Method, system, equipment and medium for AI on-line identification of scaffold building specifications

Abstract

The invention relates to the technical field of petrochemical engineering construction, and discloses a method, a system, equipment and a medium for AI online identification of scaffold building specifications. The method comprises the following steps: constructing a target detection network model taking a scaffold construction specification as a target; training the target detection network model through training data; collecting scaffold construction video data; inputting the video data into a trained target detection network model, and identifying whether the video data accord with a scaffold building specification; and if not, risk early warning is sent to the scaffold mobile terminal. And the safety of scaffold construction on a petrochemical engineering construction operation site is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of petrochemical construction, and particularly to a method, system, device, and medium for AI online recognition of scaffolding erection specifications. Background Art

[0002] Nowadays, in petrochemical enterprises, scaffolding is a common construction equipment, and scaffolding needs to be erected for civil construction, equipment installation, anti-corrosion insulation, and process construction of the device. With the construction of ten-million-ton oil refining and million-ton ethylene plants in petrochemical enterprises, the amount of high-altitude operation in engineering construction and installation is large, the process technology is becoming more and more complex, the difficulty and amount of scaffolding erection required are gradually increasing, and the safety requirements are becoming higher; the construction technology and safety risk prevention and control of scaffolding erection are also increasing.

[0003] The current scaffolding is often managed through traditional work tickets and paper acceptance forms throughout the process from erection, acceptance to use. Inevitably, there are various deficiencies in use: the management difficulty is large, the manual consumption is heavy, resulting in a waste of human resources; the manual verification is not accurate enough, and misjudgment and missed inspection are likely to occur, resulting in potential safety hazards; the feedback of manual monitoring is often not timely enough to ensure the timeliness of feedback information; through manual monitoring, there is a lack of accurate and detailed electronic data records, and it is impossible to effectively trace and supervise the reasons for problems afterwards, resulting in supervision loopholes.

[0004] Therefore, it is necessary to invent a method for identifying the standard operation of scaffolding to strictly monitor the process of scaffolding construction in real time, reduce waste of manpower, and eliminate potential safety hazards. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems in the prior art that the manual work is heavy, the management difficulty is large, misjudgment and missed inspection are likely to occur, it is difficult to ensure timeliness during the process of scaffolding erection and use, and there is a lack of electronic data records and supervision loopholes.

[0006] To achieve the above purpose, in the first aspect, the present invention provides a method for AI online recognition of scaffolding erection specifications, and the method includes:

[0007] Construct a target detection network model with scaffolding erection specifications as the target;

[0008] Train the target detection network model with training data;

[0009] Collect video data of scaffolding erection;

[0010] Input the video data into the trained target detection network model to identify whether it meets the scaffolding erection specifications; if not, send a risk warning to the scaffolding mobile terminal.

[0011] Furthermore, the target detection network model adopts the YOLO algorithm.

[0012] Furthermore, the target detection network model sequentially includes a CSPDarkNet53 neural network, an SPP+PAN network, and a YOLO-Head module.

[0013] Furthermore, the scaffolding erection specifications include face recognition, scaffolding tagging specifications, scaffolding fitting installation specifications, isolation specifications, and rectification specifications.

[0014] Furthermore, the collection of scaffolding erection video data includes:

[0015] Collecting scaffolding erection video data through a handheld terminal and a camera.

[0016] Furthermore, identifying whether the scaffolding erection complies with the specifications; if not, issuing a risk warning, including:

[0017] Performing face recognition on the personnel in the scaffolding erection video data to determine the personnel category and whether the construction personnel's certificates are within the valid period;

[0018] Identifying the scaffolding tags in the scaffolding erection video data and identifying the color of the tags so that the scaffolding mobile terminal can determine whether the scaffolding signs are correctly hung;

[0019] Identifying the installation of scaffolding fittings in the scaffolding erection video data, identifying whether the erected scaffolding is provided with wall connecting members, cross braces, and skirting boards, and whether life ropes, safety nets, and anti-falling devices are installed synchronously, and whether there is a situation where double hooks are disengaged simultaneously;

[0020] Performing hard isolation identification on the scaffolding in the scaffolding erection video data to identify whether there are hard isolation measures at the scaffolding operation site during the erection and rectification states;

[0021] Performing climbing identification on the scaffolding in the scaffolding erection video data to identify whether anyone is climbing on the scaffolding with a no-climbing sign hung;

[0022] Performing rectification identification on the scaffolding in the scaffolding erection video data to identify whether the scaffolding with signs hung and without hard isolation measures is replaced with a rectification sign within a predetermined time.

[0023] Furthermore, after identifying whether the scaffolding erection complies with the specifications, it further includes:

[0024] Using the scaffolding erection video data as training data to perform iterative training on the target detection network model.

[0025] Second aspect, the present invention provides a system for AI online recognition of scaffolding erection specifications, the system comprising:

[0026] A model construction module for constructing a target detection network model with scaffolding erection specifications as the target;

[0027] A model training module for training the target detection network model with training data;

[0028] A data acquisition module for acquiring scaffolding erection video data;

[0029] A specification recognition module for inputting the video data into the trained target detection network model to recognize whether it conforms to the scaffolding erection specifications; if not, a risk warning is sent to the scaffolding mobile terminal.

[0030] Third aspect, the present invention provides a computer device, the computer device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the method for AI online recognition of scaffolding erection specifications as described above when executing the computer program.

[0031] Fourth aspect, the present invention provides a computer-readable storage medium, the computer-readable storage medium comprising a stored computer program; wherein, the computer program controls the device where the computer-readable storage medium is located to execute the method for AI online recognition of scaffolding erection specifications as described above when running.

[0032] The method, system, device, and medium for AI online recognition of scaffolding erection specifications of the present invention, compared with the prior art, have the beneficial effects that: the present invention utilizes image recognition technology to intelligently analyze the video data collected by the camera, recognizes key components during the scaffolding erection process, and accurately determines whether they are operating at the specified positions according to the specified operation procedures, ensuring the safety of the entire scaffolding erection process; when it is determined that the scaffolding is abnormal, the determination result is timely sent to the mobile terminal, ensuring the real-time nature of the scaffolding safety monitoring; the electronic data records collected during the work process can effectively supervise and track the construction process; at the same time, the artificial intelligence recognizes the scaffolding work process, which can save labor and reduce production costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 is a flowchart of the method for AI online recognition of scaffolding erection specifications provided by an embodiment of the present invention;

[0034] Figure 2 is the network structure of the target detection algorithm of the method for AI online recognition of scaffolding erection specifications provided by an embodiment of the present invention;

[0035] Figure 3 It is the working process diagram of the system for AI online recognition of scaffolding erection specifications provided by the embodiments of the present invention;

[0036] Figure 4 It is the structural block diagram of the system for AI online recognition of scaffolding erection specifications provided by the embodiments of the present invention;

[0037] Figure 5 It is the structural diagram of the computer device provided by the embodiments of the present invention. Specific Embodiments

[0038] The following combines the accompanying drawings and embodiments to further describe in detail the specific embodiments of the present invention. The following embodiments are used to illustrate the present invention, but are not used to limit the scope of the present invention.

[0039] It should be noted that the step numbers in the text are only for the convenience of explaining specific embodiments and do not serve to limit the execution order of the steps. The method provided in this embodiment can be executed by a related server, and the server is used as the execution subject in the following for illustration.

[0040] As Figure 1 shown, the embodiments of the present invention provide a method for AI online recognition of scaffolding erection specifications, including steps S11 to S14:

[0041] Step S11, constructing a target detection network model with scaffolding erection specifications as the target;

[0042] As Figure 2 shown, using the YOLO algorithm model, constructing a target detection network model that sequentially includes a CSPDarkNet53 neural network, an SPP+PAN network, and a YOLO-Head module;

[0043] During the scaffolding construction process, due to the high requirement for real-time information transmission, when selecting a detector, a backbone network with strong image feature extraction ability will be considered. The structure cannot be too large, as a too large structure will seriously affect the detection speed, nor can it be too small, as a too small structure will not extract target features well. The backbone network in the selected target detection network model in this solution selects the Darknet53 network with a CSP structure as the BackBone. After adding the CSP structure, the problem of duplicate image information collected by mobile terminals during the scaffolding erection process is solved, and in the network gradient optimization process of subsequent model training, the number of model parameters and the number of floating-point operations per second are greatly reduced, resulting in an improvement in the inference speed of the final model.

[0044] After improving the MAP metric through CSPDarkNet53, the Neck part uses the SPP+PAN network to extract feature maps of different levels, in order to expand the receptive field and fuse the information of feature maps of different scales, so as to better perform feature fusion. Based on FPN, PAN adds bottom-up feature fusion.

[0045] The Head part uses the YOLO algorithm to perform object detection on the image using multi-scale features, accurately detect abnormal information during the scaffolding construction process from the video, and give its position and category information.

[0046] Step S12: Train the object detection network model with training data;

[0047] Collect a large amount of image information of different types of illegal operations, input it into the object detection network model constructed in step S11 for model training. Each time, a small batch of samples is extracted and input into the object detection network model to obtain predicted values. Calculate the loss function value between the predicted values and the true values, and then perform gradient backpropagation. According to the obtained results, optimize and adjust the parameters of the model, and continuously update the model to gradually adapt to the scaffolding construction detection.

[0048] The training data used for training the object detection network model of the present invention can not only collect a large amount of image information of different types of illegal operations, but also adopt the training data in the existing general object recognition process.

[0049] Step S13: Collect scaffolding construction video data;

[0050] After the scaffolding construction starts, the handheld terminals, other sensors, and cameras at the construction site simultaneously start to collect the on-site construction video in real time, and transmit the data to the server side or the AI platform, and the server or the AI platform executes AI online recognition of the scaffolding construction specifications.

[0051] Step S14: Input the video data into the trained object detection network model to identify whether it conforms to the scaffolding construction specifications; if not, send a risk warning to the scaffolding mobile terminal.

[0052] Cut the video data collected in step S13, take frames at every fixed 5s interval, and input them into the object detection network model obtained after training in step S12 for detection; as Figure 3 shown, the detection items include:

[0053] Identify the personnel in the scaffolding construction video data, judge whether the personnel appearing in the video are employees of petrochemical enterprises, whether there are corresponding leading operation leaders, technical responsible persons and supervisors, and whether the construction personnel certificates are within the validity period;

[0054] Identify the colors, shapes, and patterns on the signs hung at the scaffolding construction site, so that the mobile end of the scaffolding can determine whether the signs are correctly hung on the scaffolding, achieving the supervision goal of "there must be a sign for every scaffold".

[0055] Identify whether the erected scaffolding is provided with wall connecting members, tipping braces, and skirting boards, and whether life ropes, safety nets, and anti-falling devices are installed synchronously; and whether there is a situation where the double hooks are disengaged simultaneously. During the work of the scaffolder, the double hooks of the safety belt are tied to different fixed points, and the double hooks shall not be disengaged simultaneously during climbing up and down and horizontal movement.

[0056] If the scaffolding is not completed in erection, rectification, and acceptance, it must be rigidly isolated; identify the rigid isolation of the scaffolding in the scaffolding erection video data to determine whether there are rigid isolation measures at the scaffolding operation site during erection and rectification.

[0057] At the same time, climbing or using unaccepted scaffolding is prohibited. Identify whether anyone climbs on the scaffolding with a no-climbing sign hung in the scaffolding erection video data.

[0058] For the scaffolding with a bottom sign, a red card, a no-climbing sign hung, and no rigid isolation measures, rectification shall be carried out within 24 hours, and it shall be replaced with a "green card" or "climbing entrance", etc., and identify whether it is overdue for rectification.

[0059] If any behavior that does not conform to the scaffolding erection specification, including the above situations, is detected, immediately send a risk warning to the mobile end of the scaffolding, notify the management personnel at the construction site to handle the safety hazards, and make an electronic record for archiving. Of course, the detection of the scaffolding erection specification in the present invention is not limited to the above detection items. It is only used to explain that the present invention performs real-time identification and risk warning on the video data during the scaffolding erection or acceptance process through an object detection model based on a neural network. For other items during the scaffolding erection process, the video data can also be identified through this model.

[0060] Use the video and image information corresponding to the behavior that does not conform to the scaffolding erection specification identified in step S14 as training data, incorporate it into the training data set, and perform iterative training on the object detection network model obtained in step S12 to make it adapt to the specification identification during the scaffolding erection process.

[0061] This method constructs an object detection network model with fast recognition speed, high mAP real-time detection network, and avoiding misrecognition of objects by adopting the YOLO algorithm model; iterates the object detection network model by collecting a large amount of image information of different types of illegal operations and using the training data in the existing general object recognition process to make it gradually more adaptable to the scaffolding construction process; recognizes the image information collected at the construction site of scaffolding erection to determine whether there are behaviors that do not conform to the scaffolding erection specifications, and at the same time uses the illegal images that do not conform to the scaffolding erection specifications obtained during the execution process as training data to gradually improve the object detection network model.

[0062] As Figure 4 shown, the embodiment of the present invention also provides a system for AI online recognition of scaffolding erection specifications, which is used to execute the method for AI online recognition of scaffolding erection specifications as described above. The system includes:

[0063] A model construction module 21, which is used to construct an object detection network model with scaffolding erection specifications as the target;

[0064] Adopt the YOLO algorithm model to construct an object detection network model that sequentially includes a CSPDarkNet53 neural network, an SPP+PAN network, and a YOLO-Head module;

[0065] A model training module 22, which is used to train the object detection network model with training data;

[0066] Collect a large amount of image information of different types of illegal operations, input it into the object detection network model constructed by the model construction module 21 for model training. Each time, a small batch of samples is extracted and input into the object detection network model to obtain a predicted value, calculate the loss function value between its predicted value and the true value, then perform gradient backpropagation, and according to the obtained results, optimize and adjust the parameters of the model, continuously update the model, and make it gradually adapt to scaffolding construction detection.

[0067] A data acquisition module 23, which is used to acquire scaffolding erection video data;

[0068] After the scaffolding construction starts, the handheld terminals, other sensors, and cameras at the construction site simultaneously start to collect the on-site construction video in real time and transmit the data to the server side.

[0069] A specification recognition module 24, which is used to input the video data into the trained object detection network model to identify whether it conforms to the scaffolding erection specifications; if not, a risk warning is sent to the scaffolding mobile terminal.

[0070] Preferably, the video and image information corresponding to the behaviors that do not conform to the scaffolding erection specifications identified by the specification recognition module 24 can be used as training data, incorporated into the training data set, and iteratively trained by the target detection network model obtained by inputting into the model training module.

[0071] Taking the AI platform integrating the above system of the present invention as an example, the AI online recognition process will be described in detail. As Figure 3 shown, the working process of the present invention for one time is as follows:

[0072] The scaffolding operation process starts, and at the same time, this system starts to work; during the operation process of the system, the information exchanges and pushes between the AI platform and the scaffolding end are recorded throughout the process.

[0073] The system terminal sensor collects the working state and work ticket state information of the scaffolding. The state collection items of the scaffolding include: erection, rectification, use, demolition, deactivation, etc.; the work ticket state includes start, pause, and close. After the collection is completed, the above information, work ticket ID, and camera ID are transmitted to the target system: the AI platform together.

[0074] The AI platform obtains the face pictures of the corresponding scaffolders through the work ticket ID and camera ID, compares and verifies them with the face information collected at the camera end, and at the same time verifies whether the scaffolders wear special operation certificates as required and the certificates are within the validity period. During this process, the camera records and uploads the surrounding other face information at the same time, which is used to verify whether there is a person in charge to guide the erection and demolition of the scaffolding throughout the process, whether there is a leader of high-risk erection on duty, and whether there is a supervisor on the side for supervision.

[0075] Meanwhile, the camera monitors the scaffolding state in real time throughout the process and pushes it as soon as there is a state change. When the scaffolding changes its state, the system sends a signal to the AI platform, and transmits the current image information of whether there is hard isolation together with the work ticket ID and camera ID to the AI platform. The AI platform returns the result of whether there is hard isolation, and the recognition result of whether the target scaffolding hangs the bottom card, red card, and no-climbing sign. For the scaffolding without hard isolation measures, the rectification shall not exceed 24 hours.

[0076] When problems occur during the scaffolding operation and the operation stops, the state of the work ticket switches to pause, and the camera transmits the collected work ticket state, work ticket ID, and camera ID to the AI platform together.

[0077] After the problem troubleshooting is completed, the work process of the work ticket switches back to start, and the scaffolding end returns the camera abnormal information, work ticket ID, and camera ID to the AI platform.

[0078] During the working process, the following information is pushed every 15 minutes: work ticket ID, camera ID, and recognition results. The recognition results include original violation pictures, reasons for violations, camera abnormalities, etc. This behavior is to ensure: real-time sign management for scaffolding, "where there is a scaffold, there must be a sign"; when erecting a scaffold, connecting members, braces, and toe boards should be set up synchronously; when erecting a scaffold, life ropes, safety nets, and anti-falling devices should be installed synchronously; if the scaffold is not completed in erection, rectification, and acceptance, hard isolation must be carried out on this scaffold.

[0079] In addition to the routine push every 15 minutes, when an abnormality is found during the monitoring process, it is immediately uploaded together with the work ticket ID and camera ID. The inspection items include: during the operation of the scaffolder, the safety double hooks must be attached to different fixed points, and when climbing up and down and moving horizontally, the double hooks must not be released simultaneously; it is prohibited to climb or use a scaffold that has not been accepted.

[0080] After the work is completed, the status of the work ticket is switched to closed and uploaded to the AI platform together with the work ticket ID and camera ID. When the operation process of the work ticket ends, the system stops the operation of this work.

[0081] The technical features and technical effects of the system proposed in the embodiments of the present invention are the same as those of the method proposed in the embodiments of the present invention, and will not be elaborated here. Each module in the above system can be implemented in whole or in part by software, hardware, and their combinations. The above-mentioned modules can be embedded in the processor of the computer device in hardware form or be independent of it, or can be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above modules.

[0082] As Figure 5 shown, the embodiments of the present invention also provide a computer device. The figure is a structural block diagram of a preferred embodiment of a computer device provided by the present invention. The computer device includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the method for AI online recognition of scaffolding erection specifications as described above.

[0083] Preferably, the computer program can be divided into one or more modules / units (such as computer program 1, computer program 2,...). The one or more modules / units are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, and this instruction segment is used to describe the execution process of the computer program in the computer device.

[0084] The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor, or the processor may also be any conventional processor. The processor is the control center of the computer device and connects various parts of the computer device through various interfaces and lines.

[0085] The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function, etc., and the data storage area can store relevant data, etc. In addition, the memory may be a high-speed random access memory, or may also be a non-volatile memory, such as a plug-in hard disk, a SmartMedia Card (SMC), a Secure Digital (SD) card, a Flash Card, etc., or the memory may also be other volatile solid-state storage devices.

[0086] It should be noted that the above computer device may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that Figure 5 The structural block diagram is only an example of the computer device and does not constitute a limitation on the computer device. It may include more or fewer components than shown, or combine certain components, or different components.

[0087] In summary, the method, system, device, and medium for providing the AI online recognition scaffolding erection specification in the embodiments of the present invention utilize image recognition technology to intelligently analyze the video data collected by the camera, identify key components during the scaffolding erection process, and accurately judge whether they are operating at the specified positions according to the specified operation procedures, ensuring the safety of the entire scaffolding erection process; when it is determined that the scaffolding is abnormal, the judgment result is sent to the mobile terminal in a timely manner, ensuring the real-time nature of the scaffolding safety monitoring; the electronic data records collected during the work process can effectively supervise and track the construction process; at the same time, the artificial intelligence recognizes the work process of the scaffolding, which can save labor and reduce production costs.

[0088] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the counting principle of the present invention, several improvements and replacements can be made, and these improvements and replacements should also be regarded as the protection scope of the present invention.

[0089] In the description of this specification, the description of terms such as "one embodiment", "some embodiments", "embodiment", "example", "specific example" or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0090] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present invention.

Claims

1. An AI online method for identifying scaffolding erection specifications, characterized in that, the method includes: Constructing an object detection network model with scaffolding erection specifications as the target; Training the object detection network model with training data; Collecting scaffolding erection video data; Inputting the video data into the trained object detection network model to identify whether it conforms to the scaffolding erection specifications; if not, sending a risk warning to the scaffolding mobile terminal.

2. The AI online method for identifying scaffolding erection specifications according to claim 1, characterized in that, the object detection network model adopts the YOLO algorithm.

3. The AI online method for identifying scaffolding erection specifications according to claim 2, characterized in that, the object detection network model sequentially includes a CSPDarkNet53 neural network, an SPP+PAN network, and a YOLO-Head module.

4. The AI online method for identifying scaffolding erection specifications according to claim 1, characterized in that, the scaffolding erection specifications include face recognition, scaffolding signage specifications, scaffolding fitting installation specifications, isolation specifications, and rectification specifications.

5. The AI online method for identifying scaffolding erection specifications according to claim 1, characterized in that, the collection of scaffolding erection video data includes: Collecting scaffolding erection video data through a handheld terminal and a camera.

6. The AI online method for identifying scaffolding erection specifications according to claim 4, characterized in that, identifying whether the scaffolding erection conforms to the specifications; if not, sending a risk warning, including: Performing face recognition on the personnel in the scaffolding erection video data to determine the personnel category and whether the construction personnel's certificates are within the valid period; Identifying the scaffolding signage in the scaffolding erection video data and identifying the color of the signage so that the scaffolding mobile terminal can determine whether the scaffolding is correctly hung with a sign; Identifying the installation of scaffolding fittings in the scaffolding erection video data, identifying whether the erected scaffolding is provided with tie-ins, braces, and toe boards, and whether the life rope, safety net, and anti-fall device are installed synchronously, and whether there is a situation where the double hooks are disengaged simultaneously; Performing hard isolation identification on the scaffolding in the scaffolding erection video data to identify whether there are hard isolation measures at the scaffolding operation site during the erection and rectification states; Performing climbing identification on the scaffolding in the scaffolding erection video data to identify whether there are people climbing on the scaffolding with a no-climbing sign hung; Performing rectification identification on the scaffolding in the scaffolding erection video data to identify whether the scaffolding with a sign hung and without hard isolation measures is replaced with a rectification sign within a predetermined time.

7. The AI online method for identifying scaffolding erection specifications according to claim 1, characterized in that, after identifying whether the scaffolding erection conforms to the specifications, it further includes: Using the scaffolding erection video data as training data to perform iterative training on the object detection network model.

8. An AI online system for identifying scaffolding erection specifications, characterized in that, the system includes: A model construction module for constructing an object detection network model targeted at scaffolding erection specifications; A model training module for training the object detection network model with training data; A data acquisition module for acquiring scaffolding erection video data; A specification recognition module for inputting the video data into the trained object detection network model to identify whether it conforms to the scaffolding erection specifications; if not, a risk warning is sent to the scaffolding mobile terminal.

9. A computer device, characterized in that it includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements the method for AI online recognition of scaffolding erection specifications according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that the computer-readable storage medium includes a stored computer program; wherein, when the computer program runs, it controls the device where the computer-readable storage medium is located to execute the method for AI online recognition of scaffolding erection specifications according to any one of claims 1 to 7.

Images