Method and device for eliminating detection errors of near-surface defects using shear speckle interferometry

Abstract

The present invention discloses a method and device for eliminating detection errors in near-surface defects using shear speckle interferometry. The method sequentially involves setting the shear magnitude, rotating and adjusting the shear direction, rotating and adjusting the slit aperture, dynamically loading and collecting a light intensity map, calculating and screening a shear phase map, locating abnormal fringes, summarizing detection results in each direction, and calculating the defect distribution. Shear phase maps in different shear directions are used as defect detection targets, distributing the defect detection errors introduced by shear along different directions. The defect detection errors introduced by shear are eliminated by taking the intersection of the measurement results in each shear direction, achieving high-precision detection of near-surface defects in materials. Shear-induced defect detection errors can be directly and accurately eliminated without the need for shear calibration and compensation, resulting in the determination of the defect distribution of the test piece. The entire process is convenient and fast, with a low detection threshold.

Description

Technical Field

[0001] The present invention relates to the technical field of photoelectric detection, and in particular to a method for eliminating detection errors of near-surface defects using shear speckle interferometry, and also to a device for eliminating detection errors of near-surface defects using shear speckle interferometry. The device aims to eliminate shear amount errors in defect detection results by utilizing the differences in shear amount error directions in shear phase images under different shear directions. Background Art

[0002] With the rapid development of materials processing technology, composite materials, with their advantages such as lightweight and customizable mechanical properties, have become a hot topic in materials research and development, and are widely used in industries such as aerospace. However, defective products are inevitable in composite material processing. Furthermore, internal damage to the material is common, as some components are subjected to long-term working environments such as high temperature and high pressure. These near-surface defects are hidden within the material and cannot be directly observed. Shear speckle interferometry can be used to obtain a shear phase image of the test piece, from which the location and size of the defects can be analyzed, facilitating industrial analysis and repair.

[0003] The basic principle of dynamic detection of near-surface defects using shear speckle interferometry is to extract the spatial carrier phase based on each frame of speckle information generated after the laser beam irradiates the material, and then calculate the shear phase map of each frame in real time, and then mark the near-surface defect information of the material from it. However, in the shear phase map produced by shear speckle interferometry technology, the abnormal fringes and defects do not completely overlap. This is because shearing in shear speckle interferometry involves the splitting, dislocation, and superposition of light information, which causes the abnormal fringes in the shear phase map to also be dislocated and expanded. If the abnormal fringes in the shear phase map are directly used to represent the defect distribution, it will cause a large defect detection error. This error is described by the defect detection error introduced by the shear amount.

[0004] In shear speckle interferometry, the shear effect on the misalignment and expansion of abnormal fringes is inevitable, and the direction and magnitude of the misalignment and expansion of fringes cannot be observed from the shear phase image, making it difficult to perform post-compensation for the shear effect from an algorithmic perspective. Most scientific research chooses to ignore this error and roughly uses the distribution of abnormal fringes to describe the defect position, which leads to an overall bias in the defect detection results and inaccurate positioning of the center point. Some scholars have also attempted to calibrate the shear effect of the experimental device and then compensate for the defect detection results accordingly. Although this can reduce the defect detection error introduced by the shear effect to a certain extent, the entire system needs to be recalibrated every time the size and direction of the shear effect are adjusted. The calibration process is cumbersome and has high technical barriers. The direction of the shear effect is also difficult to determine. Therefore, the method of post-compensation for the shear effect is not applicable to a large range of areas. In summary, when using shear speckle interferometry for near-surface defect detection of materials, there is a defect detection error introduced by the shear effect. This error cannot be eliminated in principle, and the existing technology is not effective in handling this error. Summary of the Invention

[0005] In order to overcome the defects of the prior art, the technical problem to be solved by the present invention is to provide a method for eliminating detection errors of near-surface defects using shear speckle interferometry, which can eliminate the defect detection errors introduced by shear amounts in all directions and achieve high-precision measurement of near-surface defects using shear speckle interferometry without the need for shear amount calibration.

[0006] The technical solution of the present invention is: a method for eliminating detection errors of near-surface defects using shear speckle interferometry, comprising the following steps:

[0007] (1) Setting the shearing amount: In the shearing speckle interferometer system, the initial shearing amount is applied to the shearing speckle interferometer system by tilting the shearing mirror, and the shearing amount and spatial carrier are introduced at the same time;

[0008] (2) Rotation adjustment of shearing direction: The shearing direction is adjusted by rotating the shear mirror along the optical axis;

[0009] (3) Rotate and adjust the slit diaphragm angle: rotate the slit diaphragm along the optical axis to make the filtering direction of the slit consistent with the direction of the spatial carrier introduced by the shearing mirror;

[0010] (4) Dynamically load the test piece and continuously collect the light intensity map in the current shear direction, apply a slowly changing external force to the test piece, and use a camera to continuously record the light intensity information;

[0011] (5) Calculate and screen the shear phase diagram of the current shear direction: Based on the light intensity diagram continuously collected in the current shear direction, calculate the shear phase diagram under each external force magnitude through the spatial carrier technology. Among all the shear phase diagrams under different external force magnitudes, select a suitable shear phase diagram as the result of the current shear direction according to the index, so that the abnormal fringes in the shear phase diagram are sufficiently prominent and their range tends to be stable as the load increases;

[0012] (6) locating abnormal fringes from the shear phase image: locating the position of the abnormal fringes in the shear phase image filtered out in step (5), and obtaining the two-dimensional distribution of the abnormal fringes in the pixel coordinate system;

[0013] (7) Determine the position of the abnormal fringes in each shear direction: Select enough shear directions and repeat steps (2) to (6) to obtain the two-dimensional distribution of the abnormal fringes in each shear direction;

[0014] (8) Summarize and solve the defect pixel coordinate system distribution: Take the intersection of the two-dimensional distribution of abnormal stripes under each shearing direction pixel by pixel, and use the different error directions introduced by the shearing amount under different shearing directions to eliminate them by taking the intersection method to obtain the two-dimensional distribution of defects in the pixel coordinate system;

[0015] (9) Calculate the distribution of defects in the world coordinate system.

[0016] The present invention establishes a high-precision dynamic detection method for near-surface defects of shear speckle interferometry in a rotating shear direction by sequentially setting the shear amount, rotating to adjust the shear direction, rotating to adjust the slit diaphragm, dynamically loading and collecting light intensity maps, calculating and screening shear phase maps, locating abnormal stripes, summarizing detection results in each direction, and calculating defect distribution. The present invention uses shear phase maps under different shear directions as defect detection objects, so that the defect detection error introduced by the shear amount is distributed along different directions. The defect detection error introduced by the shear amount is eliminated by taking the intersection of the measurement results of each shear direction, thereby achieving high-precision detection of near-surface defects of the material. The present invention can directly and accurately eliminate the defect detection error introduced by the shear amount and obtain the defect distribution of the test piece by detecting multiple shear directions of the test piece without the need for shear amount calibration and compensation. The whole process is convenient and fast, and the detection threshold is low. In summary, the present invention eliminates the defect detection error introduced by the shear amount by rotating the shear direction, improves the accuracy of the shear speckle interferometry technology when performing defect detection through the shear phase map, and is more conducive to industrial detection and material maintenance.

[0017] A device for detecting and eliminating errors of near-surface defects using shear speckle interferometry is also provided, comprising:

[0018] a shearing amount setting module configured to apply an initial shearing amount to the shearing speckle interferometry system by tilting the shearing mirror, and simultaneously introduce the shearing amount and the spatial carrier;

[0019] A rotational shear direction adjustment module is configured to achieve rotation of the shear direction along the optical axis by rotating the shear mirror along the optical axis to adjust the shear direction;

[0020] A slit diaphragm angle adjustment module is configured to rotate the slit diaphragm along the optical axis so that the filtering direction of the slit is consistent with the direction of the spatial carrier introduced by the shearing mirror;

[0021] The DUT dynamic loading and acquisition module is configured to apply a gradually changing external force to the DUT and continuously record light intensity information using a camera;

[0022] The shear phase image calculation and screening module is configured to calculate the shear phase image under each external force magnitude based on the continuously collected light intensity image under the current shear direction using spatial carrier technology. Among all the shear phase images under different external force magnitudes, an appropriate shear phase image is selected as the result for the current shear direction based on the indicators, so that the abnormal fringes in the shear phase image are sufficiently prominent and their range tends to be stable as the load increases;

[0023] A shear phase image abnormal fringe positioning module is configured to locate the position of abnormal fringes in the shear phase image under each shear direction and obtain the two-dimensional distribution of the abnormal fringes in the pixel coordinate system;

[0024] The module for determining the position of abnormal fringes in each shearing direction is configured to intersect the two-dimensional distribution of abnormal fringes in each shearing direction pixel by pixel. Taking advantage of the different error directions introduced by the shearing amount in different shearing directions, these fringes are eliminated by taking the intersection method to obtain the two-dimensional distribution of defects in the pixel coordinate system.

[0025] The defect pixel coordinate system distribution summary and solution module is configured to intersect the two-dimensional distribution of abnormal stripes in each shear direction pixel by pixel. The error direction introduced by the shear amount in different shear directions is different, and the two-dimensional distribution of defects in the pixel coordinate system is eliminated by taking the intersection method;

[0026] The defect world coordinate system distribution calculation module is configured to calculate the specific size and location of the defect. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 The figure is a flow chart of the method for eliminating detection errors of near-surface defects using shearing speckle interferometry according to the present invention.

[0028] Figure 2 It is the result of locating abnormal stripes under three different shearing directions by instance segmentation neural network.

[0029] Figure 3 It is the two-dimensional distribution of defects after the intersection of the abnormal stripe positioning results under three different shear directions.

[0030] Figure 4 It is a structural schematic diagram of the detection error elimination system of shearing speckle interferometry near-surface defects according to the present invention.

[0031] Among them: 1- shear speckle interferometry module in the rotational shear direction, 2- CMOS camera, 3- first lens, 4- second lens, 5- imaging lens, 6- slit aperture, 7- test piece, 8- laser, 9- beam splitter, 10- reflector, 11- shear mirror, 12- computer. DETAILED DESCRIPTION

[0032] like Figure 1 As shown, the method for eliminating detection errors of near-surface defects using shear speckle interferometry includes the following steps:

[0033] (1) Setting the shearing amount: In the shearing speckle interferometer system, the initial shearing amount is applied to the shearing speckle interferometer system by tilting the shearing mirror, and the shearing amount and spatial carrier are introduced at the same time;

[0034] (2) Rotation adjustment of shearing direction: The shearing direction is adjusted by rotating the shear mirror along the optical axis;

[0035] (3) Rotate and adjust the slit diaphragm angle: rotate the slit diaphragm along the optical axis to make the filtering direction of the slit consistent with the direction of the spatial carrier introduced by the shearing mirror;

[0036] (4) Dynamically load the test piece and continuously collect the light intensity map in the current shear direction, apply a slowly changing external force to the test piece, and use a camera to continuously record the light intensity information;

[0037] (5) Calculate and screen the shear phase diagram of the current shear direction: Based on the light intensity diagram continuously collected in the current shear direction, calculate the shear phase diagram under each external force magnitude through the spatial carrier technology. Among all the shear phase diagrams under different external force magnitudes, select a suitable shear phase diagram as the result of the current shear direction according to the index, so that the abnormal fringes in the shear phase diagram are sufficiently prominent and their range tends to be stable as the load increases;

[0038] (6) locating abnormal fringes from the shear phase image: locating the position of the abnormal fringes in the shear phase image filtered out in step (5), and obtaining the two-dimensional distribution of the abnormal fringes in the pixel coordinate system;

[0039] (7) Solve the abnormal stripe position under each shear direction: Select enough shear directions and repeat steps (2)-(6) to obtain the two-dimensional distribution of abnormal stripes under each shear direction (such as Figure 3 shown);

[0040] (8) Summarize and solve the defect pixel coordinate system distribution: Take the intersection of the two-dimensional distribution of abnormal stripes under each shearing direction pixel by pixel, and use the different error directions introduced by the shearing amount under different shearing directions to eliminate them by taking the intersection method to obtain the two-dimensional distribution of defects in the pixel coordinate system;

[0041] (9) Calculate the distribution of defects in the world coordinate system.

[0042] The present invention establishes a high-precision dynamic detection method for near-surface defects of shear speckle interferometry in a rotating shear direction by sequentially setting the shear amount, rotating to adjust the shear direction, rotating to adjust the slit diaphragm, dynamically loading and collecting light intensity maps, calculating and screening shear phase maps, locating abnormal stripes, summarizing detection results in each direction, and calculating defect distribution. The present invention uses shear phase maps under different shear directions as defect detection objects, so that the defect detection error introduced by the shear amount is distributed along different directions. The defect detection error introduced by the shear amount is eliminated by taking the intersection of the measurement results of each shear direction, thereby achieving high-precision detection of near-surface defects of the material. The present invention can directly and accurately eliminate the defect detection error introduced by the shear amount and obtain the defect distribution of the test piece by detecting multiple shear directions of the test piece without the need for shear amount calibration and compensation. The whole process is convenient and fast, and the detection threshold is low. In summary, the present invention eliminates the defect detection error introduced by the shear amount by rotating the shear direction, improves the accuracy of the shear speckle interferometry technology when performing defect detection through the shear phase map, and is more conducive to industrial detection and material maintenance.

[0043] Preferably, in step (1), the shearing amount is introduced by tilting the shearing mirror. The shearing amount is selected so that the interference fringes at the defect appear within the loading range, and the interference fringes at different defects do not overlap as much as possible. The image side shearing amount is set to 5mm and then fine-tuned.

[0044] Preferably, in step (2), a rolling table is used to rotate the shear mirror along the optical axis so that the shearing amount remains unchanged and only the shearing direction is changed.

[0045] Preferably, in step (3), the slit diaphragm is rotated along the optical axis using a rolling stage so that the filtering direction and the shearing direction of the slit are consistent, so as to perform spatial carrier phase extraction.

[0046] Preferably, in step (4), an internal loading method of applying air pressure loading to the test piece is used, and a CMOS camera with a small pixel size is selected as the camera.

[0047] Preferably, in step (5), when selecting a suitable shear phase diagram, the ratio of the width of the outer circle stripes in the abnormal stripes to the width of the entire stripes is calculated, and the shear phase diagram corresponding to the minimum load when the ratio is stably lower than 10% is the final shear phase diagram under this set of shear directions.

[0048] Preferably, in step (6), the positioning of abnormal fringes in the shear phase image uses instance segmentation neural network technology to locate the pixel-level distribution of the abnormal fringes; a sufficient number of shear phase images in each shear direction are obtained through experiments, abnormal defects are manually marked and a data set is produced, and the corresponding instance segmentation neural network is used to complete the training.

[0049] Preferably, in step (7), the shear direction is measured every 15°, with a total of 24 groups of measurements.

[0050] Preferably, in step (8), the abnormal stripe detection results in each shearing direction are stored in the form of a mask, and then the intersection is calculated pixel by pixel; in step (9), the coordinate system conversion process is completed using camera calibration technology.

[0051] Those skilled in the art will appreciate that all or part of the steps in the above-described method can be implemented by instructing the relevant hardware through a program. The program can be stored in a computer-readable storage medium. When executed, the program includes the steps of the above-described method. The storage medium can be ROM / RAM, a magnetic disk, an optical disk, a memory card, etc. Therefore, corresponding to the method of the present invention, the present invention also includes a device for detecting and eliminating errors in near-surface defects using sheared speckle interferometry. The device is generally represented in the form of functional modules corresponding to the steps of the method. The device includes:

[0052] a shearing amount setting module configured to apply an initial shearing amount to the shearing speckle interferometry system by tilting the shearing mirror, and simultaneously introduce the shearing amount and the spatial carrier;

[0053] A rotational shear direction adjustment module is configured to achieve rotation of the shear direction along the optical axis by rotating the shear mirror along the optical axis to adjust the shear direction;

[0054] A slit diaphragm angle adjustment module is configured to rotate the slit diaphragm along the optical axis so that the filtering direction of the slit is consistent with the direction of the spatial carrier introduced by the shearing mirror;

[0055] The DUT dynamic loading and acquisition module is configured to apply a gradually changing external force to the DUT and continuously record light intensity information using a camera;

[0056] The shear phase image calculation and screening module is configured to calculate the shear phase image under each external force magnitude based on the continuously collected light intensity image under the current shear direction using spatial carrier technology. Among all the shear phase images under different external force magnitudes, an appropriate shear phase image is selected as the result for the current shear direction based on the indicators, so that the abnormal fringes in the shear phase image are sufficiently prominent and their range tends to be stable as the load increases;

[0057] A shear phase image abnormal fringe positioning module is configured to locate the position of abnormal fringes in the shear phase image under each shear direction and obtain the two-dimensional distribution of the abnormal fringes in the pixel coordinate system;

[0058] The module for determining the position of abnormal fringes in each shearing direction is configured to intersect the two-dimensional distribution of abnormal fringes in each shearing direction pixel by pixel. Taking advantage of the different error directions introduced by the shearing amount in different shearing directions, these fringes are eliminated by taking the intersection method to obtain the two-dimensional distribution of defects in the pixel coordinate system.

[0059] The defect pixel coordinate system distribution summary and solution module is configured to intersect the two-dimensional distribution of abnormal stripes in each shear direction pixel by pixel. The error direction introduced by the shear amount in different shear directions is different, and the two-dimensional distribution of defects in the pixel coordinate system is eliminated by taking the intersection method;

[0060] The defect world coordinate system distribution calculation module is configured to calculate the specific size and location of the defect.

[0061] The structural diagram of the shear speckle interferometry near-surface defect detection error elimination system of the present invention is as follows: Figure 4 As shown. The laser 8 illuminates the test piece 7 and generates speckle. The speckle information passes through the slit diaphragm 6 and the imaging lens 5 and enters the 4f system composed of the second lens 4 and the first lens 3. The speckle information is divided into two parts by the beam splitter 9. One part is used as reference information and returns along the original path through the reflector 10. The other part is tilted by the shearing mirror 11 and a shear amount is introduced. The two light beams converge at the first lens 3 and interfere with each other. The CCD 2 records the interference results and completes the measurement data in the shear direction. By rotating the slit diaphragm 6 and the shearing mirror 11 along the optical axis, shear speckle interferometry measurement in different shear directions can be achieved. Finally, the computer 12 performs the corresponding image processing to obtain high-precision detection results of near-surface defects of the test piece 7.

[0062] A specific embodiment of the present invention is described in detail below. A high-precision dynamic detection method for near-surface defects using shear speckle interferometry in a rotating shear direction is implemented as follows:

[0063] The process of establishing a high-precision dynamic detection method for near-surface defects using shear speckle interferometry in the rotating shear direction is as follows: Figure 1 As shown, the specific implementation steps are:

[0064] Step 1: Set the shearing amount

[0065] In this example, the shearing amount is applied by adjusting the pitch and yaw of the shearing mirror. The shearing amount is set to be approximately 10 mm, and the initial direction is along the positive direction of the x-axis.

[0066] Step 2: Rotate to adjust the shearing amount

[0067] In the example, the shear mirror is fixed on a rolling table to achieve rolling in the shearing direction.

[0068] Step 3: Rotate and adjust the slit aperture angle

[0069] In the embodiment, the slit diaphragm is fixed on a rolling stage, and the filtering direction and the shearing direction of the slit diaphragm are kept consistent in each shearing direction.

[0070] Step 4: Dynamic loading and acquisition of the device under test

[0071] In this example, an air pump was used to apply air pressure behind the DUT, starting at 0 mmHg and gradually increasing the pressure. Measurements were taken every 5 mmHg. A BFS-U3-200S6C-C camera was used for continuous recording, with a resolution of 2736 x 1824 and a pixel size of 2.4 μm.

[0072] Step 5: Shear phase map calculation and screening

[0073] In this example, spatial carrier phase extraction technology is used to extract the phase of each light intensity image. The phase image under each external force is then subtracted from the phase image under 0 mmHg to obtain the shear phase image under each external force. The width of the outer ring of the abnormal stripes under each external force is calculated, and the width of the abnormal stripes is calculated. The ratio of the two gradually decreases and stabilizes with increasing load. The shear phase image corresponding to the minimum load at which this ratio stabilizes below 5% is the final shear phase image for that set of shear directions. In this example, the shear phase image under a load of 65 mmHg is selected.

[0074] Step 6: Locate abnormal fringes in the shear phase image

[0075] In this example, a large number of shear phase images in each shear direction are collected, and the positions of abnormal stripes are manually marked. A data set is created and trained using the solov2 instance segmentation neural network model to achieve automatic detection of abnormal stripes.

[0076] Step 7: Determine the position of abnormal stripes in each shear direction

[0077] In this example, the shear direction starts from the positive direction of the x-axis and is measured every 15°, with a total of 24 groups in one circle. Repeat steps (2)-(6) to obtain the two-dimensional distribution of abnormal fringes under each shear direction, as shown in Figure 2 shown.

[0078] Step 8: Summarize and solve the defect pixel coordinate distribution

[0079] In this example, a total of 24 groups of detection results in each shearing direction are represented in the form of masks, and the intersection is taken pixel by pixel to eliminate the error introduced by the shearing amount and obtain the true pixel coordinate system distribution of the defect.

[0080] Step 9: Solving the distribution of defects in the world coordinate system

[0081] In this example, a calibration plate is used to calibrate the entire shear speckle interferometry roll detection system, converting the two-dimensional distribution of defects from the pixel coordinate system to the world coordinate system. The mean absolute error of the defect area in the world coordinate system is only 5.0874mm. 2 , accounting for 2.3% of the total defect area.

[0082] The beneficial effects of the present invention are as follows:

[0083] 1. The method for eliminating detection errors of near-surface defects using shear speckle interferometry disclosed in the present invention can accurately and efficiently eliminate shear amount errors from shear phase images without the need for calibrating the shear amount or complex post-processing algorithms, thereby achieving high-precision detection of near-surface defects in materials.

[0084] 2. The shear speckle interferometry near-surface defect detection error elimination device disclosed in the present invention has a simple structure. By adding a rolling table to a conventional shear speckle interferometry system, high-precision roll detection can be achieved, and the hardware modification cost is low.

[0085] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention are still within the scope of protection of the technical solution of the present invention.

Claims

1. A method for eliminating detection errors of near-surface defects using shear speckle interferometry, characterized by: It includes the following steps: (1) Setting the shearing amount: In the shearing speckle interferometer system, the initial shearing amount is applied to the shearing speckle interferometer system by tilting the shearing mirror, and the shearing amount and spatial carrier are introduced at the same time; (2) Rotation adjustment of shearing direction: The shearing direction is adjusted by rotating the shear mirror along the optical axis; (3) Rotate and adjust the slit diaphragm angle: rotate the slit diaphragm along the optical axis to make the filtering direction of the slit consistent with the direction of the spatial carrier introduced by the shearing mirror; (4) Dynamically load the test piece and continuously collect the light intensity map in the current shear direction, apply a slowly changing external force to the test piece, and use a camera to continuously record the light intensity information; (5) Calculate and screen the shear phase diagram of the current shear direction: Based on the light intensity diagram continuously collected in the current shear direction, calculate the shear phase diagram under each external force magnitude through the spatial carrier technology. Among all the shear phase diagrams under different external force magnitudes, select a suitable shear phase diagram as the result of the current shear direction according to the index, so that the abnormal fringes in the shear phase diagram are sufficiently prominent and their range tends to be stable as the load increases; (6) locating abnormal fringes from the shear phase image: locating the position of the abnormal fringes in the shear phase image filtered out in step (5), and obtaining the two-dimensional distribution of the abnormal fringes in the pixel coordinate system; (7) Determine the position of the abnormal fringes in each shear direction: Select enough shear directions and repeat steps (2) to (6) to obtain the two-dimensional distribution of the abnormal fringes in each shear direction; (8) Summarize and solve the defect pixel coordinate system distribution: Take the intersection of the two-dimensional distribution of abnormal stripes under each shearing direction pixel by pixel, and use the different error directions introduced by the shearing amount under different shearing directions to eliminate them by taking the intersection method to obtain the two-dimensional distribution of defects in the pixel coordinate system; (9) Calculate the distribution of defects in the world coordinate system.

2. The method for eliminating detection errors of near-surface defects using shear speckle interferometry according to claim 1, characterized in that: In the step (1), a shearing amount is introduced by tilting the shear mirror. The shearing amount is selected so that the interference fringes at the defect appear within the loading range and the interference fringes at different defects do not overlap as much as possible. The image side shearing amount is set to 5 mm and then fine-tuned.

3. The method for eliminating detection errors of near-surface defects using shear speckle interferometry according to claim 2, characterized in that: In the step (2), a rolling table is used to rotate the shear mirror along the optical axis so that the shearing amount remains unchanged and only the shearing direction is changed.

4. The method for eliminating detection errors of near-surface defects using shearing speckle interferometry according to claim 3, characterized in that: In the step (3), the slit diaphragm is rotated along the optical axis using a rolling stage so that the filtering direction and the shearing direction of the slit are consistent, so as to perform spatial carrier phase extraction.

5. The method for eliminating detection errors of near-surface defects using shearing speckle interferometry according to claim 4, characterized in that: In the step (4), an internal loading method of applying air pressure loading is applied to the test piece, and a CMOS camera with a small pixel size is selected as the camera.

6. The method for eliminating detection errors of near-surface defects using shearing speckle interferometry according to claim 5, characterized in that: In the step (2), the shear direction is measured every 15°, and a total of 24 groups of measurements are made; in the step (5), when selecting a suitable shear phase diagram, the ratio of the width of the outer circle stripes in the abnormal stripes to the width of the entire stripes is calculated, and the shear phase diagram corresponding to the minimum load when the ratio is stably lower than 10% is the final shear phase diagram under this group of shear directions.

7. The method for eliminating detection errors of near-surface defects using shearing speckle interferometry according to claim 6, characterized in that: In the step (6), the abnormal stripes in the shear phase image are located using the instance segmentation neural network technology to locate the pixel-level distribution of the abnormal stripes; a sufficient number of shear phase images in each shear direction are obtained through experiments, abnormal defects are manually marked and a data set is prepared, and the corresponding instance segmentation neural network is used to complete the training.

8. The method for eliminating detection errors of near-surface defects using shearing speckle interferometry according to claim 7, characterized in that: In the step (8), the abnormal stripe detection results under each shearing direction are stored in the form of a mask, and then the intersection is calculated pixel by pixel; in the step (9), the coordinate system conversion process is completed using camera calibration technology.

9. A device for eliminating detection errors of near-surface defects using shear speckle interferometry, characterized by: It includes: a shearing amount setting module configured to apply an initial shearing amount to the shearing speckle interferometry system by tilting the shearing mirror, and simultaneously introduce the shearing amount and the spatial carrier; A rotational shear direction adjustment module is configured to achieve rotation of the shear direction along the optical axis by rotating the shear mirror along the optical axis to adjust the shear direction; A slit diaphragm angle adjustment module is configured to rotate the slit diaphragm along the optical axis so that the filtering direction of the slit is consistent with the direction of the spatial carrier introduced by the shearing mirror; The DUT dynamic loading and acquisition module is configured to apply a gradually changing external force to the DUT and continuously record light intensity information using a camera; The shear phase image calculation and screening module is configured to calculate the shear phase image under each external force magnitude based on the continuously collected light intensity image under the current shear direction using spatial carrier technology. Among all the shear phase images under different external force magnitudes, an appropriate shear phase image is selected as the result for the current shear direction based on the indicators, so that the abnormal fringes in the shear phase image are sufficiently prominent and their range tends to be stable as the load increases; A shear phase image abnormal fringe positioning module is configured to locate the position of abnormal fringes in the shear phase image under each shear direction and obtain the two-dimensional distribution of the abnormal fringes in the pixel coordinate system; The module for determining the position of abnormal fringes in each shearing direction is configured to select a sufficient number of shearing directions and repeatedly execute the module for rotating and adjusting the shearing direction, the module for rotating and adjusting the slit aperture angle, the module for dynamically loading and collecting the test piece, the module for calculating and screening the shearing phase map, and the module for locating abnormal fringes in the shearing phase map to obtain the two-dimensional distribution of abnormal fringes in each shearing direction; The defect pixel coordinate system distribution summary and solution module is configured to intersect the two-dimensional distribution of abnormal stripes in each shear direction pixel by pixel. The error direction introduced by the shear amount in different shear directions is different, and the two-dimensional distribution of defects in the pixel coordinate system is eliminated by taking the intersection method; The defect world coordinate system distribution calculation module is configured to calculate the specific size and location of the defect.

Images