Aircraft fuselage defect in-situ automatic detection system and method

Abstract

The invention discloses an aircraft fuselage defect in-situ automatic detection system and method, and belongs to the field of aircraft overhaul and maintenance, the system comprises a visual system module, an ultrasonic system module and a data processing and storage module; the visual system module is used for performing marker detection and ultrasonic probe position automatic detection, realizing alignment of an airplane digital-analog coordinate system and a camera coordinate system, and calculating the detailed position of the probe position on a target airplane; the ultrasonic system module adopts phased array C scanning ultrasonic equipment and is used for detecting internal defects in a fuselage skin-grating structure to obtain an ultrasonic detection result; the data processing and storage module is a computing device with a computing function and a certain amount of storage medium and is used for being responsible for achieving the target detection and computing functions of the visual system and the ultrasonic system. According to the invention, the position coordinates of the detection data can be automatically obtained and stored in a standardized manner, the detection area can be visualized, and the labor investment can be greatly reduced.

Description

Technical Field

[0001] The present invention relates to the field of aircraft maintenance and repair, and specifically to an in-situ automatic detection system and method for aircraft fuselage defects. Background Art

[0002] The lightweight and high-strength composite load-bearing skin is one of the typical features of advanced civil and military aircraft. However, the layers and structures of the composite skin are prone to damage such as skin delamination and debonding between the skin and stringers under the action of impact loads, fatigue cycles, and over-threshold loads. At present, it is necessary to conduct a comprehensive non-destructive inspection of the fuselage skin every dozens to hundreds of flight hours to detect possible skin damage.

[0003] Currently, the commonly used flaw detection method is to detect and record aircraft defects by means of manual detection + manual judgment + manual marking + manual recording. The operator pushes and sweeps the probe to detect and evaluate the detection results. After detecting a defect, the shape and area of the defect are marked, and then information such as the position, depth, and area of the defect is recorded in a record book by hand.

[0004] However, the manual recording method has certain drawbacks. First of all, the manual recording method is time-consuming and laborious and cannot conform to the trend of promoting automatic detection. Secondly, it is inevitable that there will be problems such as missed areas, repeated areas, and inaccurate defect position judgment in manual detection. In addition, there are problems with non-unique references and poor standardization in the information such as the defect position, defect area, and defect depth recorded manually. Finally, the paper-based defect record data needs to be processed twice to be saved as digital data, which is not convenient for statistical analysis of defects and detection situations.

[0005] Therefore, those skilled in the art have provided an in-situ automatic detection system and method for aircraft fuselage defects to solve the problems raised in the above background art. Summary of the Invention

[0006] The purpose of the present invention is to provide an in-situ automatic detection system and method for aircraft fuselage defects, which can automatically obtain the position coordinates of detection data and store them in a standardized manner, and realize the function of visualizing the detection area, and can greatly reduce manual input to solve the problems raised in the above background art.

[0007] To achieve the above purpose, the present invention provides the following technical solutions:

[0008] An in-situ automatic detection system for aircraft fuselage defects includes a vision system module, an ultrasonic system module, and a data processing and storage module;

[0009] The visual system module is used for marker detection and automatic detection of the ultrasonic probe position, to align the aircraft digital mock-up coordinate system and the camera coordinate system, and calculate the detailed position of the probe position on the target aircraft.

[0010] The ultrasonic system module adopts a phased array C-scan ultrasonic device to detect internal defects in the fuselage skin-grid structure and obtain ultrasonic detection results.

[0011] The data processing and storage module is a computing device with computing functions and a certain amount of storage media, responsible for implementing the target detection and computing functions of the visual system and the ultrasonic system, and providing a database to store all damage entries.

[0012] As a further solution of the present invention: The visual system module includes one or more networked cameras and several coded markers attached to specific positions on the fuselage. Among them, the carrier platform of the networked cameras includes a drone platform and an external fixed bracket, and each networked camera needs to be calibrated to determine the internal parameter matrix.

[0013] As a further solution of the present invention: The ultrasonic system module includes one or more networked ultrasonic non-destructive testing devices. The networked ultrasonic non-destructive testing devices are used to detect damage points such as delamination and debonding of the aircraft skin and grid structure, and the probes of the networked ultrasonic non-destructive testing devices include one of a phased array ultrasonic system and an A-scan probe.

[0014] As a further solution of the present invention: The data processing and storage module includes a networked host server and a display. Among them, the networked host server is used to provide storage and computing services, and the display is used to present the results of the detection progress and defect statistics.

[0015] As a further solution of the present invention: The data processing and storage module needs to store a high-fidelity three-dimensional 1:1 digital model of the aircraft under repair. This model stores the curved surface key points and grid information on the aircraft surface, and all key points have unique coordinates in the model coordinate system.

[0016] As a further solution of the present invention: The specific steps for the visual system module to align the aircraft digital mock-up coordinate system and the camera coordinate system are as follows:

[0017] S101: Affix coded markers at preset positions on the fuselage surface. Their positions and distributions meet two requirements: The first requirement is that each networked camera can clearly capture at least three coded markers simultaneously; The second requirement is that the coded markers need to accurately correspond to the key points with known coordinates on one of the surfaces of the digital model.

[0018] S102: For the camera s of interest, the projection transformation corresponding to its known internal parameter is denoted as proj C→S , and at one of its poses, a photo is taken;

[0019] S103: Run the marker detection program to ensure that n visual positioning marker points can be clearly detected within its field of view, and their coordinates are (X M,i , Y M,i , Z M,i ), where 1 ≤ i ≤ n and n ≥ 3, and their imaging pixel position coordinates on the camera s are (X s,i , Y s,i );

[0020] S104: Run the spatial transformation matrix solving program, input the camera internal parameters and point pairs, and solve the orthogonal transformation matrix R and translation vector T from the camera s coordinate system to the model coordinate system. Their parameters form the affine transformation parameters, denoted as aff M→S .

[0021] As a further solution of the present invention: The specific steps for the vision system module to obtain the position of the defect are as follows:

[0022] S201: According to the internal and external parameters of the camera imaging, solve the curved surface imaging positions of each key point {x1, x2,..., x n} of the digital model under the imaging system, according to

[0023] {x 1p , x 2p ,..., x np} = proj C→S (aff M→C ({x1, x2,..., x n})); (1)

[0024] S202: When using the probe for detection, ensure that the probe appears clearly in at least one camera s, and detect the pixel position x s = (X s,i , Y s,i ) of the probe through the trained object detection neural network algorithm;

[0025] S203: According to the sorting method and the Euclidean distance index, find the three imaging points [x 1p , x 2p ,..., x np} that are closest to x s in {x 1p , x 2p , x 3p ;

[0026] S204: Solve the centroid method parameters:

[0027]

[0028] S205: Calculate the real coordinates of the probe in the model coordinate system:

[0029]

[0030] As a further solution of the present invention: the specific process of presenting the detection progress result by the digital model is:

[0031] S301: Execute steps S101 to S104 and S201 to S205 frame by frame to obtain a sequence of running trajectory points of the probe on the surface of the aircraft, and connect them into a curve;

[0032] S302: Calculate the probe scanning area according to the actual width of the probe;

[0033] S303: Using a blank digital model, fill the area scanned by all probes in a single test with a certain color, or map it with a test result image of the ultrasound system module; the area not scanned by the probe is not filled with color.

[0034] As a further solution of the present invention: the specific steps of adding non-destructive testing data entries in the data processing and storage module are:

[0035] S401: Get the C-scan image from the ultrasonic system module, run the defect target automatic detection program, if a defect is detected, get the defect location, area, depth, time range, and run S402 to S404, otherwise wait for the next ultrasonic system input result;

[0036] S402: For one push-sweep trajectory of the probe, the trajectory of the probe during the push-sweep process and its coverage area in the model coordinate system are obtained through steps S101 to S104 and S201 to S205;

[0037] S403: aligning the time range outputted in S401 with the position of the probe in the time range in S402, and obtaining the model coordinate system position corresponding to the probe when the defect is detected;

[0038] S404: The position of the defect in the model coordinate system obtained in S403 and the defect depth, area, and shape information obtained in S401 are aggregated to form a defect detection record, which is stored in the database.

[0039] The present application also discloses an in-situ automated detection method for aircraft fuselage defects, which uses an in-situ automated detection system for aircraft fuselage defects and includes the following steps:

[0040] The marker detection and the automatic detection of the ultrasonic probe position are carried out through the vision system module to align the aircraft digital model coordinate system and the camera coordinate system, and calculate the detailed position of the probe position on the target aircraft;

[0041] The detection of internal defects in the fuselage skin-grille structure is realized through the ultrasonic system module, and the ultrasonic detection results are obtained;

[0042] The data processing and storage module realizes the target detection and calculation functions of the vision system and the ultrasonic system, and provides a database responsible for storing all damage entries.

[0043] Compared with the prior art, the beneficial effects of the present invention are:

[0044] 1. This application can realize the functions of automatically obtaining the position coordinates of detection data and storing them in a standardized manner, as well as visualizing the detection area. The overall automation degree is high, which can greatly reduce the manual input.

[0045] 2. Through the in-situ automatic detection of fuselage defects, this application can effectively avoid problems such as area omission, area duplication, and inaccurate defect position interpretation, improving the detection accuracy and efficiency.

[0046] 3. Compared with the traditional manual detection, the in-situ automatic detection of fuselage defects in this application effectively avoids the problems of non-unique reference objects and poor standardization.

[0047] 4. The defect record data of this application can be directly saved without the need for secondary processing to be saved as digital data, which is convenient for statistical analysis of defects and detection situations. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a structural block diagram of an in-situ automatic detection system for aircraft fuselage defects;

[0049] Figure 2 is a detection flow chart of an in-situ automatic detection system for aircraft fuselage defects.

[0050] In the figure: 110, vision system module; 111, network camera; 112, unmanned aerial vehicle platform; 113, first data transmission module; 114, coded marker; 120, ultrasonic system module; 121, probe; 122, damage point; 123, second data transmission module; 130, data processing and storage module; 131, network host server; 132, digital model; 133, non-destructive testing data entry; 134, model coordinate system; 140, target aircraft. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0052] As mentioned in the background art of the present application, through research, it has been found that the existing flaw detection methods usually record the detection and recording of aircraft defects by means of manual detection + manual discrimination + manual marking + manual recording. However, the manual recording method has certain drawbacks. First of all, the manual recording method is time-consuming and laborious and cannot conform to the trend of promoting automated detection. Secondly, it is inevitable that problems such as regional omission, regional duplication, and inaccurate defect position interpretation occur in manual detection. In addition, there are problems with non-unique references and poor standardization in the information such as defect positions, defect areas, and defect depths recorded manually. Finally, the paper-based defect record data needs to be processed twice to be saved as digital data, which is not convenient for statistical analysis of defects and detection situations.

[0053] To solve the above defects, the present application discloses an in-situ automated detection system and method for aircraft fuselage defects, which can realize the automated acquisition of position coordinates of detection data and its standardized storage, as well as the visualization of the detection area, and can greatly reduce manual input.

[0054] The following will introduce in detail how the solution of the present application solves the above technical problems in conjunction with the accompanying drawings.

[0055] Please refer to Figure 1 , in the embodiments of the present invention, an in-situ automated detection system for aircraft fuselage defects includes a vision system module 110, an ultrasonic system module 120, and a data processing and storage module 130; the vision system module 110 is used for marker detection and automatic detection of the position of the ultrasonic probe, to align the aircraft digital model coordinate system and the camera coordinate system, and calculate the detailed position of the probe 121 on the target aircraft 140; the ultrasonic system module 120 uses a phased array C-scan ultrasonic device to detect internal defects in the fuselage skin-grid structure and obtain ultrasonic detection results; the data processing and storage module 130 is a computing device with computing functions and a certain amount of storage media, which is used to be responsible for realizing the target detection and computing functions of the vision system and the ultrasonic system, and providing a database to store all damage entries. The present application can realize the automated acquisition of position coordinates of detection data and its standardized storage, as well as the visualization of the detection area, with a high overall degree of automation, and can greatly reduce manual input.

[0056] In this embodiment, the vision system module 110 includes one or more networked cameras 111 and several coded markers 114 attached to specific positions on the fuselage. Among them, the carrier platform of the networked cameras 111 includes a drone platform 112 and an external fixed bracket. Each networked camera 111 needs to be calibrated to determine the internal parameter (projection parameter) matrix. The internal parameter matrix is a parameter matrix that describes the internal attributes of the camera. It establishes the transformation relationship from three-dimensional camera coordinates to two-dimensional homogeneous image coordinates. These parameters are only determined by the physical attributes of the camera itself, such as focal length, principal point coordinates (optical center), etc., and are independent of the external environment. Therefore, once the camera is manufactured and calibrated, its internal parameter matrix usually remains unchanged during the use of the camera. The vision system module 110 sends the detection data to the data processing and storage module 130 through the first data transmission module 113.

[0057] In this embodiment, the ultrasonic system module 120 includes one or more networked ultrasonic non-destructive testing devices. The networked ultrasonic non-destructive testing devices are used to detect damage points 122 such as delamination and debonding of the aircraft skin and grille structure. Moreover, the probe 121 of the networked ultrasonic non-destructive testing device includes one of a phased array ultrasonic system (equipped with a wheeled or flat-pushing probe) and an A-scan probe. The phased array ultrasonic system is an advanced ultrasonic imaging and detection technology. It uses phased array technology to control the phase and amplitude of multiple transmitting and receiving elements (usually piezoelectric wafers) to achieve beam orientation, focusing, and scanning, so as to obtain the internal structure information of the object to be detected and display it in the form of an image. The A-scan probe is a basic ultrasonic detection tool. It is usually placed on the surface of the object to be inspected, emits ultrasonic waves in a fixed direction in a fixed-point manner, and receives the echo signals along the axis of the sound beam. On the display screen, the vertical coordinate represents the amplitude (amplitude) of the echo signal, and the horizontal coordinate represents the information of the depth where the echo target is located. The ultrasonic system module 120 sends the detection data to the data processing and storage module 130 through the second data transmission module 123.

[0058] In this embodiment, the data processing and storage module 130 includes a networked host server 131 and a display. Among them, the networked host server 131 is used to provide storage and computing services, and the display is used to present the results of the detection progress and defect statistics.

[0059] In this embodiment, the data processing and storage module 130 needs to store a high-fidelity three-dimensional 1:1 digital model 132 of the aircraft under maintenance. This model stores the curved surface key points and grid information on the aircraft surface, and all key points have unique coordinates in the model coordinate system 134. The curved surface key points and grid information on the aircraft surface play a crucial role in the aircraft design, manufacturing, analysis, and maintenance processes. Among them, the curved surface key points on the aircraft surface usually refer to those points that have important impacts on the aircraft's aerodynamic performance, structural strength, and appearance. The grid information on the aircraft surface refers to the information required for dividing the aircraft surface into a series of small units (such as triangles, quadrilaterals, etc.) for numerical analysis (such as computational fluid dynamics analysis, finite element analysis, etc.).

[0060] In this embodiment, the specific steps for the vision system module 110 to align the aircraft digital model coordinate system and the camera coordinate system are as follows: S101: Affix coded markers 114 at preset positions on the fuselage surface, and their positions and distributions meet two requirements: The first requirement is that each networked camera 111 can clearly capture at least three coded markers 114 simultaneously; The second requirement is that the coded markers 114 need to accurately correspond to the key points with known coordinates on one of the surfaces of the digital model 132; S102: For the concerned camera s, the projection transformation corresponding to its known internal parameter is represented as proj C→S , at one of the poses, take a photo; S103: Run the marker detection program to ensure that n visual positioning marker points can be clearly detected within its field of view, and their coordinates are (X M,i , Y M,i , Z M,i ), 1 ≤ i ≤ n, n ≥ 3, and their imaging pixel position coordinates on this camera s are (X s,i , Y s,i ); S104: Run the spatial transformation matrix solving program, input the camera internal parameters and point pairs, and solve the orthogonal transformation matrix R from the camera s coordinate system to the model coordinate system 134 and the translation vector T (camera external parameters), and their parameters form affine transformation parameters, denoted as aff M→S . In the process of the vision system module 110 aligning the aircraft digital model coordinate system and the camera coordinate system, the marker detection program is a crucial link. The main function of this program is to accurately identify and locate the preset coded markers 114 in the taken photo, and then obtain the imaging pixel position coordinates of these markers on the photo. These coordinate information are the basis for subsequent calculation of the camera external parameters and realization of coordinate system alignment. Affine transformation is a method for geometrically transforming a graph, including translation, scaling, rotation, and skew, etc., which can perform various linear transformations and displacements on the graph while maintaining the "flatness" and "parallelism" of the graph. The parameters of affine transformation are usually used to define the specific form of this transformation.

[0061] In this embodiment, the specific steps for the vision system module 110 to obtain the position of the defect are as follows:

[0062] S201: According to the internal parameters of the camera imaging (the internal parameters are the properties of the camera itself and are independent of the scene, such as focal length, principal point, distortion coefficient, and image size), and the external parameters (the external parameters describe the position and orientation of the camera relative to the world coordinate system, such as rotation matrix and translation vector), solve the imaging positions of each key point {x1, x2,..., x n} on the curved surface of the imaging system, and according to

[0063] {x 1p , x 2p ,..., x np} = proj C→S (aff M→C ({x1, x2,..., x n})); (1)

[0064] S202: When using the probe 121 for detection, ensure that the probe 121 appears clearly in at least one camera s, and detect the pixel position x s = (X s,i , Y s,i ) of the probe 121 through the trained object detection neural network algorithm;

[0065] S203: According to the sorting method and the Euclidean distance index (the Euclidean distance is a commonly used method for calculating the distance between two points, which is mathematically represented as the straight-line distance between two points), find the three imaging points [x 1p , x 2p ,..., x np} that are closest to x s ; 1p , x 2p , x 3p ;

[0066] S204: Solve the parameters of the centroid method:

[0067]

[0068] S205: Calculate the true coordinates of the probe 121 in the model coordinate system 134:

[0069]

[0070] In this embodiment, the specific process for the digital model 132 to present the detection progress result is as follows:

[0071] S301: Execute steps S101 to S104 and S201 to S205 frame by frame to obtain a sequence of running trajectory points of the probe 121 on the surface of the aircraft, and connect them into a curve;

[0072] S302: Calculate the scanning area of ​​the probe 121 according to the actual width of the probe 121;

[0073] S303: using the blank digital model 132, fill the area scanned by all probes 121 in a single test with a certain color, or map it with the test result picture of the ultrasound system module 120; the area not scanned by the probe 121 is not filled with color.

[0074] In this embodiment, the specific steps of adding the nondestructive testing data entry 133 in the data processing and storage module 130 are:

[0075] S401: Obtain a C-scan image from the ultrasonic system module 120, run the defect target automatic detection program, if a defect is detected, obtain the defect location, area, depth, time range, and run S402 to S404, otherwise wait for the next ultrasonic system input result;

[0076] S402: For one push-sweep trajectory of the probe 121, the trajectory of the probe 121 during the push-sweep process and its coverage area in the model coordinate system 134 are obtained through steps S101 to S104 and S201 to S205;

[0077] S403: aligning the time range outputted from S401 with the position of the probe 121 in the time range in S402, and obtaining the position of the model coordinate system 134 corresponding to the probe 121 when the defect is detected;

[0078] S404: The position of the defect in the model coordinate system 134 obtained in S403 and the defect depth, area, and shape information obtained in S401 are aggregated to form a defect detection record, which is stored in the database.

[0079] In this embodiment, the visual system module 110, the ultrasonic system module 120, and the data processing and storage module 130 need to be connected to the Internet through 5G, WiFi or other local or global networks to realize data transmission and reception sharing; after the hardware part of the detection system is built based on the above requirements, combined with Figure 2 Introduce data analysis and automatic detection processes at the algorithm level.

[0080] The present application also discloses an in-situ automatic detection method for aircraft fuselage defects, which uses an in-situ automatic detection system for aircraft fuselage defects and includes the following steps: performing marker detection and automatic detection of the position of the ultrasonic probe through the vision system module 110 to align the aircraft digital model coordinate system and the camera coordinate system, and calculating the detailed position of the probe 121 on the target aircraft 140; detecting internal defects in the fuselage skin-grid structure through the ultrasonic system module 120 to obtain ultrasonic detection results; and implementing the target detection and calculation functions of the vision system and the ultrasonic system, as well as providing a database responsible for storing all damage entries through the data processing and storage module 130.

[0081] The present invention can realize the functions of automatically obtaining the position coordinates of detection data and storing them in a standardized manner, as well as visualizing the detection area. The overall degree of automation is high, which can greatly reduce the manual input. Through the in-situ automatic detection of fuselage defects, the present application can effectively avoid problems such as area omission, area duplication, and inaccurate judgment of defect positions, improving the detection accuracy and efficiency. Compared with traditional manual detection, the in-situ automatic detection of fuselage defects in the present application effectively avoids the problems of non-unique reference objects and poor standardization. For the defect record data in the present application, it can be directly saved without the need for secondary processing to be saved as digital data, which is convenient for statistical analysis of defects and detection situations.

[0082] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these changes and modifications.

[0083] The above is only the preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and all should be covered by the protection scope of the present invention.

Claims

1. An in-situ automatic detection system for aircraft fuselage defects, characterized in that, It includes a vision system module, an ultrasonic system module, and a data processing and storage module; The vision system module is used for marker detection and automatic detection of the ultrasonic probe position, to align the aircraft digital mock-up coordinate system and the camera coordinate system, and calculate the detailed position of the probe position on the target aircraft; The ultrasonic system module uses a phased array C-scan ultrasonic device to detect internal defects in the fuselage skin-grid structure and obtain ultrasonic detection results; The data processing and storage module is a computing device with computing functions and a certain amount of storage media, which is used to be responsible for realizing the target detection and computing functions of the vision system and the ultrasonic system, and providing a database to store all damage entries.

2. The in-situ automatic detection system for aircraft fuselage defects according to claim 1, wherein The vision system module includes one or more networked cameras and several coded markers attached to specific positions on the fuselage. Among them, the carrier platform of the networked cameras includes a drone platform and an external fixed bracket, and each networked camera needs to be calibrated to determine the internal parameter matrix.

3. An in-situ automated inspection system for aircraft fuselage defects according to claim 2, characterized in that, The ultrasonic system module includes one or more networked ultrasonic non-destructive testing devices, which are used to detect damage points such as delamination and debonding of the aircraft skin and grid structure, and the probes of the networked ultrasonic non-destructive testing devices include one of a phased array ultrasonic system and an A-scan probe.

4. The in-situ automatic inspection system for aircraft fuselage defects according to claim 3, characterized in that, The data processing and storage module includes a networked host server and a display. Among them, the networked host server is used to provide storage and computing services, and the display is used to present the results of the detection progress and defect statistics.

5. An in-situ automatic detection system for aircraft fuselage defects according to claim 4, characterized in that, The data processing and storage module needs to store a high-fidelity three-dimensional 1:1 digital model of the aircraft under maintenance. This model stores the curved surface key points and grid information on the aircraft surface, and all key points have unique coordinates in the model coordinate system.

6. An in-situ automatic detection system for aircraft fuselage defects according to claim 5, characterized in that, The specific steps for the vision system module to align the aircraft digital mock-up coordinate system and the camera coordinate system are as follows: S101: Affix coded markers at preset positions on the fuselage surface, and their positions and distributions meet two requirements: The first requirement is that each networked camera can clearly capture at least three coded markers simultaneously; The second requirement is that the coded markers need to accurately correspond to the key points with known coordinates on one of the curved surfaces of the digital model; S102: For the camera s of interest, the projection transformation corresponding to its known internal parameter is denoted as proj C→S , and take a photo at one of its poses; S103: Run the marker detection program to ensure that n visual positioning marker points can be clearly detected within its field of view, with their coordinates being (X M,i , Y M,i , Z M,i ), where 1 ≤ i ≤ n and n ≥ 3, and their imaging pixel position coordinates on this camera s are (X s,i , Y s,i ); S104: Run the program for solving the spatial transformation matrix. Input the camera internal parameters and the point pairs to solve the orthogonal transformation matrix R and the translation vector T from the camera s coordinate system to the model coordinate system. The parameters thereof form the affine transformation parameters, denoted as aff M→S 。 7. An in-situ automatic detection system for aircraft fuselage defects according to claim 6, characterized in that, The specific steps for the vision system module to obtain the position of the defect are as follows: S201: According to the internal and external parameters of camera imaging, solve the surface imaging positions of each key point {x1, x2,..., x n} in the imaging system, and according to {x 1p ,x 2p ,...,x np} = proj C→S (aff M→C ({x1,x2,...,x n})); (1) S202: When using the probe for detection, ensure that the probe appears clearly in at least one camera s, and detect the pixel position x where the probe is located through the trained object detection neural network algorithm s =(X s,i , Y s,i ); S203: According to the sorting method and the Euclidean distance metric, find the three imaging points [x 1p , x 2p ,..., x np} in {x s that are closest to x 1p , x 2p , x 3p ; S204: Solve the centroid method parameters: S205: Calculate the true coordinates of the probe in the model coordinate system:

8. An in-situ automatic detection system for aircraft fuselage defects according to claim 7, characterized in that, The specific process for the digital model to present the detection progress results is as follows: S301: Run steps S101 - S104, S201 - S205 frame by frame to obtain a sequence of running trajectory points of the probe on the aircraft surface and connect them into a curve; S302: Calculate the probe scanning area according to the actual width of the probe; S303: Use the blank digital model to fill the area scanned by all probes in a single detection with a certain color, or map the detection result picture of the ultrasonic system module; the area not scanned by the probe is not filled with color.

9. An in-situ automatic detection system for aircraft fuselage defects according to claim 8, characterized in that, The specific steps for adding non-destructive testing data entries in the data processing and storage module are as follows: S401: Get the C-scan image from the ultrasonic system module, run the defect target automatic detection program, if a defect is detected, get the defect location, area, depth, time range, and run S402 to S404, otherwise wait for the next ultrasonic system input result; S402: For one push-sweep trajectory of the probe, the trajectory of the probe during the push-sweep process and its coverage area in the model coordinate system are obtained through steps S101 to S104 and S201 to S205; S403: aligning the time range outputted in S401 with the position of the probe in the time range in S402, and obtaining the model coordinate system position corresponding to the probe when the defect is detected; S404: The position of the defect in the model coordinate system obtained in S403 and the defect depth, area, and shape information obtained in S401 are aggregated to form a defect detection record, which is stored in the database.

10. An in-situ automatic detection method for aircraft fuselage defects, characterized in that, The in-situ automated detection system for aircraft fuselage defects as described in claims 1 to 9 comprises the following steps: The visual system module is used to detect markers and automatically detect the position of the ultrasonic probe, align the aircraft digital model coordinate system and the camera coordinate system, and calculate the detailed position of the probe on the target aircraft; The ultrasonic system module is used to detect internal defects in the fuselage skin-grid structure and obtain ultrasonic test results; The data processing and storage module realizes the target detection and calculation functions of the visual system and ultrasonic system, and provides a database to store all damage entries.

Images