Online welding defect detection method and device based on molten pool image recognition

Abstract

The invention discloses an online welding defect detection method and device based on molten pool image recognition, and belongs to the technical field of industrial automation and machine vision. The method comprises the steps that a single-frame molten pool image in the welding process is obtained; outputting a defect identification result of each frame of molten pool image through a defect identification model; setting a judgment period, obtaining an identification result of each frame of molten pool image in the judgment period, and performing defect judgment; if the defect exists in the period, triggering an alarm mechanism of the welding equipment, including combining an alarm mark, a defect position starting point, a defect position section, a defect type and confidence into defect information; obtaining production context information corresponding to the judgment period, and binding the production context information with the defect information to generate a structured message; and uploading the structured message to a system end, and carrying out defect reminding and visual display. According to the invention, accurate identification of defects can be realized, data connection between defect identification information and a production management system is broken through, and comprehensive guarantee can be provided for welding operation.

Description

Technical Field

[0001] This invention relates to the fields of industrial automation and machine vision technology, and in particular to an online detection method and apparatus for welding defects based on molten pool image recognition. Background Technology

[0002] In the field of robotic welding automation, traditional methods for inspecting welding quality mainly rely on manual visual inspection or post-weld non-destructive testing (such as X-rays and ultrasound). These methods suffer from problems such as high latency, low efficiency, high subjectivity, and inability to provide real-time feedback to guide production. In recent years, deep learning has made breakthroughs in image recognition and semantic segmentation, but its direct application to the welding field still faces the following challenges: the welding process is subject to interference from strong light, spatter, and smoke, resulting in unstable image features; defects occur randomly, lacking an effective periodic statistical judgment mechanism; and there is a lack of linkage between camera data and the welding robot control system, making it difficult to provide timely feedback on inspection results. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art and provide an online detection method and device for welding defects based on molten pool image recognition. This method can acquire images in real time during the welding process, perform feature recognition using deep learning, and combine multi-cycle statistics and IoT status data for judgment feedback, thereby ensuring accurate online monitoring of the quality of robot welding.

[0004] To achieve the above objectives, the present invention is implemented using the following technical solution: In a first aspect, the present invention provides an online detection method for welding defects based on molten pool image recognition, comprising: In response to the start of the welding equipment, the molten pool video stream during the welding process is acquired, and the molten pool video stream is decoded into a single frame molten pool image; The molten pool image is input into a pre-constructed defect recognition model, and the defect recognition result of each frame of the molten pool image is output. Set a judgment period, obtain the recognition result of the molten pool image for each frame within the judgment period, and perform defect judgment; If the judgment cycle has a defect, the welding equipment alarm mechanism is triggered, and the welding equipment alarm mechanism includes: An alarm flag is generated for the judgment period. The welding position corresponding to the start of the judgment period is taken as the defect position start point, and a preset distance is added forward to form the defect position segment. The alarm flag, defect position start point, defect position segment, defect type and its confidence level are merged into defect information. Obtain the production context information corresponding to the judgment period, and bind it with the defect information to generate a structured message; The structured message is uploaded to the system for defect alerts and visual display.

[0005] Optionally, the defect determination includes: The percentage of frames of the melt pool image that are identified as having defects within the judgment period is calculated relative to the total number of frames in the judgment period. If the percentage is greater than the defect percentage threshold, it is determined that there is a defect within the judgment period; otherwise, it is determined that there is no defect within the judgment period.

[0006] Optionally, the defect determination includes: The judgment period is divided into multiple sub-periods. The proportion of the number of frames of the melt pool image with defects identified in each sub-period to the total number of frames in the corresponding sub-period is counted. If the proportion is greater than the defect proportion threshold, it is determined that there is a defect in the sub-period; otherwise, it is determined that there is no defect in the sub-period. The percentage of defective sub-cycles to the total number of sub-cycles is calculated. If the percentage is greater than the cycle percentage threshold, a serious defect is determined, triggering the welding equipment shutdown mechanism.

[0007] Optionally, the process of constructing the defect identification model includes: Collect images of the molten pool under various welding conditions and with various welding materials; The molten pool image is preprocessed, and defects and defect types are labeled after preprocessing to generate a training dataset; A defect identification model is constructed based on a lightweight convolutional neural network. The model parameters are initialized and trained using the training dataset. The model parameters are adjusted by an optimizer to minimize the loss function, resulting in the final defect identification model, which is then deployed on edge devices.

[0008] Optionally, the preprocessing of the molten pool image includes: The molten pool image is scaled to a calibrated size and then normalized and enhanced. The enhancement process includes rotation, flipping, cropping, and brightness adjustment.

[0009] Optionally, obtaining the production context information corresponding to the judgment period includes: Obtain the equipment code of the welding equipment and the weld number, product serial number, and start / stop status of the welding equipment corresponding to the judgment period from the IoT platform.

[0010] Secondly, the present invention provides an online welding defect detection device based on molten pool image recognition, comprising: The image acquisition module is configured to acquire a video stream of the molten pool during the welding process in response to the start of the welding equipment, and decode the video stream of the molten pool into a single frame image of the molten pool. The defect identification module is configured to input the molten pool image into a pre-constructed defect identification model and output the defect identification result for each frame of the molten pool image; The defect judgment module is configured to set a judgment period, obtain the recognition result of each frame of the melt pool image within the judgment period, and perform defect judgment. The defect handling module is configured to trigger a welding equipment alarm mechanism if a defect exists during the judgment period. The welding equipment alarm mechanism includes: An alarm flag is generated for the judgment period. The welding position corresponding to the start of the judgment period is taken as the defect position start point, and a preset distance is added forward to form the defect position segment. The alarm flag, defect position start point, defect position segment, defect type and its confidence level are merged into defect information. Obtain the production context information corresponding to the judgment period, and bind it with the defect information to generate a structured message; The structured message is uploaded to the system for defect alerts and visual display.

[0011] Thirdly, the present invention provides an electronic device, including a processor and a storage medium; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps according to the method described above.

[0012] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0013] Fifthly, the present invention provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the above-described method.

[0014] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: This invention provides an online detection method and device for welding defects based on molten pool image recognition. By conducting online monitoring and analysis of the weld pool, it can complete the judgment, location and data reporting in a very short time, shifting the quality control point from final inspection to process monitoring, improving the timeliness and initiative of quality control, and greatly reducing rework and scrap costs.

[0015] By deeply integrating defect identification with IoT platforms and system endpoints, defect information is automatically bound to real-time production context information. Each defect can be precisely traced back to "which equipment, which time, which product serial number, and which weld," breaking down information barriers and enabling full-process traceability and in-depth value mining of quality data. This provides unprecedented data support for root cause analysis, process optimization, predictive maintenance, and quality improvement. Managers can accurately pinpoint the source of problems and analyze the correlation between defects and process parameters, thereby achieving data-driven scientific decision-making.

[0016] Employing a deep learning-based visual model, its powerful feature self-learning capability can capture subtle defect patterns that are difficult for the human eye and traditional algorithms to detect. Combined with time-cycle-based intelligent defect judgment logic, rather than relying on single-frame image judgment, it effectively filters out false alarms caused by instantaneous interference. Thus, it can maintain extremely high detection accuracy, recall, and stability even in complex and harsh industrial environments, reducing reliance on the experience and subjective judgment of inspection personnel. Attached Figure Description

[0017] Figure 1 This is a schematic flowchart of the online detection method for welding defects based on molten pool image recognition provided in an embodiment of the present invention; Figure 2 This is a schematic diagram showing the relative positions of the camera and the welding torch provided in an embodiment of the present invention; Figure 3 This is an architecture diagram of the online welding defect monitoring system provided in an embodiment of the present invention. Detailed Implementation

[0018] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0019] Example 1

[0020] like Figure 1 As shown, this embodiment of the invention provides an online detection method for welding defects based on molten pool image recognition, which can be widely applied to quality monitoring of welding production processes in industries such as engineering machinery and shipbuilding. The method includes the following steps: Step S1: In response to the start of the welding equipment, acquire the video stream of the molten pool during the welding process and decode the video stream of the molten pool into a single frame image of the molten pool.

[0021] Specifically, in this implementation, video data is acquired using a high-resolution industrial camera or a professional weld pool camera. The camera is deployed at a specific location on the welding torch to ensure it can capture a complete image of the weld pool, such as... Figure 2As shown, the camera can be mounted on the welding torch. In other alternative implementations, the camera position can be deployed according to the actual scenario requirements.

[0022] Step S2: Input the molten pool image into the pre-built defect recognition model and output the defect recognition result of each frame of the molten pool image.

[0023] Specifically, in the implementation method, the process of constructing the defect identification model includes: Collect images of the molten pool under various welding conditions and with various welding materials; The molten pool image is preprocessed, and defects and defect types are labeled after preprocessing to generate a training dataset; A defect identification model is built based on a lightweight convolutional neural network (such as EfficientNetV2). The model parameters are initialized and trained using a training dataset. The model parameters are adjusted by an optimizer to minimize the loss function, resulting in the final defect identification model. This model is then deployed to edge devices (e.g., converted to a PyTorch .pth file for edge deployment).

[0024] The main purpose of preprocessing the molten pool image is to expand the dataset and improve the robustness of the model. Preprocessing includes, but is not limited to, scaling the molten pool image to a calibrated size, normalizing it, and enhancing it. Enhancement processes include rotation, flipping, cropping, and brightness adjustment.

[0025] Step S3: Set the judgment period, obtain the recognition results of each frame of the molten pool image within the judgment period, and perform defect judgment.

[0026] Specifically, in this embodiment, defect judgment includes: The percentage of frames in the molten pool image that are identified as having defects within the judgment period is calculated relative to the total number of frames in the judgment period. If the percentage is greater than the defect percentage threshold, it is determined that there is a defect within the judgment period; otherwise, it is determined that there is no defect within the judgment period.

[0027] The judgment period can be flexibly set according to the welding speed. The final judgment is made by analyzing the recognition results of multiple consecutive frames within a judgment period, rather than relying on the result of a single frame, thereby effectively filtering out instantaneous interference and reducing the false alarm rate.

[0028] Furthermore, this embodiment also adds a defect judgment method, including: The judgment period is divided into multiple sub-periods. The proportion of the number of frames of the melt pool image with defects identified in each sub-period to the total number of frames in the corresponding sub-period is counted. If the proportion is greater than the defect proportion threshold, it is determined that there is a defect in the sub-period; otherwise, it is determined that there is no defect in the sub-period. The percentage of defective sub-cycles to the total number of sub-cycles is counted. If the percentage is greater than the cycle percentage threshold, a serious defect is determined, triggering the welding equipment shutdown mechanism.

[0029] By dividing the judgment period into sub-periods, we can avoid false alarms caused by relying on single-frame results. On the other hand, we can determine the severity of defects based on the proportion of defective sub-periods and stop the machine in time to prevent losses in severe cases.

[0030] Step S4: If a defect is found in the judgment cycle, the welding equipment alarm mechanism is triggered. The welding equipment alarm mechanism includes: An alarm flag is generated for each judgment cycle. The welding position corresponding to the start of the judgment cycle is taken as the starting point of the defect position. A preset distance is added forward to form the defect position segment (e.g., synchronously acquiring welding speed signals, and through integration calculation, accurately converting the image time point of the identified defect into the physical position on the weld (e.g., "starting from 120mm and continuing for 30mm"), and generating a standardized position identifier). The alarm flag, defect position starting point, defect position segment, defect type and its confidence level are combined into defect information. Obtain the production context information corresponding to the judgment cycle and bind it with the defect information to generate a structured message (such as a JSON message). The production context information includes: obtaining the equipment code of the welding equipment and the weld number, product serial number, and start / stop status of the welding equipment corresponding to the judgment cycle from the IoT platform.

[0031] Upload structured messages (via a high-performance message queue such as Apache Kafka) to the system (such as a quality management system, MES system, or Kanban system) for defect alerts and visual display.

[0032] Furthermore, it can record all logs of system operation, inference process and data reporting, and can store the images and video streams triggered when an alert is triggered locally for subsequent review and model optimization.

[0033] Based on the online detection methods provided above, the following deployment can be made in actual implementation: The EfficientNetV2 model was trained on the server using historical welding image data, and the trained model file was deployed to the industrial control computer on site. Subsequently, system parameters (such as camera IP, IoT device encoding, Kafka topic, etc.) were configured.

[0034] After the system starts up, the field inference program begins to run and continuously acquires the device status through the IoT interface. When the welding device status signal returned by the IoT (such as state_1) is "1" (started), the system immediately triggers the camera connection and inference threads.

[0035] After the images captured by the camera are inferred by the model, the results are sent to the judgment logic module for periodic statistics. Once a defect is determined, the system immediately calls the send_warn method, combines the current weld number, defect mark, location information, and other data, and reports it to the cloud platform via the Kafka sender, thus completing a complete online detection and early warning process.

[0036] Once the system is deployed, its overall workflow is an automated, closed-loop process, such as... Figure 3 As shown, the specific steps are as follows: Step 1: System initialization and standby monitoring.

[0037] 1. The system starts up and the edge computing device loads the deep learning model and configuration parameters (such as camera IP, IoT interface address, and Kafka topic).

[0038] 2. The image processing and inference thread and the result display thread have been initialized and are in standby mode.

[0039] 3. The system begins to cyclically query the IoT platform to obtain the status signal of the specified welding equipment (such as state_1).

[0040] Step 2: Production Start-up and System Triggering.

[0041] 4. When the welding robot starts working, the device status signal on the IoT platform changes to "1" (start status).

[0042] 5. Upon receiving the "start" signal via the API interface, the system immediately sends a command to the control module.

[0043] 6. The control module then triggers the industrial camera to start acquiring data and initiates a series of threads, including image reception, inference and judgment, and data reporting.

[0044] Step 3: Real-time image acquisition and preprocessing.

[0045] 7. Industrial cameras capture high-definition video streams of the welding molten pool area in real time.

[0046] 8. The image receiving module captures single-frame images from the video stream at a preset frame rate (e.g., 30 FPS).

[0047] 9. Perform standardization preprocessing on the captured images (such as size scaling, color space conversion, and normalization) to meet the input requirements of deep learning models.

[0048] Step 4: Online inference of the deep learning model.

[0049] 10. The preprocessed image is fed into a trained lightweight deep learning model (such as EfficientNetV2) deployed on an edge device for forward inference.

[0050] 11. The model outputs the classification result of the current frame image (such as "GOOD" or "BAD") and the corresponding confidence score.

[0051] Step 5: Intelligent cycle decision-making and positioning.

[0052] 12. The intelligent judgment logic module initiates statistical decision-making based on a time window (e.g., 4.5 seconds): * Record the inference results for all frames within a period.

[0053] * At the end of the cycle, calculate the proportion of "defective" frames to the total number of frames.

[0054] * If the proportion exceeds the set threshold (e.g., 60%), the weld section is determined to have a defect, and a warning flag (warn_flag=1) is generated.

[0055] 13. At the same time, the system automatically accumulates and records the starting position and length of the defect (e.g., s:150; l:30) based on the welding speed integral, thereby achieving online precise positioning of the defect.

[0056] Step 6: Data fusion and real-time reporting.

[0057] 14. The result reporting module obtains the latest production context information from the IoT platform, including equipment number, weld number, product serial number, batch number, etc.

[0058] 15. Automatically bind the defect judgment results (early warning signs, defect type, confidence level, location information) with production information and assemble them into a structured message (such as JSON format).

[0059] 16. Push messages to the Quality Management System (QMS), Manufacturing Execution System (MES), and Production Kanban System in real time and asynchronously through a high-performance message queue (Kafka).

[0060] Step 7: Visualize and persist the results.

[0061] 17. After receiving the results, the management system updates the visual dashboard in real time, triggers audible and visual alarms, and reminds production personnel.

[0062] 18. The system will locally save the images, video clips and all related data when an alert is triggered, forming a quality traceability archive.

[0063] 19. Real-time video footage with overlaid information is displayed on the industrial control computer screen for on-site monitoring personnel to view.

[0064] Step 8: Production halt and system reset.

[0065] 20. When the welding operation is completed, the IoT device status signal changes to "0" (stopped state).

[0066] 21. After the system detects the "stop" signal, it automatically stops the camera acquisition, inference, and reporting threads and releases resources.

[0067] 22. The defect location counter is reset to zero, awaiting the start of the next work cycle.

[0068] Step 9: Continuous model optimization (performed offline and asynchronously).

[0069] 23. The system periodically transmits newly collected defect sample data (after manual verification) back to the data center.

[0070] 24. Use new data to incrementally train and optimize the deep learning model, and update the model parameters.

[0071] 25. Redeploy the optimized model to the edge to complete the iterative upgrade of the model and realize the self-evolution of the system's detection capabilities.

[0072] Example 2

[0073] This invention provides an online welding defect detection device based on molten pool image recognition, comprising: The image acquisition module is configured to acquire the molten pool video stream during the welding process in response to the start of the welding equipment, and decode the molten pool video stream into a single frame molten pool image; The defect identification module is configured to input the molten pool image into a pre-built defect identification model and output the defect identification result for each frame of the molten pool image; The defect judgment module is configured to set a judgment period, obtain the recognition result of each frame of the melt pool image within the judgment period, and perform defect judgment. The defect handling module is configured to trigger a welding equipment alarm mechanism if a defect is detected during the judgment period. The welding equipment alarm mechanism includes: An alarm flag is generated for the judgment cycle. The welding position corresponding to the start of the judgment cycle is taken as the defect location start point, and a preset distance is added forward to form the defect location segment. The alarm flag, defect location start point, defect location segment, defect type and its confidence level are merged into defect information. Obtain the production context information corresponding to the judgment cycle and bind it with the defect information to generate a structured message; The structured messages are uploaded to the system for defect alerts and visual display.

[0074] Example 3

[0075] Based on the online detection method provided in Embodiment 1, this embodiment of the invention provides an electronic device, including a processor and a storage medium; Storage media are used to store instructions; The processor is used to perform operations according to instructions to execute the steps according to the method described above.

[0076] Example 4

[0077] Based on the online detection method provided in Embodiment 1, this embodiment of the invention provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above method.

[0078] Example 5

[0079] Based on the online detection method provided in Embodiment 1, this embodiment of the invention provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the above-described method.

[0080] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied 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.

[0081] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0082] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0083] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0084] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. An online detection method for welding defects based on molten pool image recognition, characterized in that, include: In response to the start of the welding equipment, the molten pool video stream during the welding process is acquired, and the molten pool video stream is decoded into a single frame molten pool image; The molten pool image is input into a pre-constructed defect recognition model, and the defect recognition result of each frame of the molten pool image is output. Set a judgment period, obtain the recognition result of the molten pool image for each frame within the judgment period, and perform defect judgment; If the judgment cycle has a defect, the welding equipment alarm mechanism is triggered, and the welding equipment alarm mechanism includes: An alarm flag is generated for the judgment period. The welding position corresponding to the start of the judgment period is taken as the defect position start point, and a preset distance is added forward to form the defect position segment. The alarm flag, defect position start point, defect position segment, defect type and its confidence level are merged into defect information. Obtain the production context information corresponding to the judgment period, and bind it with the defect information to generate a structured message; The structured message is uploaded to the system for defect alerts and visual display.

2. The online detection method for welding defects based on molten pool image recognition according to claim 1, characterized in that, The defect assessment includes: The percentage of frames of the melt pool image that are identified as having defects within the judgment period is calculated relative to the total number of frames in the judgment period. If the percentage is greater than the defect percentage threshold, it is determined that there is a defect within the judgment period; otherwise, it is determined that there is no defect within the judgment period.

3. The online detection method for welding defects based on molten pool image recognition according to claim 1, characterized in that, The defect assessment includes: The judgment period is divided into multiple sub-periods. The proportion of the number of frames of the melt pool image with defects identified in each sub-period to the total number of frames in the corresponding sub-period is counted. If the proportion is greater than the defect proportion threshold, it is determined that there is a defect in the sub-period; otherwise, it is determined that there is no defect in the sub-period. The percentage of defective sub-cycles to the total number of sub-cycles is calculated. If the percentage is greater than the cycle percentage threshold, a serious defect is determined, triggering the welding equipment shutdown mechanism.

4. The online detection method for welding defects based on molten pool image recognition according to claim 1, characterized in that, The process of constructing the defect identification model includes: Collect images of the molten pool under various welding conditions and with various welding materials; The molten pool image is preprocessed, and defects and defect types are labeled after preprocessing to generate a training dataset; A defect identification model is constructed based on a lightweight convolutional neural network. The model parameters are initialized and trained using the training dataset. The model parameters are adjusted by an optimizer to minimize the loss function, resulting in the final defect identification model, which is then deployed on edge devices.

5. The online detection method for welding defects based on molten pool image recognition according to claim 4, characterized in that, The preprocessing of the molten pool image includes: The molten pool image is scaled to a calibrated size and then normalized and enhanced. The enhancement process includes rotation, flipping, cropping, and brightness adjustment.

6. The online detection method for welding defects based on molten pool image recognition according to claim 1, characterized in that, The step of obtaining the production context information corresponding to the judgment period includes: Obtain the equipment code of the welding equipment and the weld number, product serial number, and start / stop status of the welding equipment corresponding to the judgment period from the IoT platform.

7. An online detection device for welding defects based on molten pool image recognition, characterized in that, include: The image acquisition module is configured to acquire a video stream of the molten pool during the welding process in response to the start of the welding equipment, and decode the video stream of the molten pool into a single frame image of the molten pool. The defect identification module is configured to input the molten pool image into a pre-constructed defect identification model and output the defect identification result for each frame of the molten pool image; The defect judgment module is configured to set a judgment period, obtain the recognition result of each frame of the melt pool image within the judgment period, and perform defect judgment. The defect handling module is configured to trigger a welding equipment alarm mechanism if a defect exists during the judgment period. The welding equipment alarm mechanism includes: An alarm flag is generated for the judgment period. The welding position corresponding to the start of the judgment period is taken as the defect position start point, and a preset distance is added forward to form the defect position segment. The alarm flag, defect position start point, defect position segment, defect type and its confidence level are merged into defect information. Obtain the production context information corresponding to the judgment period, and bind it with the defect information to generate a structured message; The structured message is uploaded to the system for defect alerts and visual display.

8. An electronic device, characterized in that, Including processor and storage media; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-6.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method described in any one of claims 1-6.

Images