain the safety protection equipment recognition model; Based on the processed training set and verification set of the employee's facial information, the initial face recognition model is trained to obtain a face recognition model.

4. The method according to claim 3, characterized in that The step of training the initial safety protection equipment identification model based on the processed training set and verification set of the safety protection equipment to obtain the safety protection equipment identification model includes: Inputting the training set into the input end of the initial safety protection equipment recognition model in batches, and performing image data enhancement processing; Extracting image features of the image data based on a neural network of the initial safety protection equipment recognition model; After the initial safety protection equipment identification model outputs a predicted value, the predicted value and the label value are calculated based on a preset loss function, and a weight parameter of the initial safety protection equipment identification model is adjusted according to the calculation result; The initial safety protection equipment identification model is verified based on the verification set, and when the test loss is minimum, the weight parameter is retained to obtain the safety protection equipment identification model.

5. The method according to claim 3, characterized in that The step of training the initial face recognition model based on the processed training set and verification set of the employee face information to obtain the face recognition model includes: Converting the image data of the training set into corresponding sequence data, and inputting the sequence data into the initial face recognition model; Based on the encoder and decoder of the initial face recognition model, the sequence data is linearly transformed and an output sequence is generated; Calculating an output score based on the output sequence, and adjusting parameters of the encoder and decoder according to the output score; The adjusted encoder and decoder are verified based on the verification set to obtain the face recognition model.

6. The method according to claim 1, characterized in that After the step of classifying the target person's violation behavior and determining the determination result based on the person information, the violation duration and the digital indicator data, the method further includes: Submitting the determination decision result to the confidence evaluation node; Based on the confidence evaluation node, a confidence evaluation is performed on the personnel information and violation duration of the decision-making result; When the evaluation score is higher than a preset threshold, the determination decision result is uploaded to the performance system.

7. A workshop supervision decision-making device, characterized in that: The workshop supervision decision device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the workshop supervision decision method according to any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the workshop supervision decision-making method according to any one of claims 1 to 6 are implemented.

Images