An aircraft state identification method, device, equipment and storage medium

Abstract

The application discloses an aircraft state recognition method, device and equipment and a storage medium. The method comprises the following steps: acquiring a to-be-recognized image collected by a target station; acquiring an image recognition scene matched with the to-be-recognized image according to the number of manufacturing areas included in the collection angle of view of the target station and the number of selectable aircraft states in each manufacturing area; and acquiring an image recognition result matched with the to-be-recognized image by using a target recognition model matched with the image recognition scene, wherein the recognition result comprises respective aircraft states in different manufacturing areas. The technical scheme of the embodiment of the application provides an aircraft in-position state recognition method applied to a 5G large aircraft park, and realizes the recognition of the in-position state of the aircraft to improve the accuracy of aircraft state recognition.

Description

Technical Field

[0001] This invention relates to the field of aircraft manufacturing technology, and in particular to an aircraft status identification method, device, equipment, and storage medium. Background Technology

[0002] Traditional machine vision methods utilize various algorithmic factors and customize parameters based on target features. They excel at solving quantitative problems primarily based on measurement and simple problems involving determining presence or absence, but for complex detection requirements, they suffer from high false negative or false positive rates.

[0003] In industrial inspection, especially when there are size benchmarks for application scenarios, a combination of image processing and deep learning can be used. Image processing is mainly responsible for rotation correction, image filtering enhancement, and target segmentation, while deep learning is mainly responsible for classification. This can solve the problem of complex scene detection that traditional vision cannot solve, or improve development efficiency by using deep learning annotation methods.

[0004] Existing cloud-based industrial vision deep learning tools implement the annotation, training, and testing process online. Relying on a large library of deep learning models for various scenarios accumulated in the cloud, users upload and annotate images, and the cloud automatically matches the optimal deep learning model and parameters for training, requiring no user intervention. However, existing machine vision methods are not suitable for the aircraft manufacturing field, and existing methods for recognizing aircraft manufacturing conditions cannot meet the high-precision requirements of aircraft condition recognition. Summary of the Invention

[0005] This invention provides an aircraft status identification method, apparatus, device, and storage medium to provide an aircraft on-site status identification method applied to 5G large aircraft parks, thereby improving the accuracy of aircraft status identification.

[0006] According to one aspect of the present invention, an aircraft state recognition method is provided, the method comprising:

[0007] Acquire the image to be identified from the target workstation; different workstations correspond to different acquisition perspectives;

[0008] Based on the number of manufacturing areas included in the acquisition view of the target workstation, and the number of selectable aircraft states in each manufacturing area, an image recognition scene matching the image to be recognized is obtained;

[0009] Using a target recognition model that matches the image recognition scene, obtain image recognition results that match the image to be recognized;

[0010] The identification results include the aircraft status corresponding to different manufacturing regions.

[0011] According to another aspect of the present invention, an aircraft status identification device is provided, the device comprising:

[0012] The image acquisition module is used to acquire the image to be recognized collected at the target workstation; different workstations correspond to different acquisition perspectives.

[0013] The image recognition scene acquisition module is used to acquire an image recognition scene that matches the image to be recognized based on the number of manufacturing areas included in the acquisition view of the target workstation and the number of selectable aircraft states in each manufacturing area.

[0014] The image recognition result acquisition module is used to acquire the image recognition result matching the image to be recognized by using a target recognition model that matches the image recognition scene;

[0015] The identification results include the aircraft status corresponding to different manufacturing regions.

[0016] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0017] At least one processor; and

[0018] A memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the aircraft state recognition method according to any embodiment of the present invention.

[0020] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the aircraft state recognition method according to any embodiment of the present invention.

[0021] The technical solution of this invention involves acquiring an image to be identified from a target workstation; wherein different workstations correspond to different acquisition perspectives; based on the number of manufacturing areas included in the acquisition perspective of the target workstation and the number of selectable aircraft states in each manufacturing area, an image recognition scene matching the image to be identified is obtained; a target recognition model matching the image recognition scene is used to obtain an image recognition result matching the image to be identified; wherein the recognition result includes the aircraft states corresponding to different manufacturing areas, providing a method for identifying the aircraft's on-site status in a 5G large aircraft park, thereby improving the accuracy of aircraft status recognition.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1a A flowchart of an aircraft state recognition method is provided in Embodiment 1 of the present invention;

[0025] Figure 1b This is a flowchart of an aircraft state recognition method under different image recognition scenarios provided in Embodiment 1 of the present invention;

[0026] Figure 2a This is a flowchart of another aircraft state recognition method provided in Embodiment 2 of the present invention;

[0027] Figure 2b This is a flowchart of an aircraft state reproduction method provided in Embodiment 2 of the present invention;

[0028] Figure 3 This is a schematic diagram of the structure of an aircraft status identification device provided in Embodiment 3 of the present invention;

[0029] Figure 4 This is a schematic diagram of the structure of an electronic device that implements the aircraft state recognition method of this invention. Detailed Implementation

[0030] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0031] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification, claims and accompanying drawings of this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such processes, methods, products or devices.

[0032] Example 1

[0033] Figure 1a This is a flowchart illustrating an aircraft state recognition method according to Embodiment 1 of the present invention. This embodiment is applicable to situations within a manufacturing area. The method can be executed by an aircraft state recognition device, which can be implemented in hardware and / or software. This device can be configured in a server with machine vision recognition capabilities. Figure 1a As shown, the method includes:

[0034] S110. Acquire the image to be identified from the target workstation.

[0035] In this embodiment, different workstations correspond to different acquisition angles. Different workstations can refer to different network cameras within the aircraft manufacturing workshop. The shooting area of ​​each network camera can be its acquisition angle, and each network camera can correspond to a fixed acquisition angle. The image to be identified can refer to the image corresponding to the acquisition angle of the target workstation.

[0036] In this embodiment, the address of the target network camera can be accessed to obtain the image to be identified captured by the target network camera at the current processing time.

[0037] S120. Based on the number of manufacturing areas included in the acquisition view of the target workstation and the number of selectable aircraft states in each manufacturing area, obtain an image recognition scene that matches the image to be recognized.

[0038] The manufacturing area can refer to the aircraft manufacturing area within a workshop. Optional aircraft states can refer to the stages of aircraft manufacturing, such as wings, fuselage, or wing-fuselage assembly. The image recognition scene can refer to the scene category in the image to be recognized.

[0039] In this embodiment, the recognition method may vary depending on the image recognition scene corresponding to the image to be recognized. Specifically, the image recognition scene matching the image to be recognized can be obtained based on the number of manufacturing areas included in the acquisition view of the target workstation and the number of selectable aircraft states in each manufacturing area.

[0040] Optionally, in one specific implementation of this embodiment, if it is determined that the number of manufacturing areas is unique and the number of selectable aircraft states is unique, the image recognition scenario is determined to be a single-target single-area scenario; if it is determined that the number of manufacturing areas is not unique and the number of selectable aircraft states is unique, the image recognition scenario is determined to be a single-target multi-area scenario; if it is determined that the number of manufacturing areas is unique and the number of selectable aircraft states is not unique, the image recognition scenario is determined to be a multi-target single-area scenario; if it is determined that the number of manufacturing areas is not unique and the number of selectable aircraft states is not unique, the image recognition scenario is determined to be a multi-target multi-area scenario.

[0041] S130. Using a target recognition model that matches the image recognition scene, obtain an image recognition result that matches the image to be recognized.

[0042] The identification results include the aircraft status corresponding to different manufacturing areas. The aircraft status output by the target identification model can be in the form of "0", "1", "2", and "3", etc. The target identification model can be a pre-trained identification model based on a machine vision recognition algorithm.

[0043] In this embodiment, a target recognition model matching the image recognition scene can be determined based on the image recognition scene, and then the target recognition model can be used to recognize the image to be recognized to obtain an image recognition result matching the image to be recognized.

[0044] Optionally, a target recognition model matching a single target, single region scene can be used to detect whether a single region in the image to be recognized contains an aircraft, and to obtain a recognition result indicating whether the single region contains or does not contain an aircraft; a target recognition model matching a single target, multi-region scene can be used to detect whether each manufacturing region in the image to be recognized contains an aircraft, and to obtain a recognition result indicating whether each manufacturing region contains or does not contain an aircraft; a target recognition model matching a multi-target, single-region scene can be used to detect whether a single region in the image to be recognized contains an aircraft, and when the single region contains an aircraft, to obtain a recognition result indicating the type of aircraft state in the single region; a target recognition model matching a multi-target, multi-region scene can be used to detect whether each manufacturing region in the image to be recognized contains an aircraft, and to obtain a recognition result indicating the type of aircraft state in the manufacturing region containing an aircraft.

[0045] For example, Figure 1bThis is a flowchart illustrating an aircraft state recognition method under different image recognition scenarios according to Embodiment 1 of the present invention. The method involves acquiring an image to be recognized from a webcam stream and then detecting and recognizing the image according to different image recognition scenarios. For a single-target, single-region scenario, the presence or absence of an aircraft within the single region is determined. If an aircraft is present, the output is "1"; otherwise, it is "0". In this scenario, the aircraft state within the manufacturing region can be a complete aircraft. For a single-target, multi-region scenario, the manufacturing region can be divided into several regions, and each region can be determined to contain an aircraft. The corresponding aircraft state for each region is then output, for example, "1" if an aircraft is present, and "0" if no aircraft is present. For a multi-target, single-region scenario, the presence or absence of an aircraft within a single region is determined. If no aircraft is present, the output is "0". If an aircraft is present within a single region, features of the aircraft within the manufacturing region are extracted to identify the aircraft state category, such as "State 1" and "State 2". For multi-target, multi-region scenarios, manufacturing areas can be divided, and it can be determined whether each manufacturing area contains an aircraft. For manufacturing areas without aircraft, the corresponding "0" is output. For manufacturing areas with aircraft, feature extraction is performed to identify the aircraft's state category, such as "State 3" and "State 4". It should be noted that "State 1", "State 2", "State 3" and "State 4" here are only examples and not limitations.

[0046] The technical solution of this invention involves acquiring an image to be identified from a target workstation; wherein different workstations correspond to different acquisition perspectives; based on the number of manufacturing areas included in the acquisition perspective of the target workstation and the number of selectable aircraft states in each manufacturing area, an image recognition scene matching the image to be identified is obtained; a target recognition model matching the image recognition scene is used to obtain an image recognition result matching the image to be identified; wherein the recognition result includes the aircraft states corresponding to different manufacturing areas, providing a method for identifying the aircraft's on-site status in a 5G large aircraft park, thereby improving the accuracy of aircraft status recognition.

[0047] Based on the above technical solution, after obtaining the image recognition result matching the image to be recognized, the method may further include: obtaining the true location box of the aircraft in the manufacturing area, and obtaining at least one predicted location box of the aircraft in the manufacturing area; determining the confidence level of each predicted location box based on the distance between the at least one predicted location box and the true location box; determining the predicted location box with the highest confidence level as the target box matching the aircraft in the manufacturing area, and adding the target box to the image recognition result.

[0048] For example, in a single-target, single-region scenario, the bounding box coordinates (x1, y1, x2, y2) of each aircraft can be labeled, with each coordinate corresponding to a label; based on the obtained predicted location box, the confidence score is calculated, regression analysis is performed, and the location of the aircraft's bounding box is obtained.

[0049] For single-target, multi-region scenarios, a (c, x, y, w, h) coordinate can be used to detect an aircraft in a specific region. Assuming the image to be identified comprises ten regions, each region is represented by one (c, x, y, w, h) coordinate. The target recognition model can then output ten bounding boxes at once, each represented by one (c, x, y, w, h) coordinate. Since convolution operations are strongly position-dependent, the location of the original aircraft remains on the feature map after convolution. Therefore, the image is divided into ten regions, and the results are distributed across these ten regions.

[0050] For multi-target single-region scenes, the regression coordinates and one-hot classification are N*(c, x, y, w, h, one-hot). The output is represented by a one-hot vector, for example, [0, 0, 1, 0, 0, 0], where any dimension that is 1 represents the class to which the image to be identified belongs. Non-maximum suppression (NMS) is used to filter bounding boxes, using overlap to suppress overlapping boxes. Since each target only appears once, other boxes with the same label are also removed, and the predicted location box with the highest confidence is found as the target box.

[0051] For multi-target, multi-region scenarios, add grid-based offsets and anchor-based offsets. The anchor-based offset = target position - anchor position, and the grid-based offset = target position - grid position. Change the predicted values ​​from x, y, w, h to t. x , t y , t w , t h , where t x =log((bbox) x -c x ) / (1-(bbox x -c x ))), t y =log((bbox) y -c y ) / (1-(bbox y -c y ))), t w =log(gt w / p w ), t h=log(gt h / p h ), where t x , t y , t w , t h c is the predicted value. x and c y p represents the coordinates of the top-left corner of the grid. w and p h The width and height are the anchor's dimensions. The anchor is a predefined frame with fixed width and height.

[0052] In multi-target, multi-region scenarios, the offset can be predicted, and the final target position can be obtained based on the offset of the width and height of the anchor box and the prior position of the grid. Since aircraft size and status vary at different workstations in aircraft state recognition, we pre-prepare several bounding boxes with high probability of occurrence from the dataset and use them as a benchmark for prediction. The image is divided into 13*13 regions, each with 5 anchors, and each anchor corresponds to a category. Therefore, the output size should be [N, 13, 13, 125].

[0053] The detection head is divided into three parts: 13*13*3*(4+1+30), 26*26*3*(4+1+30), and 52*52*3*(4+1+30). That is, each grid has 9 prior boxes: 3 large, 3 medium, and 3 small. Because 32x downsampling results in a larger receptive field per point, it predicts large targets; 8x downsampling results in a smaller receptive field per point, so it predicts small targets. Each branch predicts 3 boxes, and each box predicts a 5-tuple + 30 one-hot vector classes, so the total size is 3*(4+1+30). A total of (13*13+26*26+52*52)*3=10467 can be predicted >> (13*13*5)=845, thus improving the model's capability.

[0054] Example 2

[0055] Figure 2a This is a flowchart of another aircraft state recognition method provided in Embodiment 2 of the present invention. Based on the above embodiments, this embodiment adds an operation after obtaining the image recognition result matching the image to be recognized. For example... Figure 2a As shown, the method includes:

[0056] S210. Acquire the image to be identified from the target workstation; where different workstations correspond to different acquisition perspectives.

[0057] S220. Based on the number of manufacturing areas included in the acquisition view of the target workstation and the number of selectable aircraft states in each manufacturing area, obtain an image recognition scene that matches the image to be recognized.

[0058] S230. Using a target recognition model that matches the image recognition scene, obtain an image recognition result that matches the image to be recognized; wherein, the recognition result includes the aircraft status corresponding to different manufacturing areas.

[0059] S240. Save the image recognition result to the target file and upload it to the digital twin platform to realize the reproduction of the target workstation aircraft status.

[0060] The target file can typically be a JSON file.

[0061] S250. According to a preset time interval, periodically obtain the aircraft status corresponding to the target workstation and update the recognition results in the digital twin platform.

[0062] For example, Figure 2b This is a flowchart of an aircraft state reproduction method provided in Embodiment 2 of the present invention. The method requests a server address to obtain video frames and frame numbers for the current n workstations, saves the images to be recognized to a test folder, calls a prediction function, executes a trained model, and performs aircraft state recognition on the images. When an aircraft is recognized, the method outputs the recognized aircraft workstation and the numbers 0, 1, 2, 3, etc., representing different aircraft states. The method is set to output only the target bounding boxes with the highest confidence of the same type of target each time to avoid the problem of multiple bounding boxes for a single target. The output image recognition results are saved in a JSON file and displayed on a webpage. The JSON file is also saved to a database and uploaded to a digital twin platform to achieve aircraft state reproduction. The server is set to start automatically upon system boot, refresh the webpage every ten minutes to obtain workstation information, and update it to the digital twin platform in real time to ensure consistency between the on-site state and the digital twin platform. Data information is stored in the database, with one record saved every ten minutes.

[0063] The technical solution of this invention provides a method for recognizing aircraft status in a 5G large aircraft park by acquiring an image to be recognized from a target workstation, wherein different workstations correspond to different acquisition perspectives; based on the number of manufacturing areas included in the acquisition perspective of the target workstation and the number of selectable aircraft states in each manufacturing area, an image recognition scene matching the image to be recognized is obtained; a target recognition model matching the image recognition scene is used to obtain an image recognition result matching the image to be recognized; wherein the recognition result includes the aircraft states corresponding to different manufacturing areas, the image recognition result is saved to a target file and uploaded to a digital twin platform to reproduce the aircraft state of the target workstation; according to a preset time interval, the aircraft state corresponding to the target workstation is acquired periodically, and the recognition result in the digital twin platform is updated, thereby improving the accuracy of aircraft status recognition and increasing aircraft manufacturing efficiency.

[0064] Optionally, due to potential occlusion issues in aircraft manufacturing workshops, oversampling can be performed using the SMOTE algorithm. This allows for analysis and simulation of minority class samples, and the simulated new samples are added to the dataset to address the imbalanced sample problem. For example, in cases of occlusion, for each region, two quintuples (c, x, y, w, h) can be used: one for regressing large targets and the other for regressing small targets. Similarly, a one-hot vector is added, and four images are randomly used, randomly scaled to add many small targets, and then randomly distributed and stitched together to enrich the detection dataset.

[0065] Example 3

[0066] Figure 3 This is a schematic diagram of the structure of an aircraft status identification device provided in Embodiment 3 of the present invention. Figure 3 As shown, the device includes: an image acquisition module 310, an image recognition scene acquisition module 320, and an image recognition result acquisition module 330. Wherein:

[0067] The image acquisition module 310 is used to acquire the image to be recognized collected at the target workstation; wherein, different workstations correspond to different acquisition perspectives;

[0068] The image recognition scene acquisition module 320 is used to acquire an image recognition scene that matches the image to be recognized based on the number of manufacturing areas included in the acquisition view of the target workstation and the number of selectable aircraft states in each manufacturing area.

[0069] The image recognition result acquisition module 330 is used to acquire the image recognition result matching the image to be recognized by using a target recognition model that matches the image recognition scene;

[0070] The identification results include the aircraft status corresponding to different manufacturing regions.

[0071] The technical solution of this invention involves acquiring an image to be identified from a target workstation; wherein different workstations correspond to different acquisition perspectives; based on the number of manufacturing areas included in the acquisition perspective of the target workstation and the number of selectable aircraft states in each manufacturing area, an image recognition scene matching the image to be identified is obtained; a target recognition model matching the image recognition scene is used to obtain an image recognition result matching the image to be identified; wherein the recognition result includes the aircraft states corresponding to different manufacturing areas, providing a method for identifying the aircraft's on-site status in a 5G large aircraft park, thereby improving the accuracy of aircraft status recognition.

[0072] Optionally, the image recognition scene acquisition module 320 can be used for:

[0073] If the number of manufacturing areas is determined to be unique, and the number of selectable aircraft states is also unique, then the image recognition scenario is determined to be a single-target, single-area scenario.

[0074] Correspondingly, the image recognition result acquisition module 330 can be specifically used to: detect whether a single region in the image to be recognized contains an aircraft through the target recognition model, and obtain the recognition result of whether the single region contains or does not contain an aircraft.

[0075] Optionally, the image recognition scene acquisition module 320 can also be used for:

[0076] If it is determined that the number of manufacturing areas is not unique, but the number of selectable aircraft states is unique, then the image recognition scenario is determined to be a single-target multi-area scenario.

[0077] Correspondingly, the image recognition result acquisition module 330 can also be used for:

[0078] The target recognition model detects whether each manufacturing area in the image to be recognized contains an aircraft, and obtains the recognition result of whether each manufacturing area contains or does not contain an aircraft.

[0079] Optionally, the image recognition scene acquisition module 320 can also be used for:

[0080] If the number of manufacturing areas is determined to be unique, and the number of selectable aircraft states is not unique, then the image recognition scenario is determined to be a multi-target single-area scenario.

[0081] Correspondingly, the image recognition result acquisition module 330 can also be used for:

[0082] The target recognition model detects whether a single region in the image to be recognized contains an aircraft, and when an aircraft is contained in the single region, the type of aircraft status in the single region is identified.

[0083] Optionally, the image recognition scene acquisition module 320 can also be used for:

[0084] If it is determined that the number of manufacturing areas is not unique and the number of selectable aircraft states is not unique, then the image recognition scenario is determined to be a multi-target, multi-region scenario.

[0085] Correspondingly, the image recognition result acquisition module 330 can also be used for:

[0086] The target recognition model detects whether each manufacturing area in the image to be recognized contains an aircraft, and obtains the type recognition result of the aircraft status in the manufacturing area containing the aircraft.

[0087] Optionally, the aircraft status recognition device further includes a target bounding box acquisition and addition module, used after acquiring the image recognition result matching the image to be recognized:

[0088] Obtain the true location box of the aircraft in the manufacturing area, and obtain at least one predicted location box of the aircraft in the manufacturing area;

[0089] The confidence level of each predicted location box is determined based on the distance between the at least one predicted location box and the actual location box.

[0090] The predicted location box with the highest confidence is determined as the target box for matching the aircraft within the manufacturing area, and the target box is added to the image recognition result.

[0091] Optionally, the aircraft status recognition device further includes a recognition result update module, used to update the image after obtaining an image recognition result matching the image to be recognized:

[0092] The image recognition results are saved to the target folder and uploaded to the digital twin platform to reproduce the state of the aircraft at the target workstation;

[0093] According to a preset time interval, the aircraft status corresponding to the target workstation is acquired periodically, and the recognition results in the digital twin platform are updated.

[0094] The aircraft status identification device provided in the embodiments of the present invention can execute the aircraft status identification method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method execution.

[0095] Example 4

[0096] Figure 4 A schematic diagram of an electronic device 400 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0097] like Figure 4 As shown, the electronic device 400 includes at least one processor 401 and a memory, such as a read-only memory (ROM) 402 and a random access memory (RAM) 403, communicatively connected to the at least one processor 401. The memory stores computer programs executable by the at least one processor. The processor 401 can perform various appropriate actions and processes based on the computer program stored in the ROM 402 or loaded into the RAM 403 from storage unit 408. The RAM 403 may also store various programs and data required for the operation of the electronic device 400. The processor 401, ROM 402, and RAM 403 are interconnected via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0098] Multiple components in electronic device 400 are connected to I / O interface 405, including: input unit 406, such as keyboard, mouse, etc.; output unit 407, such as various types of displays, speakers, etc.; storage unit 408, such as disk, optical disk, etc.; and communication unit 409, such as network card, modem, wireless transceiver, etc. Communication unit 409 allows electronic device 400 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0099] Processor 401 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 401 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 401 performs the various methods and processes described above, such as aircraft state recognition methods.

[0100] In some embodiments, the aircraft state identification method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 408. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 400 via ROM 402 and / or communication unit 409. When the computer program is loaded into RAM 403 and executed by processor 401, one or more steps of the aircraft state identification method described above may be performed. Alternatively, in other embodiments, processor 401 may be configured to perform the aircraft state identification method by any other suitable means (e.g., by means of firmware).

[0101] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0102] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0103] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0104] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0105] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0106] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0107] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0108] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. An aircraft state recognition method, characterized in that, The method comprises: acquiring a to-be-recognized image collected by a target station; different stations correspond to different collection angles; acquiring an image recognition scene matched with the to-be-recognized image according to the number of manufacturing areas included in the collection angle of the target station and the number of optional aircraft states in each manufacturing area; the optional aircraft state refers to the manufacturing stage state of the aircraft; acquiring an image recognition result matched with the to-be-recognized image by using a target recognition model matched with the image recognition scene; the image recognition result includes the aircraft states in different manufacturing areas respectively; acquiring an image recognition scene matched with the to-be-recognized image according to the number of manufacturing areas included in the collection angle of the target station and the number of optional aircraft states in each manufacturing area comprises: if it is determined that the number of manufacturing areas is unique and the number of optional aircraft states is unique, determining that the image recognition scene is a single-target single-area scene; if it is determined that the number of manufacturing areas is not unique and the number of optional aircraft states is unique, determining that the image recognition scene is a single-target multi-area scene; if it is determined that the number of manufacturing areas is unique and the number of optional aircraft states is not unique, determining that the image recognition scene is a multi-target single-area scene; if it is determined that the number of manufacturing areas is not unique and the number of optional aircraft states is not unique, determining that the image recognition scene is a multi-target multi-area scene; the aircraft states in the image recognition result represent the presence or absence of the aircraft in the single-target single-area scene and the single-target multi-area scene; the aircraft states in the image recognition result represent the presence or absence of the aircraft and the manufacturing stage state of the aircraft when the aircraft is present in the multi-target single-area scene and the multi-target multi-area scene.

2. The method of claim 1, wherein, in the case that the image recognition scene is the single-target single-area scene, acquiring an image recognition result matched with the to-be-recognized image by using a target recognition model matched with the image recognition scene comprises: detecting whether the single area in the to-be-recognized image contains an aircraft by using the target recognition model and acquiring the recognition result of whether the single area contains an aircraft.

3. The method of claim 1, wherein, in the case that the image recognition scene is the single-target multi-area scene, acquiring an image recognition result matched with the to-be-recognized image by using a target recognition model matched with the image recognition scene comprises: detecting whether each manufacturing area in the to-be-recognized image contains an aircraft by using the target recognition model and acquiring the recognition result of whether each manufacturing area contains an aircraft.

4. The method of claim 1, wherein, in the case that the image recognition scene is the multi-target single-area scene, acquiring an image recognition result matched with the to-be-recognized image by using a target recognition model matched with the image recognition scene comprises: detecting whether the single area in the to-be-recognized image contains an aircraft by using the target recognition model and acquiring the type recognition result of the aircraft state in the single area when the single area contains an aircraft.

5. The method of claim 1, wherein, In a case where the image recognition scenario is the multi-target multi-region scenario, a target recognition model matched with the image recognition scenario is adopted to obtain an image recognition result matched with the to-be-identified image, including: Whether the to-be-identified image contains an airplane in each manufacturing region is detected through the target recognition model, and a kind recognition result of an airplane state in a manufacturing region containing an airplane is obtained.

6. The method of claim 1, wherein, After the image recognition result matched with the to-be-identified image is obtained, further including: A real position box of the airplane in the manufacturing region is obtained, and at least one predicted position box of the airplane in the manufacturing region is obtained; According to distances between the at least one predicted position box and the real position box, a confidence of each predicted position box is determined; The predicted position box with the maximum confidence is determined as a target box matched with the airplane in the manufacturing region, and the target box is added to the image recognition result.

7. The method of claim 1, wherein, After the image recognition result matched with the to-be-identified image is obtained, further including: The image recognition result is saved to a target file and uploaded to a digital twin platform to realize reproduction of an airplane state of a target work station; According to a preset time interval, an airplane state corresponding to the target work station is obtained in a timely manner, and an identification result in the digital twin platform is updated.

8. An aircraft state recognition device, characterized by including: A to-be-identified image acquisition module is configured to acquire a to-be-identified image collected by a target work station; different work stations correspond to different collection angles; An image recognition scenario acquisition module is configured to acquire an image recognition scenario matched with the to-be-identified image according to a number of manufacturing regions included in the collection angle of the target work station and a number of selectable airplane states in each manufacturing region; the selectable airplane state refers to a manufacturing stage state of an airplane; An image recognition result acquisition module is configured to acquire an image recognition result matched with the to-be-identified image by using a target recognition model matched with the image recognition scenario; the image recognition result includes airplane states corresponding to different manufacturing regions respectively; The image recognition scenario acquisition module is specifically configured to: if it is determined that the number of manufacturing regions is unique and the number of selectable airplane states is unique, determine that the image recognition scenario is a single-target single-region scenario; if it is determined that the number of manufacturing regions is not unique and the number of selectable airplane states is unique, determine that the image recognition scenario is a single-target multi-region scenario; if it is determined that the number of manufacturing regions is unique and the number of selectable airplane states is not unique, determine that the image recognition scenario is a multi-target single-region scenario; and if it is determined that the number of manufacturing regions is not unique and the number of selectable airplane states is not unique, determine that the image recognition scenario is a multi-target multi-region scenario. The airplane states in the image recognition result represent whether there is an airplane in a case of the single-target single-region scenario and the single-target multi-region scenario; the airplane states in the image recognition result represent whether there is an airplane and a manufacturing stage state of the airplane when there is an airplane in a case of the multi-target single-region scenario and the multi-target multi-region scenario.

9. An electronic device, comprising: The electronic device includes: at least one processor; and A memory connected in communication with the at least one processor; wherein The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to perform the aircraft state identification method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for causing a processor to implement the aircraft state identification method according to any one of claims 1-7 when executed by the processor.

Images