A device and method for detecting micro-cracks on the inner and outer surfaces of a blade and a computer device

Abstract

A device, method, and computer equipment for detecting internal and external microcracks in blades are disclosed. The device includes a projector stabilization mechanism; an infrared camera mechanism, a first industrial camera mechanism, and a second industrial camera mechanism, arranged circumferentially and connected to the projector stabilization mechanism for adjusting camera yaw and rotational degrees of freedom; and a connecting component for connecting the projector stabilization mechanism and a robotic arm. The method includes: adjusting the detection positions of the projector, infrared camera, and industrial cameras; developing a segmented image information acquisition scheme based on the non-destructive blade at the detection position; capturing infrared thermal images, visible light images, and 3D point cloud images of the blade to be detected using the set acquisition scheme to obtain a 3D information dataset for a single blade; and inputting the information dataset into a microcrack detection model to detect and distinguish internal and external microcracks in the blade to be detected. This invention features a compact structure, high detection efficiency, and reduced false detection rate.

Description

Technical Field

[0001] This invention belongs to the field of aero-engine turbine blade inspection technology, specifically relating to a device, method, and computer equipment for detecting internal and external microcracks in blades based on infrared and vision. Background Technology

[0002] Turbine blades are critical hot-end components of aero-engines, characterized by "small structure, high load, small space, and high temperature," making them one of the most frequently failing parts in aero-engines. Turbine blades rotate at high speeds for extended periods under extreme conditions such as high temperature and high pressure, experiencing continuous reciprocating stress, which can lead to fatigue and overload damage. This results in microcracks on the blade surface or inside, severely impacting the engine's safe lifespan.

[0003] Microcracks in aero-engine turbine blades are classified into surface cracks and internal cracks. However, existing detection methods have a high detection rate for surface defects but struggle to distinguish between internal and external cracks. Furthermore, most non-destructive testing methods have low detection rates and high false identification rates for microcracks. Due to the thin-walled and complex curved surface structure of aero-engine turbine blades, traditional detection methods are insufficient to detect microcracks. Therefore, more flexible detection equipment and methods are needed. Summary of the Invention

[0004] To overcome the shortcomings of existing technologies, this invention provides a device, method, and computer equipment for detecting internal and external microcracks in blades based on infrared and vision. An innovation is made at the end effector of the robotic arm, allowing the projector and each camera to adjust their pose, resulting in a compact structure and enabling efficient detection. By performing segmented detection of the blades and weighted calculation of two matching degrees, this method achieves a higher detection rate and a lower false detection rate compared to traditional calculation methods.

[0005] A device for detecting microcracks inside and outside blades includes:

[0006] Projector stabilization mechanism for fixing the projector's position in the horizontal and vertical directions;

[0007] An infrared camera mechanism, a first industrial camera mechanism, and a second industrial camera mechanism are arranged circumferentially and connected to a projector stabilization mechanism to achieve adjustments in camera yaw and rotational degrees of freedom and determine the optimal shooting position.

[0008] Connecting components are used to connect the projector stabilization mechanism and the robotic arm so that the robotic arm can control the following motion.

[0009] Furthermore, the projector stabilization mechanism includes a housing, a transmission structure, a longitudinal clamping structure, and a transverse clamping structure;

[0010] A lateral clamping structure is arranged inside the housing for lateral positioning of the projector.

[0011] The transmission structure is fixed to the housing and is used to provide power to the longitudinal clamping structure.

[0012] A longitudinal clamping structure is arranged on the housing for vertical positioning of the projector.

[0013] Furthermore, the infrared camera mechanism, the first industrial camera mechanism, and the second industrial camera mechanism have the same structure, including motor A, motor B, a fixed frame, a frame connecting frame, control component A, a frame, and control component B; the fixed frame is connected to one arm of the three-jaw bracket, the three-jaw bracket is connected to the projector stabilization mechanism, the frame is rotatably mounted on the frame connecting frame, the motor is fixed on the fixed frame, the frame connecting component is fixed on the output shaft of motor A and can rotate relative to the fixed frame, the axis of motor B is perpendicular to the axis of motor A, motor B is fixed on the fixed frame, control component A is fixed on the output shaft of motor B, and control component B is connected to control component A so that the frame can be rotated around the frame connecting frame after motor B is started.

[0014] A method for detecting microcracks inside and outside blades includes the following steps:

[0015] S1. Adjust the detection positions of the projector, infrared camera, and industrial camera;

[0016] S2. Based on the non-destructive blades at the detection position, formulate a block image information acquisition scheme, and divide the blades into i blocks, where i takes the values ​​1, 2, ..., n, and n is a positive integer;

[0017] S3. Using the set acquisition scheme, capture infrared thermal image, visible light image and three-dimensional point cloud image of the leaf to be detected to obtain i three-dimensional information datasets for a single leaf.

[0018] One of the industrial cameras, in conjunction with a projector, projects stripes to perform 3D reconstruction and obtain 3D point cloud data.

[0019] S4. Input the information dataset into the microcrack detection model to detect and distinguish the internal and external microcracks of the blade to be detected.

[0020] Further, the detection process in step S4 includes: inputting the i-th 3D information data into a convolutional neural network, performing feature matching to obtain the matching degree p1 between visible light and infrared thermal images, and the matching degree p2 between visible light and point clouds, with corresponding preset matching degree thresholds L1 and L2, respectively. The matching degree weight optimization formula is: p i =max{p i , αp1+βp2}, i takes the values ​​1 and 2, where α and β are the weighting coefficients;

[0021] Determine the matching degrees p1 and L1 of the blade to be detected, and the magnitudes of the matching degrees p2 and L2 of the blade to be detected;

[0022] If p1 > L1 and p2 > L2, it is determined that the detected blade has a microcrack and it is a surface microcrack;

[0023] If p1 > L1 and p2 < L2, it is determined that the detected blade has a microcrack and it is an internal microcrack;

[0024] If p1 < L1 and p2 < L2, it is determined that the detected blade has no defect.

[0025] A computer device includes:

[0026] A memory for storing a program for implementing the above-mentioned method for detecting internal and external microcracks of a blade;

[0027] A processor for loading and executing the program stored in the memory.

[0028] The beneficial effects of the present invention compared with the prior art are:

[0029] A device for detecting internal and external microcracks of a blade based on infrared and vision according to the present invention can be used for detecting microcracks of turbine blades of aeroengines and detecting microcracks of metals with similar curved surface features. By controlling seven servo motors, the position of the camera can be freely adjusted. When the robotic arm moves, the light source of the projector can maintain stable projection, making its degree of freedom higher than that of traditional end detection devices.

[0030] By performing block detection on the blade and weighted calculation of two matching degrees, compared with traditional calculation methods, the detection rate of the method of the present invention is improved and the false detection rate is reduced.

[0031] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments: Description of the Drawings

[0032] Figure 1 For the advantages of the device for detecting internal and external microcracks of a blade based on infrared and vision according to the present invention;

[0033] Figure 2 A schematic diagram of the connection of the projector stabilization mechanism, the infrared camera mechanism and two sets of industrial camera mechanisms;

[0034] Figure 3 A schematic diagram of the connection of the transmission structure and the longitudinal clamping structure;

[0035] Figure 4 A schematic diagram of the connection of the projector stabilization mechanism and the three-jaw bracket;

[0036] Figure 5 A schematic diagram of the transverse clamping structure;

[0037] Figure 6This is a schematic diagram showing the connection between the connecting component and the projector stabilization mechanism;

[0038] Figure 7 This is a schematic diagram of an infrared camera mechanism or an industrial camera mechanism.

[0039] Figure 8 This is a flowchart of the blade internal and external microcrack detection method based on infrared and vision according to the present invention;

[0040] Figure 9 Flowchart of image preprocessing for a method to detect microcracks inside and outside blades;

[0041] Figure 10 This is a segmented inspection diagram of the method for detecting microcracks inside and outside the blade. Detailed Implementation

[0042] The embodiments of the technical solution of the present invention will now be described in detail with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention. Unless otherwise stated, the technical or scientific terms used in this application should have the ordinary meaning understood by those skilled in the art.

[0043] like Figure 1 As shown, this embodiment of the blade microcrack detection device based on infrared and vision includes:

[0044] Projector stabilization mechanism 1 is used to fix the projector inside it in the horizontal and vertical directions.

[0045] Infrared camera mechanism 2, first industrial camera mechanism 3 and second industrial camera mechanism 5 are arranged circumferentially and connected to projector stabilization mechanism 1 to achieve adjustment of camera tilt and rotation degrees of freedom and determine the optimal shooting position.

[0046] The connecting component 6 is used to connect the projector stabilizing mechanism 1 and the robotic arm 4 so that it can be controlled by the robotic arm 4 to achieve following motion.

[0047] This implementation innovates on the actuator at the end of the robotic arm, enabling each camera to adjust its pose for efficient detection.

[0048] This embodiment designs a robotic arm end effector with high degrees of freedom and flexibility, which effectively improves detection accuracy and reduces false detection rate.

[0049] As a possible implementation method, such as Figures 1-5 As shown, the projector stabilization mechanism 1 includes a housing 30, a transmission structure K, a longitudinal clamping structure H, and a transverse clamping structure F;

[0050] A lateral clamping structure F is arranged inside the housing 30 for lateral positioning of the projector 14.

[0051] The transmission structure K is fixed to the housing 30 and is used to provide power to the longitudinal clamping structure H.

[0052] A longitudinal clamping structure H is arranged on the housing 30 for vertically positioning the projector 14.

[0053] The transmission structure K provides power to control the longitudinal clamping structure H to fix the projector 14 inside the housing 30 in the vertical direction, and the lateral clamping structure F controls the fixation of the projector 14 in the horizontal plane, thereby ultimately achieving the purpose of precise positioning of the projector 14. The lateral clamping structure F and the longitudinal clamping structure H are controlled simultaneously to achieve the best stability effect.

[0054] Furthermore, such as Figure 2 and Figure 3 As shown, the transmission structure K includes a motor 16, a transmission shaft 15, a gear pair, and a worm gear pair 13;

[0055] The motor 16 is mounted on the housing 30. The output end of the motor 16 drives the transmission shaft 15 through a gear pair. The transmission shaft 15 is rotatably mounted on the housing 30. Both ends of the transmission shaft 15 are connected to the worm gear pair 13. The output end of the worm gear is connected to the input end of the longitudinal clamping structure H.

[0056] Furthermore, such as Figure 3 As shown, the longitudinal clamping structure H includes a screw 10, a top plate 7, and a guide rail 24;

[0057] Two screws 10 are arranged on one side of the top plate 7, and a nut is screwed on each screw 10. One end of the screw 10 is connected to the output end of the transmission structure K, and the other end of the screw 10 is rotatably set at the bottom of the housing 10. One side of the top plate 7 is fixed to the nut, and the other side of the top plate 7 is slidably set on the guide rail 24 for positioning the projector 14 inside the housing 30.

[0058] Based on the above scheme, the screw 10 is mounted on the worm gear, a large gear is mounted on the transmission shaft 15, and a small gear is mounted on the output shaft of the motor 16. The small gear and the large gear mesh to realize the rotation of the transmission shaft 15, which in turn drives the worm gear to rotate, realizing the reversal. The two worm gear pairs are symmetrically arranged in the entire structure, driving the screw 10 to rotate, and the nut on it moves linearly. The top plate 7 slides on the guide rail 24 through the slider 9 on it. In this way, the top plate 7 moves up and down linearly under the action of the nut and the slider 9, thereby achieving the purpose of fixing or releasing the projector 14 vertically as needed. A limit block 23 is provided at the bottom of the screw 10, and the screw 10 can rotate around the limit block 23. The limit block 23 is fixed inside the bottom of the housing 30.

[0059] Furthermore, such as Figure 2 and Figure 5 As shown, the transverse clamping structure F includes a base plate 28, a baffle 20, a connecting rod 29, and a spring 17;

[0060] The base plate 28 is fixed to the bottom of the housing 30. Two grooves are arranged obliquely on the base plate 28. A spring 17 is arranged in the groove. Each groove corresponds to a baffle 20. The slider of the baffle 20 can slide along the length of the groove. The two ends of the spring 17 abut against the slider and the groove wall, respectively. Two connecting rods 29 are provided between the two baffles 20. The baffles 20 can slide along the connecting rods 29.

[0061] Based on the above scheme, the projector 14 is placed on two connecting rods 29. The baffles 20 on both sides tend to move in opposite directions under the action of the lower slider and spring 17, generating a clamping force to fix the projector 14 in the horizontal direction.

[0062] Optionally, the groove is a long, narrow groove, with two grooves arranged at an angle, virtually forming the two sides of an isosceles trapezoid. The baffle 20 forms an angle with the groove, creating a cross-shaped arrangement. The force component of the spring's action on the baffle 20 causes it to slide linearly along the connecting rod 29, resulting in the two baffles 20 moving closer together, thus clamping and fixing the projector 14 laterally. Ultimately, this achieves precise longitudinal and lateral positioning of the projector 14.

[0063] In the above, such as Figure 6 As shown, the housing 30 is fixed to the end effector of the robotic arm 4 via a stepped shaft 22 and a connecting component 6. Optionally, the connecting component 6 is made of a magnet, with one side bolted to the stepped shaft 22 and the other side connected to the end effector of the robotic arm 4. Alternatively, the stepped shaft 22 can be fitted inside the hollow cavity of the connecting component 6 and radially fixed by bolts, with the stepped shaft 22 connected to the end effector of the robotic arm 4 by bolts. Alternatively, the connecting component 6 can have a flange structure on the other side, connected to the end effector of the robotic arm 4 by bolts. This configuration ensures a reliable and stable connection.

[0064] In another possible implementation, such as Figure 2 and Figure 7 As shown, the infrared camera mechanism 2, the first industrial camera mechanism 3, and the second industrial camera mechanism 5 have the same structure, including motor A31, motor B32, fixing frame 25, frame connecting frame 11, control component A24, frame 26, and control component B27; the fixing frame 25 is connected to one arm of the three-jaw bracket.

[0065] The three-claw bracket is connected to the projector stabilization mechanism 1. The photo frame 26 is rotatably mounted on the photo frame connecting frame 11. The motor A31 is fixed on the fixed frame 25. The photo frame connecting piece 11 is fixed on the output shaft of the motor A31 and can rotate relative to the fixed frame 25. The axis of the motor B32 is perpendicular to the axis of the motor A31. The motor B32 is fixed on the fixed frame 25. The control piece A24 is fixed on the output shaft of the motor B32. The control piece B27 is connected to the control piece A24 so as to be able to rotate the photo frame 26 around the photo frame connecting frame 11 after the motor B32 is started.

[0066] The stepped shaft 22 is provided with a threaded post 21. The sleeve in the middle of the three-jaw bracket is threadedly connected to the threaded post 21, which can realize the purpose of manually rotating the three-jaw bracket so as to adjust the attitude of the infrared camera and industrial camera on the three-jaw bracket.

[0067] Optionally, both control components A24 and B27 are L-shaped rods, with one end hinged, denoted as hinge shaft 1. The other end of the L-shaped rod of control component A24 is connected to the output shaft of motor B32, enabling circumferential rotation of the L-shaped rod. The other end of the L-shaped rod of control component B27 is hinged to phase frame 26, denoted as hinge shaft 2. The axes of hinge shaft 1 and hinge shaft 2 are perpendicular, and the axis of hinge shaft 2 is parallel to the axis of motor A31. The rotation between phase frame 26 and phase frame connecting frame 11 is denoted as rotation shaft 1. Rotation shaft 1 is arranged such that hinge shaft 2 is perpendicular to rotation shaft 1. The rotation between control component A24 and motor B32 is denoted as rotation shaft 2. Hinges shaft 1 and 2 are perpendicular to rotation shaft 2. With this setup, when motor B32 starts, it drives control component A24 to rotate circumferentially (left and right disturbance), which in turn drives control component B27 to pull phase frame 26 to rotate circumferentially at the first rotating axis; when motor A31 starts, it drives phase frame connecting frame 11 to flip, which in turn drives phase frame 26 to rotate at the second hinge axis, thus realizing the flipping and swaying of phase frame 26 (up and down disturbance).

[0068] When using a robotic arm to detect microcracks, it was found that the end effector was not flexible enough, unable to adjust the relative pose between cameras or between the camera and the projector. Furthermore, the light source projected by the projector was unstable during robotic arm movement, making simultaneous detection impossible. Additionally, it was discovered that using an infrared camera alone for microcrack detection might result in some cracks going undetected, and the method was significantly affected by the external environment, leading to low detection accuracy. However, microcracks on the blade surface and internally have different impacts on blade life and need to be differentiated. Existing detection methods mainly rely on manual inspection, which cannot distinguish between internal and external cracks and has a high false detection rate.

[0069] Blade internal and external microcrack detection devices based on any of the above schemes, such as Figure 8As shown, a method for detecting internal and external microcracks in blades based on infrared and vision is provided. The method includes the following steps:

[0070] S1. Adjust the detection positions of the projector, infrared camera, and industrial camera;

[0071] S2. Based on the non-destructive blades at the detection position, formulate a block image information acquisition scheme, divide the blades into i blocks, where i takes the values ​​1, 2, ..., n, and n is a positive integer; where the numbers represent the order of detection.

[0072] S3. Using the set acquisition scheme, capture infrared thermal image, visible light image and three-dimensional point cloud image of the leaf to be detected to obtain i three-dimensional information datasets for a single leaf.

[0073] One of the industrial cameras, in conjunction with a projector, projects stripes to perform 3D reconstruction and obtain 3D point cloud data.

[0074] S4. Input the information dataset into the microcrack detection model to detect and distinguish the internal and external microcracks of the blade to be detected.

[0075] In this embodiment, when detecting microcracks in blades, the projector and camera can be freely adjusted by controlling seven motors (one motor 16, three motors A31, and three motors B32) to achieve the optimal detection position. This ensures the stability of the projector's light source during the robotic arm's movement, improving the device's degree of freedom and detection efficiency. After adjusting the detection position, a segmented detection technique is used on the blades, and a detection path is designed. The total matching degree is calculated according to the aforementioned matching weight formula, and the presence of microcracks is detected using the described judgment method. Compared to traditional methods, this approach has a higher detection rate and a lower false detection rate.

[0076] Specifically:

[0077] Based on the non-destructive blades at the detection position, a segmented image information acquisition scheme is formulated, such as... Figure 10 The blade is divided into i pieces, where i is 1, 2, ..., n, and n is a positive integer. The numbers represent the order in which the blades are detected.

[0078] The detected infrared thermal images, visible light images, and point cloud images are stored in a dataset, and each leaf is numbered and labeled.

[0079] For the marked segmented blades, the blades to be inspected are segmented according to their numbers, and the data in the database is updated.

[0080] Cancel the marking of the blade blocks that have been inspected.

[0081] To obtain i three-dimensional information datasets for a single leaf, the data is first processed into grayscale, and then noise reduction is achieved using bilateral filtering techniques, such as... Figure 9 As shown, it can implement noise processing of data and retain edge information.

[0082] Specifically, i in the i three-dimensional information datasets refers to the number of blocks of blade segmentation, and three-dimensional means that each block contains three information data: infrared thermal image, visible light image, and point cloud image.

[0083] The method of gradient sparse prior image decomposition is used to decompose the image into two layers, foreground and background, so as to effectively separate the texture and microcrack information in the image.

[0084] The separated microcrack information data is input into a convolutional neural network for feature matching to obtain the matching degree p1 between visible light and infrared thermal image, and the matching degree p2 between visible light and point cloud. The corresponding preset matching degree thresholds are L1 and L2 respectively.

[0085] The selection of the threshold is related to factors such as the parameters of each camera, the environment, and the number of blocks. The specific selection is determined in specific experiments.

[0086] The detection process in step S4 includes:

[0087] The i-th three-dimensional information data is input into a convolutional neural network for feature matching to obtain the matching degree p1 between visible light and infrared thermal image, and the matching degree p2 between visible light and point cloud. The corresponding preset matching degree thresholds are L1 and L2 respectively. The matching degree weight optimization formula is: p i = max{p i , αp1 + βp2}, where i takes 1, 2, and α and β are weight ratio coefficients;

[0088] Among them, the selection of the weight ratio coefficients α and β follows the principle of α > β and α + β = 1;

[0089] Judge the magnitudes of the matching degree p1 of the blade to be detected and L1, and the matching degree p2 of the blade to be detected and L2;

[0090] If p1 > L1 and p2 > L2, it is judged that the detected blade has microcracks and they are surface microcracks;

[0091] If p1 > L1 and p2 < L2, it is judged that the detected blade has microcracks and they are internal microcracks;

[0092] If p1 < L1 and p2 < L2, it is judged that the detected blade has no defects.

[0093] When the above three situations are not met, the i-th blade needs to be marked, and the database path information is exported for re-detection, and steps S1 - S4 are executed until the above three situations are satisfied.

[0094] In particular, a computer device is also provided, comprising:

[0095] The memory is used to store the computer program that implements the above-mentioned method for detecting microcracks inside and outside the blade;

[0096] A processor is used to load and execute the computer program stored in the memory. When the processor executes the program, it implements the above-described method for detecting internal and external microcracks.

[0097] An infrared vision-based device and method for detecting microcracks inside and outside blades is proposed. This device can be used to detect microcracks in aero-engine turbine blades, as well as microcracks in metals with similar curved surface features. By controlling seven servo motors, the camera position can be freely adjusted, and the projector's light source can maintain stable projection during the movement of the robotic arm, giving it greater freedom of movement than traditional end-effector detection devices. By performing segmented detection of the blades and weighted calculation of two matching degrees, this invention improves the detection rate and reduces the false detection rate compared to traditional calculation methods.

[0098] The present invention has been disclosed above with reference to preferred embodiments, but it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed structure and technical content to create equivalent embodiments without departing from the scope of the present invention, and all such modifications or alterations shall still fall within the scope of the present invention.

Claims

1. A device for detecting microcracks inside and outside a blade, characterized by, Include: Projector stabilization mechanism (1) for fixing the position of the projector inside it in the horizontal and vertical directions; Infrared camera mechanism (2), first industrial camera mechanism (3) and second industrial camera mechanism (5) are arranged circumferentially and connected to projector stabilization mechanism (1) to realize camera tilt and rotation degree of freedom adjustment and determine the best shooting position; The connecting component (6) is used to connect the projector stabilization mechanism (1) and the robotic arm (4) so ​​that it can be controlled by the robotic arm (4) to achieve following motion; The projector stabilization mechanism (1) includes a housing (30), a transmission structure (K), a longitudinal clamping structure (H), and a transverse clamping structure (F). A lateral clamping structure (F) is arranged inside the housing (30) for lateral positioning of the projector (14). The transmission structure (K) is fixed to the housing (30) and is used to provide power to the longitudinal clamping structure (H). A longitudinal clamping structure (H) is arranged on the housing (30) for vertically positioning the projector (14). The transmission structure (K) includes a motor (16), a transmission shaft (15), a gear pair and a worm gear pair (13). The motor (16) is mounted on the housing (30). The output end of the motor (16) drives the transmission shaft (15) through a gear pair. The transmission shaft (15) is rotatably mounted on the housing (30). The two ends of the transmission shaft (15) are respectively connected to the worm gear pair (13). The output end of the worm gear is connected to the input end of the longitudinal clamping structure (H). The longitudinal clamping structure (H) includes a screw (10), a top plate (7), and a guide rail (24); two screws (10) are arranged on one side of the top plate (7), each screw (10) is screwed with a nut, one end of the screw (10) is connected to the output end of the transmission structure (K), and the other end of the screw (10) is rotatably set at the bottom of the housing (30). One side of the top plate (7) is fixed to the nut, and the other side of the top plate (7) is slidably set on the guide rail (24) for positioning the projector (14) inside the housing (30). The lateral clamping structure (F) includes a bottom plate (28), a baffle plate (20), a connecting rod (29) and a spring (17); the bottom plate (28) is fixed at the bottom inside the housing (30), two grooves are arranged obliquely on the bottom plate (28), springs (17) are arranged in the grooves, one baffle plate (20) corresponds to each groove, the sliding head of the baffle plate (20) can slide along the length direction of the groove, and both ends of the spring (17) abut against the sliding head and the groove wall respectively. Two connecting rods (29) are arranged between the two baffle plates (20), and the baffle plate (20) can slide along the connecting rod (29); the infrared camera mechanism (2), the first industrial camera mechanism (3) and the second industrial camera mechanism (5) have the same structure, and include a motor A (31), a motor B (32), a fixing bracket (25), a photo frame connecting bracket (11), a control part A (24), a photo frame (26) and a control part B (27); the fixing bracket (25) is connected to one arm of the three-jaw bracket, the three-jaw bracket is connected to the projector stabilizing mechanism (1), the photo frame (26) is rotatably arranged on the photo frame connecting bracket (11), the motor A (31) is fixed on the fixing bracket (25), the photo frame connecting bracket (11) is fixed on the output shaft of the motor A (31) and can rotate relative to the fixing bracket (25), the axis of the motor B (32) is perpendicular to the axis of the motor A (31), the motor B (32) is fixed on the fixing bracket (25), the control part A (24) is fixed on the output shaft of the motor B (32), and the control part B (27) is connected to the control part A (24) for拨动 the photo frame (26) to rotate around the photo frame connecting bracket (11) after the motor B (32) is started.

2. A method for detecting microcracks inside and outside blades, characterized in that: Implemented based on the detection device described in claim 1, characterized in that: the method includes the following steps: S1. Adjust the detection positions of the projector, the infrared camera and the industrial camera; S2. Develop a segmented image information acquisition plan according to the non-destructive blades at the detection positions, divide the blades into i pieces, where i takes 1, 2,..., n, and n is a positive integer; S3. Take infrared thermal images, visible light images and three-dimensional point cloud images of the blades to be detected according to the set acquisition plan, and obtain i three-dimensional information data sets of a single blade; Among them, one industrial camera cooperates with the stripes projected by the projector for three-dimensional reconstruction to obtain three-dimensional point cloud data; S4. Input the information data set into the micro-crack detection model to detect and distinguish the internal and external micro-cracks of the blades to be detected; The detection process in step S4 includes: The i-th 3D information data is input into a convolutional neural network for feature matching to obtain the matching degree p1 between visible light and infrared thermal images, and the matching degree p2 between visible light and point clouds. The corresponding preset matching degree thresholds are L1 and L2, respectively. The matching degree weight optimization formula is: p i =max{p i , αp1+βp2}, i takes the values ​​1 and 2, where α and β are the weighting coefficients; Judge the magnitudes of the matching degrees p1 and L1 of the blade to be detected, and the matching degrees p2 and L2 of the blade to be detected; If p1 > L1 and p2 > L2, it is judged that the detected blade has micro-cracks and is surface micro-cracks; If p1 > L1 and p2 < L2, it is judged that the detected blade has micro-cracks and is internal micro-cracks; If p1 < L1 and p2 < L2, it is judged that the detected blade has no defects.

3. A method for detecting internal and external micro-cracks of blades according to claim 2, characterized in that: If the above three conditions are not met, the i-th blade needs to be marked, and the database path information needs to be exported to re-detect and execute steps S1-S4 until the above three conditions are met.

4. A computer device, characterized in that: Includes: a memory for storing a program for implementing the blade internal and external microcrack detection method as described in any one of claims 2-3; A processor for loading and executing programs stored in the memory.

Images