Method, device and storage medium for identifying violations in power projects by artificial intelligence

Abstract

The present application discloses a method, apparatus and storage medium for artificial intelligence to identify violations in power projects. The method includes: obtaining the construction area and construction information of a construction project, and marking the construction behavior of the construction area according to the construction information, wherein the construction information at least includes a construction plan and a layout of construction equipment; dividing the marked construction area based on violation construction behavior labels to obtain a number of sub-areas, wherein each sub-area contains at least one violation construction behavior label; obtaining in real time the on-site construction video collected by an image acquisition device arranged on the sub-area, and obtaining the on-site construction image corresponding to the sub-area in the on-site construction video; obtaining, based on the on-site construction image, a target image for violation determination corresponding to the sub-area; and inputting the target image into a pre-trained violation detection model to obtain a detection result corresponding to the sub-area. The detection accuracy and timeliness of violation behaviors are improved.

Description

Technical Field

[0001] This application relates to the field of power engineering construction assistance, and particularly to a method, device, and storage medium for artificial intelligence to identify violations in power projects. Background Technique

[0002] In power engineering construction projects, information management is an important support system to ensure the high efficiency, safety, and controllable quality of power projects. By building an information platform integrating data collection, storage, processing, analysis, and visual display, real-time monitoring and dynamic adjustment of key elements such as construction progress, quality, safety, and cost are realized. Specifically, it not only covers front-end management links such as project preliminary planning, design approval, material procurement, and construction preparation, but also through in-depth construction sites, realizes precise scheduling and intelligent control of personnel, machinery, and materials at the construction site, better ensuring that construction activities meet regulatory requirements and effectively preventing accidents at the project construction site.

[0003] Currently, the vast majority of construction safety accidents are due to the dangerous behaviors of construction workers. The traditional manual management method can meet the regulatory requirements of construction through manual management and monitoring, which not only requires a large amount of human resources but also has a certain lag, resulting in poor monitoring effects.

[0004] Therefore, there is an urgent need for a method to identify violations in power projects that can improve the accuracy and timeliness of monitoring violations during the construction process. Summary of the Invention

[0005] The purpose of the embodiments of this application is to provide a method, device, electronic device, and storage medium for artificial intelligence to identify violations in power projects, so as to solve the technical problem of low accuracy and timeliness in monitoring violations during the construction process in related technologies.

[0006] In the first aspect, the embodiments of this application provide a method for artificial intelligence to identify violations in power projects, including:

[0007] Obtain the construction area and construction information of the construction project, and mark the construction behaviors in the construction area according to the construction information, where the construction information at least includes the construction plan and the layout of construction equipment;

[0008] Divide the marked construction area based on the violation construction behavior labels to obtain several sub-areas, where each sub-area contains at least one violation construction behavior label;

[0009] Real-time obtain the on-site construction video collected by the image acquisition device set on the sub-area, and obtain the on-site construction image corresponding to the sub-area in the on-site construction video;

[0010] Based on the on-site construction image, obtain the target image for violation determination corresponding to the sub-region;

[0011] Input the target image into a pre-trained violation detection model to obtain the detection result corresponding to the sub-region, where the detection result includes whether there is a violation and the violation behavior.

[0012] In a second aspect, an embodiment of the present application provides an artificial intelligence device for identifying violations in a power project, including:

[0013] An information acquisition module, configured to acquire the construction area and construction information of a construction project, and mark the construction behavior of the construction area according to the construction information, where the construction information at least includes a construction plan and a construction equipment layout;

[0014] A region division module, configured to divide the marked construction area based on violation construction behavior labels to obtain a number of sub-regions, where each sub-region contains at least one violation construction behavior label;

[0015] An image acquisition module, configured to acquire in real time the on-site construction video collected by an image acquisition device arranged on the sub-region, and obtain the on-site construction image corresponding to the sub-region from the on-site construction video;

[0016] An image processing module, configured to obtain the target image for violation determination corresponding to the sub-region based on the on-site construction image;

[0017] A violation determination module, configured to input the target image into a pre-trained violation detection model to obtain the detection result corresponding to the sub-region, where the detection result includes whether there is a violation and the violation behavior.

[0018] In a third aspect, an embodiment of the present application provides an electronic device, which includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps in the method for identifying violations in a power project by artificial intelligence described in any one of the above are implemented.

[0019] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps in the method for identifying violations in a power project by artificial intelligence described in any one of the above are implemented.

[0020] The embodiments of the present application provide a method, device, electronic device and storage medium for artificial intelligence to identify violations in power projects. By obtaining the construction area and construction information of a construction project, marking the construction behaviors in the construction area according to the construction information, and dividing the marked construction area based on the violation construction behavior labels to obtain several sub-areas. Then, during violation detection, the on-site construction video collected by the image acquisition device set on the sub-area is obtained in real time, and the target image for violation determination is obtained through the processing of the on-site construction video. Finally, the pre-trained violation detection model is used to analyze and process the target image to obtain the detection result corresponding to the sub-area. This avoids the errors and untimely nature brought by manual monitoring, and at the same time improves the accuracy and timeliness of detecting violation construction behaviors in each area through the method of area division. Description of the Drawings

[0021] Figure 1 is a schematic flowchart of the steps of the method for artificial intelligence to identify violations in power projects provided by the embodiments of the present application;

[0022] Figure 2 is a schematic flowchart of the steps for dividing the construction area provided by the embodiments of the present application;

[0023] Figure 3 is a schematic diagram after the construction area is divided provided by the embodiments of the present application;

[0024] Figure 4 is a schematic flowchart of the steps for obtaining video frames provided by the embodiments of the present application;

[0025] Figure 5 is a schematic diagram of the alignment of video segments provided by the embodiments of the present application;

[0026] Figure 6 is a schematic flowchart of the process for training the violation detection model provided by the embodiments of the present application;

[0027] Figure 7 is a schematic structural diagram of the attention mechanism CBAM provided by the embodiments of the present application;

[0028] Figure 8 is a schematic structural diagram of the improved YOLOv5 provided by the embodiments of the present application;

[0029] Figure 9 is a schematic structural diagram of the device for artificial intelligence to identify violations in power projects provided by the embodiments of the present application

[0030] Figure 10 is a schematic structural diagram of the electronic device provided by the embodiments of the present application;

[0031] Figure 11 It is another schematic structural diagram of the electronic device provided by the embodiment of the present application. Detailed implementation manners

[0032] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0033] It should be understood that the various steps recorded in the method implementation manners disclosed in the present application may be executed in different orders and / or executed in parallel. In addition, the method implementation manners may include additional steps and / or omit the steps shown. The scope disclosed in the present application is not limited in this regard.

[0034] The term "including" and its variations used herein are open-ended, that is, "including but not limited to". The term "based on" is "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0035] In the related art, most construction safety accidents are caused by the dangerous behaviors of construction workers. The traditional manual management method can meet the construction specification requirements through manual management and monitoring, which not only requires a large amount of human resources, but also has a certain lag, thus resulting in poor monitoring effects.

[0036] To solve the technical problems existing in the related art, the embodiment of the present application provides a method for artificial intelligence to identify violations in power projects. Please refer to Figure 1 , Figure 1 It is a schematic flowchart of the steps of the method for artificial intelligence to identify violations in power projects provided by the embodiment of the present application. The method includes steps 101 to 105.

[0037] Step 101: Obtain the construction area and construction information of the construction project, and mark the construction behaviors in the construction area according to the construction information, where the construction information at least includes the construction plan and the layout of construction equipment.

[0038] In one embodiment, when detecting violations in a power construction project, the construction behavior during actual construction is obtained, and then it is determined whether there is a construction violation by analysis and processing. Specifically, the construction area and construction information of the construction project are obtained, and then the construction information is used to mark the construction behavior in the construction area. Among them, the obtained construction information at least includes the construction plan and the layout of construction equipment. At the same time, it may also include personnel allocation, etc.

[0039] Exemplarily, when using the construction information to mark the construction behavior in the construction area, the construction behavior is the behavior of the construction object when implementing the construction plan, such as high-altitude operation, walking, and equipment operation, etc. For the construction behavior, the location where it occurs can be determined, that is, the areas where different construction behaviors occur are different. For example, the construction behavior of walking can occur throughout the construction area, while the construction behavior of high-altitude operation requires the construction object to be in a specific position to occur. Moreover, the corresponding violation behaviors for different construction behaviors are also different. Therefore, using the construction plan in the construction information to mark the construction area can facilitate more accurate identification of different violation behaviors.

[0040] In practical applications, when marking the construction area, since the actual operation scenario is a three-dimensional space scenario, when using the construction information to mark the construction area, it includes: constructing a three-dimensional space corresponding to the construction area and filling the three-dimensional space based on the layout of the construction equipment shown; performing marking processing of the construction behavior in the filled three-dimensional space according to the construction plan, and determining the spatial area in the filled three-dimensional space when implementing the construction plan.

[0041] Therefore, when performing the marking processing, first construct the three-dimensional space corresponding to the construction area, and then fill the equipment, objects, obstacles, etc. included in the construction area in the constructed three-dimensional space. Specifically, obtain the layout of the construction equipment from the construction information, and then fill the layout of the construction equipment into the constructed three-dimensional space, that is, set the position of the construction equipment in the three-dimensional space, etc. Then, determine and mark the construction behavior according to the construction plan in the construction information to determine the spatial area in the three-dimensional space when implementing the construction plan, that is, complete the marking processing of the construction area.

[0042] Step 102: Divide the marked construction area based on the violation construction behavior labels to obtain several sub-areas, where each sub-area contains at least one violation construction behavior label.

[0043] In one embodiment, after completing the marking process of the construction area based on construction behaviors, the construction area will continue to be divided based on the illegal construction behavior labels to obtain multiple sub-areas. When marking and dividing the construction area based on the illegal construction behavior labels, the same small construction area can be marked with multiple illegal construction behavior labels. Therefore, for the obtained sub-areas, there will be at least one illegal construction behavior label.

[0044] Exemplarily, the illegal construction behavior labels used for area division can be preset based on construction projects, construction sites, and construction requirements. For example, the illegal construction behavior labels can include: not wearing a safety helmet within the construction area, not fastening the safety belt during high-altitude operations, performing edge operations at heights without protective facilities, not using the safety passage, riding on the lifting device for suspended objects or walking / standing under suspended objects. Of course, the illegal construction behavior labels can also include other labels according to the actual situation.

[0045] During the actual operation process, for different construction plans, different construction behaviors will occur, and the construction behaviors will be accompanied by situations of compliance and non-compliance. Therefore, when dividing the construction area, first obtain the possible illegal construction behaviors currently existing, and then perform area division based on the corresponding labels. Refer to Figure 2 , Figure 2 FIG. is a flowchart of the steps for dividing the construction area provided by an embodiment of the present application, where the steps include step 201 to step 203.

[0046] Step 201, obtain the illegal construction behavior labels corresponding to the execution of the construction plan;

[0047] Step 203, perform marking processing on the three-dimensional space based on the illegal construction behavior labels to obtain the marked three-dimensional space;

[0048] Step 203, obtain the area demarcation line of the three-dimensional space based on the illegal construction behavior labels marked on the space area, and divide the three-dimensional space according to the area demarcation line to obtain several sub-areas.

[0049] Specifically, when dividing the construction area, first obtain the illegal construction behavior labels corresponding to the execution of the construction plan, then perform marking processing on the three-dimensional space based on the illegal construction behavior labels to obtain the marked three-dimensional space. Next, determine the area demarcation line in the three-dimensional space based on the illegal construction behavior labels marked on the space area, and then perform area division processing on the three-dimensional space according to the area demarcation line to obtain several sub-areas.

[0050] Among them, the illegal construction behavior labels corresponding to the execution of the construction plan are preset. Since different construction behaviors correspond to corresponding illegal construction behavior labels, the corresponding relationship between the construction plan and the illegal construction behavior labels can be constructed in advance. Then, when marking is required, the corresponding illegal construction behavior labels can be directly found based on the construction plan of the construction project.

[0051] When obtaining the illegal construction behavior labels, they can be obtained through the statistical analysis of historical data. Specifically, data is obtained in historical projects, including manual collection and automated collection. Among them, manual collection is to manually extract key information by reading documents such as fault reports and fault records and enter it into the specified data system. Automated collection is for fault data that has been digitized and has a unified format. By configuring the parameters of the collection task, such as collection time, collection frequency, etc., start the collection task, and the system automatically extracts data from the data source and stores it in the specified location. And the collection process is monitored in real time to ensure the accuracy and integrity of data collection. An alarm mechanism can be set up to send an alarm and take corresponding handling measures in a timely manner when the collection task has an exception or error.

[0052] For the collected data, quality verification will also be carried out, including checks on aspects such as data integrity, accuracy, and consistency. The verified data is stored in the specified database or data warehouse. According to the type and characteristics of the data, a suitable storage method and management strategy are selected to ensure the security and accessibility of the data, which is convenient for subsequent data analysis and application.

[0053] After obtaining the relevant data, the illegal construction behavior labels can be obtained through statistical analysis. In the actual operation process, the risk factors at the construction site mainly come from the unsafe behaviors of construction personnel, the abnormal states of construction machinery, and adverse environmental conditions. Based on the historical construction fault data, statistical analysis of key construction dangerous behaviors is carried out to determine the main unsafe behaviors and corresponding risk levels of illegal construction in power projects. Among them, the risk level can be used to indicate the reminder method during the monitoring process. For example, the higher the risk level, the more obvious the reminder, such as a broadcast.

[0054] For example, assume that in a certain period across the country, there are a total of 100 production safety accidents in power projects. Among them, 40 cases of not wearing safety helmets in the construction area account for 40%; 20 cases of not wearing safety belts during high-altitude operations account for 20%; 15 cases of performing edge operations at heights without protective facilities account for 15%; 15 cases of not using the safety passage account for 15%; 10 cases of taking the hoisting device or walking or standing under the hoisted object account for 10%. Combining the statistical data and accident reports for analysis, the illegal construction behavior labels and corresponding risk levels caused by the dangerous behaviors of construction personnel can be as shown in Table 1 below:

[0055] Table 1

[0056]

[0057] When marking and processing the three-dimensional space according to the illegal construction behavior labels, by calibrating the spatial regions of the illegal construction behavior labels in the three-dimensional space, the dividing lines between different spatial regions will be used as the dividing lines for area division, and then area division processing will be carried out according to the dividing lines.

[0058] Taking the construction area as the occupied area, that is, the two-dimensional plane, after marking and area division of the illegal construction behavior, the obtained area division result can be as Figure 3 shown, referring to Figure 3 , Figure 3 which is a schematic diagram after the construction area division provided by the embodiment of the present application. Among them, M is the entire construction area, A is the area corresponding to the behavior of not wearing a safety helmet within the construction area, B is the area corresponding to the behavior of not fastening the safety belt during high-altitude operations, C is the area corresponding to the behavior of performing edge operations at high altitudes without protective facilities, D is the area corresponding to the behavior of not using the safety passage, E is the area corresponding to the behavior of riding on a hoisting device or walking and standing under a hoisted object, and A coincides with M, and E and C are within B.

[0059] Step 103, obtain in real time the on-site construction video collected by the image acquisition device arranged on the sub-region, and obtain the on-site construction image corresponding to the sub-region in the on-site construction video.

[0060] In one embodiment, after completing the division processing of the construction area, when monitoring and analyzing illegal behaviors, accurate and precise analysis can be carried out for different areas. At the same time, for the divided areas, comprehensive monitoring can be carried out within the corresponding sub-regions, and then by obtaining the corresponding images or videos for analysis, it is determined whether there is an illegal construction behavior in the corresponding sub-region.

[0061] Therefore, when performing illegal behavior analysis processing, obtain in real time the on-site construction video collected by the image acquisition device arranged on the sub-region, and obtain the on-site construction image corresponding to the sub-region in the on-site construction video. Specifically, for each sub-region, a certain number of image acquisition devices can be set and installed according to requirements to be able to collect videos and images for accurate analysis.

[0062] Exemplarily, during the monitoring process, the on-site construction video collected by the set image acquisition device is obtained in real time, and then the corresponding on-site construction image is obtained by acquiring images in the on-site construction video. When obtaining the on-site construction image, it is obtained by extracting video frames, specifically including: obtaining in real time the on-site construction videos collected by each image acquisition device set on the sub-region, and performing video frame extraction on the on-site construction videos to obtain a number of video frames, where the number of video frames corresponding to each image acquisition device is the same; performing similarity screening on the video frames corresponding to each image acquisition device to obtain one video frame corresponding to each image acquisition device, and summarizing the one video frame obtained by each image acquisition device to obtain the on-site construction image corresponding to the sub-region.

[0063] Among them, during the processing, for the obtained on-site construction video, video frame extraction can be performed according to the set frame rate, and the number of video frames extracted from the on-site construction video corresponding to each image acquisition device is the same. Then, for the video frames corresponding to each image acquisition device, one video frame that can represent the image collected by the image acquisition device will be obtained through similarity screening, and then the one video frame corresponding to all image acquisition devices in the sub-region will be summarized to obtain the on-site construction image.

[0064] Furthermore, when detecting and analyzing illegal construction behaviors, it is usually for the construction object, that is, the constructor. Therefore, when extracting video frames, it can be referred to Figure 4 , Figure 4 FIG. 10 is a flowchart of a step for obtaining video frames provided by an embodiment of the present application, where this step includes steps 401 to 404.

[0065] Step 401, obtain in real time the on-site construction videos collected by each image acquisition device set on the sub-region, and detect the construction object in the on-site construction videos;

[0066] Step 402, intercept the video segment of the on-site construction video according to the detection result to obtain a video segment containing the construction object, where the video segment contains a time stamp;

[0067] Step 403, perform video alignment processing on the video segment corresponding to each image acquisition device according to the time stamp to obtain the target video segment corresponding to each image acquisition device for video frame extraction;

[0068] Step 404, perform video frame extraction on the target video segment to obtain a number of video frames.

[0069] Specifically, after obtaining the on-site construction video collected by the image acquisition device, the on-site construction video will be detected for construction objects, and the video segment of the on-site construction video will be intercepted according to the detection result to obtain the video segment containing the construction object, and the obtained video segments all contain corresponding timestamps. Then, the video segments corresponding to each image acquisition device will be processed for video alignment using the timestamps to obtain the target video segments corresponding to each image acquisition device for video frame extraction. Finally, video frame extraction processing will be performed in the target video segments at the set frame rate to obtain a number of video frames.

[0070] Exemplarily, when performing construction object detection, it is to determine whether each sub-region contains a construction object. Only when there is a construction object, the above-described construction violation behavior may occur. Since the monitoring of construction violation behavior is a real-time process, in each judgment cycle, after obtaining the on-site construction video, it is necessary to determine whether the on-site construction video contains a construction object. If it does not contain a construction object, there will be no illegal construction behavior at present. By detecting the construction object, the length of the on-site construction video collected by each image acquisition device is effectively shortened. And because there may be a certain difference in the time when each image acquisition device captures the construction object, therefore, time alignment processing can be performed on the intercepted video frames, which can effectively improve the accuracy of analysis.

[0071] When aligning multiple video frames based on timestamps, assume that the sub-region contains four image acquisition devices, and at this time the intercepted video segments are as Figure 5 shown, including video segment 1, video segment 2, video segment 3 and video segment 4, and they are arranged based on their respective corresponding timestamps. When determining the target video segment, it is intercepted through the timestamps corresponding to each video segment. As Figure 5 shown, the target video segment obtained at this time is the video segment in the time period of t1 - t2, that is, the video segments 1, 2, 3 and 4 are intercepted respectively based on the t1 - t2 moments to obtain the target video segments corresponding to each image acquisition device.

[0072] Next, when obtaining the on-site construction images, video frames will be extracted from the target video segments corresponding to each image acquisition device respectively, and then a video frame corresponding to each image acquisition device will be obtained through similarity calculation. Finally, the video frames corresponding to each image acquisition device will be summarized to obtain the on-site construction image. Taking the Figure 5 scenario shown as an example, the on-site construction image obtained at this time contains four video frames.

[0073] It should be noted that when obtaining the on-site construction video for analysis and processing, in addition to periodic acquisition, it can also be by detecting whether there are construction objects entering the sub-region. When it is determined that there are construction objects entering, subsequent analysis and determination will be carried out. Anyway, only the on-site construction video can be recorded, and there is no specific limitation.

[0074] Step 104: Based on the on-site construction image, obtain the target image for violation determination corresponding to the sub-region.

[0075] In one embodiment, after obtaining the on-site construction image, analysis will be carried out according to the on-site construction image to obtain the target image for violation determination corresponding to the sub-region. Since the number of on-site construction images corresponding to different sub-regions is different, therefore, for different situations, the target image for violation determination can be obtained through different processing methods.

[0076] Among them, if the number of video frames in the on-site construction image is one, then the on-site construction image will be used as the target image for violation determination corresponding to the sub-region; if the number of video frames in the on-site construction image is greater than one, then image stitching will be carried out based on each video frame in the on-site construction image to obtain the corresponding stitched image, and the stitched image will be used as the target image for violation determination corresponding to the sub-region.

[0077] That is to say, when the number of video frames in the on-site construction image is one, it will be directly used as the target image for violation determination. While when the number of video frames in the on-site construction is greater than one, image stitching processing will be carried out. Each video frame in the on-site construction image will be stitched to obtain a new stitched image, and then the obtained stitched image will be used as the target image for violation determination.

[0078] Step 105: Input the target image into the pre-trained violation detection model to obtain the detection result corresponding to the sub-region, where the detection result includes whether there is a violation and the violation behavior.

[0079] In one embodiment, after obtaining the target image for violation determination, the target image will be input into the pre-trained violation detection model to obtain the detection result corresponding to the sub-region, where the detection result at least includes whether there is a violation and the specific violation behavior. Since there may be multiple violation behaviors in a sub-region, therefore, while determining whether there is a violation behavior, the specific violation behavior can also be determined.

[0080] And when it is determined that there is a violation behavior, recording and feedback can also be carried out in a timely manner, specifically including: when it is determined that the detection result is a violation, corresponding prompt information will be generated according to the violation behavior, and the target image and the violation moment will be recorded. By recording the violation moment and the violation behavior, and combining with the specific construction image (i.e., the target image) for recording.

[0081] Further, referring to Figure 6 , Figure 6 is a schematic flow chart of training a violation detection model provided by an embodiment of the present application. Among them, this step includes steps 601 to 603.

[0082] Step 601, obtain the collected historical construction images, and perform behavior marking processing on the historical construction images based on the violation construction behavior labels to obtain the marked historical construction images;

[0083] Step 602, perform sample expansion processing on the marked historical construction images to obtain training images for training. Among them, the expansion processing includes flipping, rotating, scaling, and cropping, and the size of each image in the training images is the same;

[0084] Step 603, build a violation detection model to be trained based on YOLOv5, and train the violation detection model based on the training images until the training is completed to obtain a trained violation detection model.

[0085] Specifically, when training the violation detection model, first obtain the training data for training, specifically historical construction images, and then use the violation construction labels to perform behavior marking processing on the historical construction images to obtain the marked historical construction images. Then, perform sample expansion processing on the marked historical construction images to increase the number of training samples. Among them, the expansion processing includes flipping, rotating, scaling, and cropping, and the processed images need to have the same size as the original images. Finally, build a violation detection model to be trained based on YOLOv5 to train the violation detection model based on the obtained training images until the training is completed to obtain a trained violation detection model.

[0086] Exemplarily, after completing the marking processing and expansion processing of the historical construction images, the obtained training images can be divided into a training set, a validation set, a test set, etc. according to training requirements, and then a violation detection model that can perform violation behavior detection and analysis can be obtained by training, validating, and testing the model.

[0087] Taking wearing a safety helmet as an example, a total of 10,000 images were obtained through simulation and online collection in the construction environment, specifically presenting the scenario of construction workers wearing safety helmets. Three categories were clearly defined, namely not wearing a safety helmet, correctly wearing a safety helmet, and wearing a safety helmet but not in a standard way. By expanding the original data set using enhancement techniques such as flipping, rotating, scaling, and cropping, and setting the image size to 640×640 pixels, a total of 50,000 images were finally generated. Among them, 70% of the data was used as the training set of YOLOv5, and the remaining 30% of the data was used as the test set.

[0088] When building a violation detection model based on YOLOv5, in order to improve the detection ability for small targets, the K-means++ algorithm (clustering algorithm) can be introduced to solve the problem of insensitivity to small targets. At the same time, the attention mechanism CBAM can be introduced into the YOLOv5 architecture to increase the proportion of small target features, and the alpha-loU loss function can also be introduced to enhance the robustness to small data sets.

[0089] When introducing the K-means++ algorithm to solve the problem of insensitivity to small targets, the following steps are included:

[0090] Step 1: Select the clustering center, which is randomly selected from the sample set D;

[0091] Step 2: Calculate the shortest distance d(x) from c to the sample point x;

[0092] Step 3: Calculate the probability p of x becoming a clustering center, and the calculation formula is as follows:

[0093] ;

[0094] Step 4: Find C = C1, C2... C k , and repeat the above steps;

[0095] Step 5: Cluster by distance;

[0096] Step 6: Output the result.

[0097] At the same time, if the structure of the introduced attention mechanism CBAM is as Figure 7 shown, the overall process of the attention mechanism CBAM is that the input feature first passes through the channel attention module, then is weighted, and then passes through the spatial attention module, and then is weighted with the former to obtain the final result.

[0098] Combined with the structure of the attention mechanism CBAM, the outputs of relevant nodes are as follows:

[0099] ;

[0100] ;

[0101] Among them, in the formula , , and are the input feature map, channel attention output, and spatial attention output respectively, is the weighting, , are the channel attention and spatial attention respectively.

[0102] Furthermore, the introduced alpha-loU loss function is as follows:

[0103] ;

[0104] where a is the weight in the formula.

[0105] In summary, by improving the existing YOLOv5 architecture and then using the improved YOLOv5 architecture to build a violation detection model for detecting illegal construction behaviors, where the improved YOLOv5 architecture is as Figure 8 shown, specifically introducing the attention mechanism CBAM structure.

[0106] Then during training, perform adaptive median filtering on the collected images, divide the dataset into a training set, a validation set, and a test set according to a certain ratio; merge the dataset into images of 640x640x3 and use the LabelImg tool to label the images; after preparing the data, initialize the parameters of the improved YOLOv5 model; train and evaluate the model through the training set and the validation set; test the test set with the trained model; output the test results.

[0107] In addition, to verify the superiority and effectiveness of the proposed detection method in detecting violations in intelligent substations, an experimental analysis system can also be configured for it, and the specific configuration is shown in Table 2 below:

[0108] Table 2

[0109]

[0110] To verify the performance of the proposed model in the complex scenarios of real substations, detect the wearing status and out-of-bounds of safety helmets of substation personnel, and collect images in the actual scenarios of a certain intelligent substation to construct a dataset, with a total of 10,020 images. The sample ratio is shown in Table 3:

[0111] Table 3

[0112]

[0113] Divide the dataset into a training set, a validation set, and a test set, with a ratio of 8:1:1. The experimental parameters are shown in Table 4:

[0114] Table 4

[0115]

[0116] When evaluating the trained model, use precision P, recall rate R, accuracy A, mean average precision mAP, and frames per second FPS to evaluate the model, as follows:

[0117] ;

[0118] ;

[0119] ;

[0120] ;

[0121] wherein, in the formula TP is the number of correctly detected positive samples; FP is the number of correctly detected negative samples; FN is the number of undetected positive samples: TN is the number of correctly detected negative samples; is the category i average precision; N is the number of categories. In the present invention, the number of frames of images detected per second is used to evaluate the speed of the model.

[0122] In summary, the above embodiments provide a method for artificial intelligence to identify violations in power projects. By obtaining the construction area and construction information of a construction project, marking the construction behavior of the construction area according to the construction information, and dividing the marked construction area based on the violation construction behavior label to obtain several sub-areas. Then, during violation detection, the on-site construction video collected by the image acquisition device set on the sub-area is obtained in real time, and the target image for violation determination is obtained through processing the on-site construction video. Finally, the pre-trained violation detection model is used to analyze and process the target image to obtain the detection result corresponding to the sub-area. This avoids the errors and untimely nature brought by manual monitoring, and at the same time improves the accuracy and timeliness of detecting violation construction behaviors in each area through the method of area division.

[0123] According to the method described in the above embodiments, this embodiment will further describe from the perspective of the device for artificial intelligence to identify violations in power projects. The device for artificial intelligence to identify violations in power projects can be specifically implemented as an independent entity, or can be integrated in an electronic device, such as a terminal. The terminal can include a mobile phone, a tablet computer, etc.

[0124] Please refer to Figure 9 , Figure 9 which is a schematic structural diagram of the device for artificial intelligence to identify violations in power projects provided by the embodiments of the present application. As Figure 9 shown, the device 900 for artificial intelligence to identify violations in power projects provided by the embodiments of the present application includes:

[0125] An information acquisition module 901, configured to obtain the construction area and construction information of a construction project, and mark the construction behavior of the construction area according to the construction information, wherein the construction information at least includes a construction plan and a construction equipment layout;

[0126] The area division module 902 is configured to divide the marked construction area based on the illegal construction behavior labels to obtain a number of sub-areas, where each sub-area contains at least one illegal construction behavior label;

[0127] The image acquisition module 903 is configured to acquire in real time the on-site construction video collected by the image acquisition device set on the sub-area, and obtain the on-site construction image corresponding to the sub-area from the on-site construction video;

[0128] The image processing module 904 is configured to obtain the target image for illegal determination corresponding to the sub-area based on the on-site construction image;

[0129] The illegal determination module 905 is configured to input the target image into a pre-trained illegal detection model to obtain the detection result corresponding to the sub-area, where the detection result includes whether there is an illegal act and the illegal behavior.

[0130] In one embodiment, the information acquisition module 901 is further configured to:

[0131] Construct a three-dimensional space corresponding to the construction area, and fill the three-dimensional space based on the layout of the construction equipment shown;

[0132] Perform marking processing of construction behaviors in the filled three-dimensional space according to the construction plan, and determine the space area in the filled three-dimensional space when the construction plan is executed.

[0133] In one embodiment, the area division module 902 is further configured to:

[0134] Obtain the illegal construction behavior labels corresponding to the execution of the construction plan;

[0135] Perform marking processing of space areas on the three-dimensional space based on the illegal construction behavior labels to obtain the marked three-dimensional space;

[0136] Based on the illegal construction behavior labels marked on the space areas, obtain the area dividing line of the three-dimensional space, and divide the three-dimensional space according to the area dividing line to obtain a number of sub-areas.

[0137] In one embodiment, the image acquisition module 903 is further configured to:

[0138] Acquire in real time the on-site construction videos collected by each image acquisition device set on the sub-area, and perform video frame extraction on the on-site construction videos to obtain a number of video frames, where the number of video frames corresponding to each image acquisition device is the same;

[0139] Perform similarity screening on the video frames corresponding to each image acquisition device to obtain a video frame corresponding to each image acquisition device, and summarize the video frames obtained by each image acquisition device to obtain the on-site construction image corresponding to the sub-region.

[0140] In one embodiment, the image acquisition module 903 is further configured to:

[0141] Obtain in real time the on-site construction videos collected by each image acquisition device set on the sub-region, and detect the construction objects in the on-site construction videos;

[0142] Intercept video segments of the on-site construction videos according to the detection results to obtain video segments containing construction objects, where the video segments contain timestamps;

[0143] Perform video alignment processing on the video segments corresponding to each image acquisition device according to the timestamps to obtain the target video segments corresponding to each image acquisition device for video frame extraction;

[0144] Extract video frames from the target video segments to obtain a number of video frames.

[0145] In one embodiment, the image processing module 904 is further configured to:

[0146] If the number of video frames in the on-site construction image is one, use the on-site construction image as the target image for violation determination corresponding to the sub-region;

[0147] If the number of video frames in the on-site construction image is greater than one, perform image stitching based on each video frame in the on-site construction image to obtain the corresponding stitched image, and use the stitched image as the target image for violation determination corresponding to the sub-region.

[0148] In one embodiment, the device 900 for artificial intelligence to identify power project violations further includes a prompt feedback module, which is used for:

[0149] When it is determined that the detection result is a violation, generate corresponding prompt information according to the violation behavior, and record the target image and the violation moment.

[0150] In one embodiment, the device 900 for artificial intelligence to identify power project violations further includes a model training module, which is used for:

[0151] Obtain the collected historical construction images, and perform behavior marking processing on the historical construction images based on the violation construction behavior labels to obtain the marked historical construction images;

[0152] Perform sample expansion processing on the historical construction images after marker processing to obtain training images for training. Among them, the expansion processing includes flipping, rotating, scaling, and cropping, and the sizes of each image in the training images are the same;

[0153] Build a violation detection model to be trained based on YOLOv5, and train the violation detection model based on the training images until the training is completed to obtain a trained violation detection model.

[0154] In addition, please refer to Figure 10 , Figure 10 which is a schematic structural diagram of an electronic device provided by an embodiment of the present application. The electronic device may be a mobile terminal such as a smart phone, a tablet computer, or other devices. As Figure 10 shown, the electronic device 1000 includes a processor 1001 and a memory 1002. Among them, the processor 1001 is electrically connected to the memory 1002.

[0155] The processor 1001 is the control center of the electronic device 1000, connects various parts of the entire electronic device using various interfaces and lines, executes various functions of the electronic device 1000 and processes data by running or loading application programs stored in the memory 1002 and calling data stored in the memory 1002, so as to monitor the electronic device 1000 as a whole.

[0156] In this embodiment, the processor 1001 in the electronic device 1000 will load the instructions corresponding to the processes of one or more application programs into the memory 1002 according to the following steps, and the processor 1001 will run the application programs stored in the memory 1002, so as to implement any step in the method for artificial intelligence to identify violations in power projects provided in the above embodiments.

[0157] The electronic device 1000 can implement the steps in any embodiment of the method for artificial intelligence to identify violations in power projects provided by the embodiments of the present application. Therefore, it can achieve the beneficial effects that any method for artificial intelligence to identify violations in power projects provided by the embodiments of the present application can achieve. For details, please refer to the previous embodiments and will not be elaborated here.

[0158] Please refer to Figure 11 , Figure 11 which is another schematic structural diagram of an electronic device provided by an embodiment of the present application. As Figure 11 shown, Figure 11 shows the specific structural block diagram of the electronic device provided by the embodiment of the present application. The electronic device can be used to implement the method for artificial intelligence to identify violations in power projects provided in the above embodiments. The electronic device 1100 may be a mobile terminal such as a smart phone or a laptop computer.

[0159] The RF circuit 1110 is used to receive and transmit electromagnetic waves, realizing the mutual conversion between electromagnetic waves and electrical signals, so as to communicate with a communication network or other devices. The RF circuit 1110 may include various existing circuit components for performing these functions, such as antennas, radio frequency transceivers, digital signal processors, encryption / decryption chips, subscriber identity module (SIM) cards, memories, and so on. The RF circuit 1110 can communicate with various networks such as the Internet, enterprise intranets, wireless networks or communicate with other devices through wireless networks. The above-mentioned wireless networks may include cellular phone networks, wireless local area networks or metropolitan area networks. The above-mentioned wireless networks can use various communication standards, protocols and technologies, including but not limited to Global System for Mobile Communication (GSM), Enhanced Data GSM Environment (EDGE), Wideband Code Division Multiple Access (WCDMA), Code Division Access (CDMA), Time Division Multiple Access (TDMA), Wireless Fidelity (Wi-Fi) (such as Institute of Electrical and Electronics Engineers standards IEEE 802.11a, IEEE 802.11b, IEEE 802.11g and / or IEEE 802.11n), Voice over Internet Protocol (VoIP), Worldwide Interoperability for Microwave Access (Wi-Max), other protocols for email, instant messaging and short messages, and any other suitable communication protocols, and may even include those protocols that have not been developed yet.

[0160] The memory 1120 can be used to store software programs and modules, such as the program instructions / modules corresponding to the method for artificial intelligence to identify power project violations in the above embodiments. The processor 1180 executes various functional applications and the method for artificial intelligence to identify power project violations by running the software programs and modules stored in the memory 1120.

[0161] The memory 1120 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 1120 may further include a memory remotely located relative to the processor 1180, and these remote memories may be connected to the electronic device 1100 through a network. Examples of the above network include but are not limited to the Internet, intranet, local area network, mobile communication network, and combinations thereof.

[0162] The input unit 1130 can be used to receive uploaded digital or character information and generate keyboard, mouse, joystick, optical, or trackball signal inputs related to user settings and function controls. Specifically, the input unit 1130 may include a touch-sensitive surface 1131 and other input devices 1132. The touch-sensitive surface 1131, also known as a touch display screen or touchpad, can collect touch operations of the user on or near it (such as operations of the user using a finger, stylus, or any suitable object or accessory on or near the touch-sensitive surface 1131), and drive the corresponding connection device according to a pre-set program. Optionally, the touch-sensitive surface 1131 may include two parts: a touch detection device and a touch controller. Among them, the touch detection device detects the touch position of the user and the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, converts it into contact coordinates, and then sends it to the processor 1180, and can receive and execute the commands sent by the processor 1180. In addition, the touch-sensitive surface 1131 can be implemented in multiple types such as resistive, capacitive, infrared, and surface acoustic wave. In addition to the touch-sensitive surface 1131, the input unit 1130 may further include other input devices 1132. Specifically, the other input devices 1132 may include but are not limited to one or more of a physical keyboard, function keys (such as volume control keys, power on / off keys, etc.), trackball, mouse, joystick, etc.

[0163] The display unit 1140 can be used to display information input by the user or information provided to the user, as well as various graphical user interfaces of the electronic device 1100. These graphical user interfaces can be composed of graphics, text, icons, videos, and any combination thereof. The display unit 1140 may include a display panel 1141. Optionally, the display panel 1141 can be configured in the form of an LCD (Liquid Crystal Display), an OLED (Organic Light-Emitting Diode), etc. Further, the touch-sensitive surface 1131 can cover the display panel 1141. When the touch-sensitive surface 1131 detects a touch operation on or near it, it is transmitted to the processor 1180 to determine the type of touch event. Subsequently, the processor 1180 provides a corresponding visual output on the display panel 1141 according to the type of touch event. Although in the figure, the touch-sensitive surface 1131 and the display panel 1141 are implemented as two independent components to achieve input and output functions, in some embodiments, the touch-sensitive surface 1131 and the display panel 1141 can be integrated to achieve input and output functions.

[0164] The electronic device 1100 may further include at least one sensor 1150, such as a light sensor, a motion sensor, and other sensors. Specifically, the light sensor may include an ambient light sensor and a proximity sensor. Among them, the ambient light sensor can adjust the brightness of the display panel 1141 according to the brightness of the ambient light, and the proximity sensor can generate an interruption when the flip cover is closed or opened. As a type of motion sensor, the gravity acceleration sensor can detect the magnitude of acceleration in all directions (generally three axes). When stationary, it can detect the magnitude and direction of gravity, and can be used in applications for identifying the posture of the mobile phone (such as horizontal and vertical screen switching, related games, magnetometer posture calibration), vibration recognition related functions (such as pedometer, tapping), etc. As for other sensors that the electronic device 1100 can also be configured with, such as a gyroscope, a barometer, a hygrometer, a thermometer, an infrared sensor, etc., they will not be elaborated here.

[0165] The audio circuit 1160, the speaker 1161, and the microphone 1162 can provide an audio interface between the user and the electronic device 1100. The audio circuit 1160 can transmit the electrical signal converted from the received audio data to the speaker 1161, and the speaker 1161 converts it into a sound signal for output. On the other hand, the microphone 1162 converts the collected sound signal into an electrical signal, which is received by the audio circuit 1160 and then converted into audio data. After the audio data is output to the processor 1180 for processing, it is sent to another terminal, for example, via the RF circuit 1110, or the audio data is output to the memory 1120 for further processing. The audio circuit 1160 may also include an earphone jack to provide communication between the external earphone and the electronic device 1100.

[0166] The electronic device 1100 can help users receive requests, send information, etc. through a transmission module 1170 (such as a Wi-Fi module), which provides users with wireless broadband Internet access. Although the transmission module 1170 is shown in the figure, it can be understood that it does not belong to the essential components of the electronic device 1100 and can be omitted entirely within the scope of not changing the essence of the invention as needed.

[0167] The processor 1180 is the control center of the electronic device 1100, connecting various parts of the entire mobile phone through various interfaces and circuits. By running or executing software programs and / or modules stored in the memory 1120, and by calling the data stored in the memory 1120, it executes various functions of the electronic device 1100 and processes data, thereby monitoring the electronic device as a whole. Optionally, the processor 1180 may include one or more processing cores; in some embodiments, the processor 1180 may integrate an application processor and a modem processor. Among them, the application processor mainly processes the operating system, user interface, and application programs, etc., and the modem processor mainly processes wireless communications. It can be understood that the above-mentioned modem processor may not be integrated into the processor 1180.

[0168] The electronic device 1100 also includes a power supply 1190 (such as a battery) for powering each component. In some embodiments, the power supply can be logically connected to the processor 1180 through a power management system, so as to realize functions such as management of charging, discharging, and power consumption management through the power management system. The power supply 1190 may also include any components such as one or more DC or AC power supplies, a recharge system, a power failure detection circuit, a power converter or inverter, and a power status indicator.

[0169] Although not shown, the electronic device 1100 also includes a camera (such as a front camera and a rear camera), a Bluetooth module, etc., which will not be elaborated here. Specifically, in this embodiment, the display unit of the electronic device is a touch screen display, and the mobile terminal also includes a memory, and one or more programs, where one or more programs are stored in the memory and are configured to be executed by one or more processors to implement any step in the method for artificial intelligence to identify power project violations provided in the above embodiments.

[0170] In specific implementation, the above-mentioned each module can be implemented as an independent entity, or can be combined arbitrarily to be implemented as the same or several entities. For the specific implementation of the above-mentioned each module, reference can be made to the method embodiments above, which will not be elaborated here.

[0171] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructions or by controlling relevant hardware through instructions. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor. For this purpose, an embodiment of the present application provides a storage medium in which multiple instructions are stored, and when these instructions are executed by a processor, any step in the method for artificial intelligence to identify power project violations provided by the above embodiments can be implemented.

[0172] Among them, the storage medium may include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk, etc.

[0173] Since the instructions stored in the storage medium can execute the steps in any embodiment of the method for artificial intelligence to identify power project violations provided by the embodiments of the present application, the beneficial effects achievable by any method for artificial intelligence to identify power project violations provided by the embodiments of the present application can be achieved. For details, please refer to the previous embodiments and will not be elaborated here.

[0174] The above has introduced in detail a method, device, electronic device and storage medium for artificial intelligence to identify power project violations provided by the embodiments of the present application. Specific examples are used in this article to elaborate on the principle and implementation manner of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present application. Moreover, for those of ordinary skill in the technical field, without departing from the principle of the present application, several improvements and refinements can be made, and these improvements and refinements are also regarded as the protection scope of the present application.

Claims

1. A method for an artificial intelligence to identify violations in power projects, characterized in that, Including: Obtain the construction area and construction information of a construction project, and mark the construction behavior of the construction area according to the construction information, where the construction information at least includes a construction plan and a construction equipment layout; Divide the marked construction area based on the illegal construction behavior labels corresponding to the execution of the construction plan to obtain several sub-areas, where each sub-area at least contains one illegal construction behavior label; Obtain in real time the on-site construction video collected by the image acquisition device set on the sub-area, and obtain the on-site construction image corresponding to the sub-area from the on-site construction video; Based on the on-site construction image, obtain the target image for illegal determination corresponding to the sub-area; Input the target image into a pre-trained illegal detection model to obtain the detection result corresponding to the sub-area, where the detection result includes whether there is an illegal act and the illegal behavior.

2. The method according to claim 1, characterized in that, The marking of the construction behavior of the construction area according to the construction information includes: Construct a three-dimensional space corresponding to the construction area, and fill the three-dimensional space based on the construction equipment layout; Perform marking processing of the construction behavior in the filled three-dimensional space according to the construction plan, and determine the space area in the filled three-dimensional space when the construction plan is executed.

3. The method according to claim 2, wherein The dividing of the marked construction area based on the illegal construction behavior labels corresponding to the execution of the construction plan to obtain several sub-areas includes: Obtain the illegal construction behavior labels corresponding to the execution of the construction plan; Perform marking processing of the space area on the three-dimensional space based on the illegal construction behavior labels to obtain the marked three-dimensional space; Based on the illegal construction behavior labels marked by the space area, obtain the regional dividing line of the three-dimensional space, and divide the three-dimensional space according to the regional dividing line to obtain several sub-areas.

4. The method according to claim 1, wherein The obtaining in real time the on-site construction video collected by the image acquisition device set on the sub-area and obtaining the on-site construction image corresponding to the sub-area from the on-site construction video includes: Obtain in real time the on-site construction videos collected by each image acquisition device set on the sub-area, and extract video frames from the on-site construction videos to obtain several video frames, where the number of video frames corresponding to each image acquisition device is the same; Perform similarity screening of the images for the video frames corresponding to each image acquisition device to obtain one video frame corresponding to each image acquisition device, and summarize the one video frame obtained by each image acquisition device to obtain the on-site construction image corresponding to the sub-area.

5. The method according to claim 4, wherein The obtaining in real time the on-site construction videos collected by each image acquisition device set on the sub-area and extracting video frames from the on-site construction videos to obtain several video frames includes: Obtain in real time the on-site construction videos collected by each image acquisition device set on the sub-area, and detect the construction objects in the on-site construction videos; Intercept video segments from the on-site construction video according to the detection results to obtain video segments containing construction objects, where the video segments contain timestamps; Perform video alignment processing on the video segments corresponding to each image acquisition device according to the timestamps to obtain target video segments corresponding to each image acquisition device for video frame extraction; Extract video frames from the target video segments to obtain a number of video frames.

6. The method according to claim 4, characterized in that The obtaining of the target image for violation determination corresponding to the sub-region based on the on-site construction image includes: If the number of video frames in the on-site construction image is one, use the on-site construction image as the target image for violation determination corresponding to the sub-region; If the number of video frames in the on-site construction image is greater than one, perform image stitching based on each video frame in the on-site construction image to obtain a corresponding stitched image, and use the stitched image as the target image for violation determination corresponding to the sub-region.

7. The method according to claim 1, characterized in that, After inputting the target image into a pre-trained violation detection model to obtain the detection result corresponding to the sub-region, it further includes: When it is determined that the detection result is a violation, generate corresponding prompt information according to the violation behavior, and record the target image and the violation moment.

8. The method according to claim 1, wherein The training process of the violation detection model includes: Obtain the collected historical construction images, and perform behavior marking processing on the historical construction images based on the violation construction behavior labels to obtain the marked historical construction images; Perform sample expansion processing on the marked historical construction images to obtain training images for training, where the expansion processing includes flipping, rotating, scaling, and cropping, and the sizes of each image in the training images are the same; Build a violation detection model to be trained based on YOLOv5, and train the violation detection model based on the training images until the training is completed to obtain a trained violation detection model.

9. An apparatus for identifying violations in a power project by artificial intelligence, characterized in that, It includes: An information acquisition module, configured to acquire the construction area and construction information of the construction project, and mark the construction behavior of the construction area according to the construction information, where the construction information includes at least the construction plan and the construction equipment layout; A region division module, configured to divide the marked construction area based on the violation construction behavior labels corresponding to the execution of the construction plan to obtain a number of sub-regions, where each sub-region contains at least one violation construction behavior label; An image acquisition module, configured to acquire the on-site construction video collected by the image acquisition device set on the sub-region in real time, and obtain the on-site construction image corresponding to the sub-region from the on-site construction video; An image processing module, configured to obtain the target image for violation determination corresponding to the sub-region based on the on-site construction image; A violation determination module, configured to input the target image into a pre-trained violation detection model to obtain the detection result corresponding to the sub-region, where the detection result includes whether there is a violation and the violation behavior.

10. 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, it implements the steps in the method according to any one of claims 1 to 7.

Images