A method for adaptive image detection of internal defects of a composite material

Abstract

The present application relates to the technical field of composite material nondestructive testing, and provides a composite material internal defect self-adaptive image detection method, which comprises the following steps: acquiring a polarization image of a composite material by using a polarization measurement device, and performing noise analysis to identify noise in the image, i.e., random scattering noise and high-frequency speckle noise; suppressing the random scattering noise by using the polarization characteristic difference through an IPPs method, and removing the high-frequency speckle noise in the frequency domain by using Fourier transform; and performing image edge detail enhancement by using a self-adaptive Canny algorithm, and identifying a small defect target structure in the picture.The present application solves the problems of high cost and low accuracy of the prior art, and improves the composite material internal defect detection efficiency.

Description

Technical Field

[0001] This invention relates to the field of nondestructive testing technology for composite materials, and in particular to an adaptive image detection method for internal defects in composite materials. Background Technology

[0002] In recent years, composite materials have been widely used globally in key areas such as aircraft structural design, new energy vehicles, and construction due to their advantages of high stiffness, toughness, structural robustness, and ease of processing. In 2025, fiber-based composite materials were listed as one of the key national technologies, further highlighting their important value in both defense and civilian fields. Composite materials typically have a multi-layered sandwich structure, and their manufacturing process determines their common characteristics across multiple scales and levels: macroscopically, they appear as an assembly of interwoven fiber yarns, while microscopically they can be decomposed into micron-sized fiber filaments. However, during production and use, these materials are prone to internal defects such as yarn deformation, bending, and voids. These defects can seriously affect the structural safety performance and service life. Therefore, non-destructive testing (NTD) technology is crucial for the quality and safety testing and life assessment of composite materials.

[0003] Currently common non-destructive testing (NDT) methods include magnetic particle testing (MPs), shear wave interferometry, X-ray inspection, infrared thermography, laser ultrasound, and visual inspection. Among these, ultrasonic NDT is one of the most widely used techniques; however, in practical applications, it requires calibration using a coupling medium and a reference sample, and in complex structures, it is susceptible to interference from multiple reflections, wave scattering, and signal attenuation, posing significant challenges to its testing effectiveness. X-ray computed tomography (CT), as the standard NDT method for quality inspection, provides relatively accurate information about internal structures, but its high equipment cost and potential health risks to operators limit its large-scale application. Furthermore, emerging NDT methods such as Mueller polarization measurement often suffer from reduced image contrast and resolution when dealing with composite materials due to the strong light scattering effect of the materials themselves, especially noticeable in image edge areas, making it difficult to meet high-precision testing requirements. In summary, existing NDT technologies have limitations in terms of adaptability, cost, operational safety, and testing accuracy, necessitating the development of more efficient, reliable, and applicable new testing methods. Summary of the Invention

[0004] To address the above problems, this invention combines the polarization degree index in Mueller polarization measurement with Fourier frequency domain filtering and detail enhancement to process complex noise in images step by step, identify minute defects in composite materials, solve the problems of high cost and low accuracy of existing methods, and improve the efficiency of detecting internal defects in composite materials.

[0005] According to an embodiment of the present invention, an adaptive image detection method for internal defects in composite materials is provided.

[0006] In a first aspect of the invention, an adaptive image detection method for internal defects in composite materials is provided. The method includes:

[0007] Step S01: Use a polarization measurement device to acquire a polarization image of the composite material and perform noise analysis to identify the noise in the image: random scattering noise and high-frequency speckle noise;

[0008] Step S02: Suppress random scattering noise by utilizing the difference in polarization characteristics through the IPPs method, and remove high-frequency speckle noise in the frequency domain using Fourier transform;

[0009] Step S03: Use the adaptive Canny algorithm to enhance image edge details and identify minute defective target structures in the image.

[0010] Furthermore, the noise analysis steps described in step S01 are as follows:

[0011] The polarization image of the composite material is segmented into several regions;

[0012] In Bonga spherical coordinates, the degree of polarization of each region under different combinations of ellipticity and azimuth is calculated using the Mueller matrix and the Stokes vector of the incident light.

[0013] After further separating the noisy image from the original image, the characteristics and distribution patterns of the noise are analyzed.

[0014] Furthermore, the formula for further separating the noisy image from the original image is as follows:

[0015] ,

[0016] In the formula, Represents a noisy image. Represents the original polarized image. This represents a Gaussian image.

[0017] Furthermore, the specific steps for suppressing random scattering noise using the IPPs method based on differences in polarization characteristics described in step S02 are as follows:

[0018] Calculate the coherence matrix of the Mueller matrix;

[0019] The coherence matrix is ​​decomposed to obtain its eigenvalues ​​and submatrices;

[0020] The polarization index is calculated based on eigenvalues ​​and submatrices.

[0021] Furthermore, the specific steps for removing high-frequency speckle noise in the frequency domain using Fourier transform as described in step S02 are as follows:

[0022] Perform a Fourier transform on the image to convert it from the spatial domain to the frequency domain;

[0023] A low-pass filter is applied to filter the frequency domain image;

[0024] Perform an inverse Fourier transform on the filtered frequency domain image to obtain the spatial domain image after removing high-frequency noise.

[0025] Furthermore, the adaptive Canny algorithm described in step S03 automatically finds the high threshold of strong boundary images and the low threshold of weak boundary images based on the gradient distribution of the image. The high threshold is used to filter pixels with extremely strong gradients and determine points with a high probability of being true edges, while the low threshold is used to delete weak boundaries and define the lower limit of the gradient at the boundary.

[0026] Furthermore, the identification of minute defect target structures in the image described in step S03 specifically involves:

[0027] Pixels with gradient magnitudes greater than a high threshold are identified as strong edge points, while pixels with gradient magnitudes less than a low threshold are suppressed.

[0028] For pixels whose gradient magnitude is between the high and low thresholds, they are retained if they are connected to strong edge points, otherwise they are suppressed.

[0029] In a second aspect of the invention, an apparatus for adaptive image detection of internal defects in composite materials is provided. The apparatus includes:

[0030] Noise Analysis Module: Used to acquire polarization images of composite materials using a polarization measurement device, and to perform noise analysis to identify noise in the images: random scattering noise and high-frequency speckle noise;

[0031] Denoising module: Used to suppress random scattering noise by utilizing polarization characteristic differences through the IPPs method, and to remove high-frequency speckle noise in the frequency domain using Fourier transform;

[0032] Defect recognition module: Used to enhance image edge details using the adaptive Canny algorithm and identify minute defect targets in the image.

[0033] In a third aspect of the invention, an electronic device is provided. The electronic device includes a memory and a processor, the memory storing a computer program, the processor executing the program to implement the method according to a first aspect of the invention.

[0034] In a fourth aspect of the invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method according to a first aspect of the invention.

[0035] This invention combines the polarization degree index in Mueller polarization measurement with Fourier frequency domain filtering and detail enhancement to process complex noise in images step by step, identify minute defects in composite materials, solve the problems of high cost and low accuracy of existing methods, and improve the efficiency of detecting internal defects in composite materials.

[0036] It should be understood that the description in the Summary of the Invention is not intended to limit the key or essential features of the embodiments of the present invention, nor is it intended to restrict the scope of the invention. Other features of the invention will become readily apparent from the following description.

[0037] The beneficial effects of this invention are:

[0038] 1. The IPPs method suppresses random scattering noise, significantly improves detection accuracy, and provides some transferability for the identification of other targets;

[0039] 2. Fourier transform effectively suppresses high-frequency noise at image edges, highlighting the spectral characteristics of the target region;

[0040] 3. The adaptive Canny algorithm is used to enhance the structural details of different defect regions and realize the connectivity of defect regions, which effectively improves the efficiency and conversion rate of composite material defect detection, while reducing costs. Attached Figure Description

[0041] The above and other features, advantages, and aspects of the various embodiments of the present invention will become more apparent from the accompanying drawings and the following detailed description. Wherein:

[0042] Figure 1 A flowchart of a method for adaptive image detection of internal defects in composite materials according to an embodiment of the present invention is shown;

[0043] Figure 2 A schematic diagram of a polarization measurement device according to an embodiment of the present invention is shown;

[0044] Figure 3 A schematic diagram of image segmentation according to an embodiment of the present invention is shown;

[0045] Figure 4 A three-dimensional DOP image of a segmented region according to an embodiment of the present invention is shown;

[0046] Figure 5 A noise image is shown according to an embodiment of the present invention;

[0047] Figure 6A schematic diagram of noise grayscale distribution according to an embodiment of the present invention is shown;

[0048] Figure 7 Images of IPPs according to an embodiment of the present invention are shown, (a) is P1, (b) is P2, (c) is P3, and (d) is P4;

[0049] Figure 8 A Fourier transform image according to an embodiment of the present invention is shown;

[0050] Figure 9 A schematic diagram of the frequency domain distribution of noise according to an embodiment of the present invention is shown;

[0051] Figure 10 The FFT images according to an embodiment of the present invention are shown, (a) is P1, (b) is P2, (c) is P3, and (d) is P4;

[0052] Figure 11 The results of the adaptive Canny algorithm according to an embodiment of the present invention are shown in the figure, (a) is P1, (b) is P2, (c) is P3, and (d) is P4;

[0053] Figure 12 Other algorithm results are shown in the embodiments of the present invention: (a) morphological filtering, (b) bilateral filtering, (c) Sobel algorithm, and (d) Laplace algorithm.

[0054] Figure 13 A block diagram of an apparatus for adaptive image detection of internal defects in composite materials according to an embodiment of the present invention is shown;

[0055] Figure 14 A schematic diagram of an apparatus for adaptive image detection of internal defects in composite materials according to an embodiment of the present invention is shown. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] According to an embodiment of the present invention, an adaptive image detection method for internal defects in composite materials is proposed. By combining the polarization degree index in Mueller polarization measurement with Fourier frequency domain filtering and detail enhancement, complex noise in the image is processed step by step to identify minute defects in composite materials. This solves the problems of high cost and low accuracy of existing methods and improves the detection efficiency of internal defects in composite materials.

[0058] The principles and spirit of the present invention will be explained in detail below with reference to several representative embodiments.

[0059] Figure 1 This is a schematic flowchart of an adaptive image detection method for internal defects in composite materials according to an embodiment of the present invention. The method includes:

[0060] Step S01: Use a polarization measurement device to acquire a polarization image of the composite material and perform noise analysis to identify the noise in the image: random scattering noise and high-frequency speckle noise;

[0061] Step S02: Suppress random scattering noise by utilizing the difference in polarization characteristics through the IPPs method, and remove high-frequency speckle noise in the frequency domain using Fourier transform;

[0062] Step S03: Use the adaptive Canny algorithm to enhance image edge details and identify minute defective target structures in the image.

[0063] It should be noted that although the operation of the method of the present invention has been described in a specific order in the above embodiments and figures, this does not require or imply that the operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0064] To provide a clearer explanation of the above-described adaptive image detection method for internal defects in composite materials, a specific embodiment will be used for illustration below. However, it is worth noting that this embodiment is only for better illustrating the present invention and does not constitute an improper limitation of the present invention.

[0065] The following example will further illustrate the adaptive image detection method for internal defects in composite materials.

[0066] Step S01: Use a polarization measurement device to acquire a polarization image of the composite material and perform noise analysis to identify the noise in the image: random scattering noise and high-frequency speckle noise.

[0067] This embodiment utilizes a polarization measurement device and a composite material sample to acquire a polarization image of the composite material. The invention will be further described below with reference to the accompanying drawings and embodiments. Figure 2 For the sake of simplification, the schematic diagram of the device is only used to illustrate the instrument structure used in this invention, and only the components relevant to this invention are shown.

[0068] like Figure 3 As shown, the polarization image of the composite material is divided into 7 regions (adjusted according to the size of the target region in the image, and limited by the instrument resolution and field of view), labeled 1-7 respectively. The polarization degree formula in Bonga spherical coordinates is expressed as:

[0069] ,

[0070] In the formula, Indicates ellipticity. , Indicates azimuth. , The Mueller matrix is ​​represented by... Figure 2 Measurement by the device in the middle, Indicates the Stokes vector of the incident light:

[0071] ,

[0072] Using numerical methods to , Discretize the two values ​​by dividing the two data points into 1000×1000 grids in a two-dimensional space, and calculate the values ​​for each group separately. This allows us to obtain the polarization degree distribution under all polarization states, such as Figure 4 As shown, the three-dimensional DOP distribution is displayed intuitively, which facilitates the analysis of polarization response characteristics and laws, and further analysis of its scattering and noise characteristics.

[0073] like Figure 5 As mentioned above, the image is severely scattered, such as Figure 6 As shown, the noise distribution is uneven, requiring further analysis of the image noise characteristics. The noisy image can be obtained as follows:

[0074] ,

[0075] In the formula, Represents a noisy image. Represents the original polarized image. This represents a Gaussian image. The characteristics and distribution patterns of noise are obtained using this method. The noise is mainly high-frequency noise, including random scattering noise and high-frequency speckle noise. Random scattering noise is related to the propagation path of the polarized laser within the material, while high-frequency speckle noise is caused by the coherence characteristics of the laser within the composite material's internal structure.

[0076] Step S02: Suppress random scattering noise by utilizing the difference in polarization characteristics through the IPPs method, and remove high-frequency speckle noise in the frequency domain using Fourier transform.

[0077] First, the contrast and resolution of the polarization image are improved by using the IPPs method, which suppresses random scattering noise and highlights the structural differences between the target area and the background area.

[0078] For random scattering noise, since it is equivalent to natural light, the eigenvalues ​​of the coherence matrix of the Mueller matrix are equal. The polarization degree index (IPPs) method is used to remove random scattering noise.

[0079] First, calculate the coherence matrix of the Mueller matrix, which is expressed as:

[0080] ,

[0081] In the formula, For coherence matrix, For Mueller matrix elements, Represents the tensor product. Representing the Pauli matrix:

[0082] , , , ,

[0083] Constructing 16 Dirac matrices, the eigenvalues ​​and submatrices can be obtained using coherent matrix decomposition:

[0084] ,

[0085] In the formula, These are the eigenvalues ​​of the coherence matrix. This is a submatrix of the Mueller matrix.

[0086] The IPPs method can be represented as:

[0087] ,

[0088] In the formula, Represents the trace of a matrix. This constitutes an ideal depolarization space, such as Figure 7 As shown, its combination form is:

[0089] ,

[0090] In the formula, They are respectively The proportionality coefficients must satisfy the condition that the sum of the coefficients is 1.

[0091] Furthermore, Fourier transform is used to remove high-frequency noise in the frequency domain, such as... Figures 8-9 As shown, high-frequency noise interference at the edges of details is suppressed and smoothed, as... Figure 10 As shown:

[0092] ,

[0093] In the formula, Indicates the output image. This represents the inverse Fourier transform. Indicates a low-pass filter. This represents the convolution operation. It is a Fourier transform.

[0094] For minute defects, this invention defines a plane of defect structure based on 3 pixels, and designs a low-pass filter with a cutoff frequency of 110 to ensure that the signal of this smallest defect unit is not destroyed during noise reduction.

[0095] Perform an inverse Fourier transform on the filtered frequency domain image to obtain the spatial domain image after removing high-frequency noise.

[0096] The preprocessing step, while ensuring the robustness of the original image information, significantly reduces the complexity of subsequent image processing tasks. Experimental data shows that image contrast is improved by 95%, resolution is enhanced, and high-frequency noise in the image is suppressed by 73%.

[0097] Step S03: Use the adaptive Canny algorithm to enhance image edge details and identify minute defective target structures in the image.

[0098] The Canny algorithm automatically finds thresholds for strong and weak boundaries based on the image's grayscale histogram, which serves as the algorithm's input. By traversing the image, the thresholds are dynamically optimized based on the image features of different regions. High thresholds filter out pixels with extremely strong gradients and identify points with a high probability of being true edges. Low thresholds remove weak boundaries and define lower bounds for gradients at boundaries. The thresholds are entirely determined by the gradient distribution of the image itself, which is the core of the adaptive algorithm. By using the dual thresholding mechanism in the Canny algorithm to identify edges in minute structures, the details and texture structure of defective targets are enhanced.

[0099] Specifically, the Canny operator is used for image detail enhancement, such as... Figure 11 As shown:

[0100] ,

[0101] Indicates the output image. Indicates the input image. Indicates the threshold. This represents the Gaussian convolution kernel.

[0102] ,

[0103] In the formula, For low gradient threshold, For a high gradient threshold, it can be obtained using the adaptive thresholding function in the Canny algorithm. Where the gradient magnitude is greater than... The pixels with the gradient magnitude less than 0 were identified as strong edge points. Pixels with gradient values ​​between the two are preserved if they are connected to strong edge points, otherwise they are suppressed.

[0104] After image preprocessing, the Canny operator was used for detail enhancement, connecting the defective regions of the composite material. Minor defects in the boundary regions were extracted, achieving a defect detection accuracy of 99%.

[0105] like Figure 12 As shown, compared with morphological filtering, bilateral filtering, Sobel operator, and Laplacian algorithm, the adaptive method of the present invention significantly improves the contrast of the image, enhances the image details of the defect area, and improves the accuracy of defect detection.

[0106] Based on the same inventive concept, this invention also proposes an apparatus for adaptive image detection of internal defects in composite materials. The implementation of this apparatus can be found in the implementation of the method described above; details that are repeated will not be repeated. Figure 13 As shown, the device 100 includes:

[0107] Noise analysis module 101: used to acquire polarization images of composite materials using a polarization measurement device, and to perform noise analysis to identify noise in the images: random scattering noise and high-frequency speckle noise;

[0108] Denoising module 102: used to suppress random scattering noise by utilizing the difference in polarization characteristics through the IPPs method, and to remove high-frequency speckle noise in the frequency domain by using Fourier transform;

[0109] Defect recognition module 103: Used to enhance image edge details using the adaptive Canny algorithm and identify minute defect target structures in the image.

[0110] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the described module can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0111] like Figure 14As shown, the device includes a central processing unit (CPU), which can perform various appropriate actions and processes based on computer program instructions stored in read-only memory (ROM) or loaded from storage units into random access memory (RAM). The RAM can also store various programs and data required for device operation. The CPU, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0112] Multiple components in the device are connected to the I / O interface, including: input units such as keyboards and mice; output units such as various types of displays and speakers; storage units such as disks and optical discs; and communication units such as network interface cards (NICs), modems, and wireless transceivers. The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0113] The processing unit executes the various methods and processes described above, such as method steps S01 to S03. For example, in some embodiments, method steps S01 to S03 may be implemented as a computer software program tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and / or installed on the device via ROM and / or a communication unit. When the computer program is loaded into RAM and executed by the CPU, one or more steps of method steps S01 to S03 described above may be performed. Alternatively, in other embodiments, the CPU may be configured to execute method steps S01 to S03 by any other suitable means (e.g., by means of firmware).

[0114] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload programmable logic devices (CPLDs), and so on.

[0115] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0116] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0117] Furthermore, although the operations are described in a specific order, this should be understood as requiring that such operations be performed in the specific order shown or in sequential order, or requiring that all illustrated operations be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, although several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the invention. Certain features described in the context of individual embodiments may also be implemented in combination in a single implementation. Conversely, various features described in the context of a single implementation may also be implemented individually or in any suitable sub-combination in multiple implementations.

[0118] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A method for adaptive image detection of internal defects in composite materials, characterized in that, The method includes: Step S01: Use a polarization measurement device to acquire a polarization image of the composite material and perform noise analysis to identify the noise in the image: random scattering noise and high-frequency speckle noise; Step S02: Suppress random scattering noise by utilizing the difference in polarization characteristics through the IPPs method, and remove high-frequency speckle noise in the frequency domain using Fourier transform; Step S03: Use the adaptive Canny algorithm to enhance image edge details and identify minute defective target structures in the image.

2. The method for adaptive image detection of internal defects in composite materials according to claim 1, characterized in that, The noise analysis steps described in step S01 are as follows: The polarization image of the composite material is segmented into several regions; In Bonga spherical coordinates, the degree of polarization of each region under different combinations of ellipticity and azimuth is calculated using the Mueller matrix and the Stokes vector of the incident light. After further separating the noisy image from the original image, the characteristics and distribution patterns of the noise are analyzed.

3. The method for adaptive image detection of internal defects in composite materials according to claim 2, characterized in that, The formula for further separating the noisy image from the original image is as follows: ， In the formula, Represents a noisy image. Represents the original polarized image. This represents a Gaussian image.

4. The method for adaptive image detection of internal defects in composite materials according to claim 1, characterized in that, The specific steps for suppressing random scattering noise using the IPPs method based on differences in polarization characteristics, as described in step S02, are as follows: Calculate the coherence matrix of the Mueller matrix; The coherence matrix is ​​decomposed to obtain its eigenvalues ​​and submatrices; The polarization index is calculated based on eigenvalues ​​and submatrices.

5. The method for adaptive image detection of internal defects in composite materials according to claim 1, characterized in that, The specific steps for removing high-frequency speckle noise in the frequency domain using Fourier transform as described in step S02 are as follows: Perform a Fourier transform on the image to convert it from the spatial domain to the frequency domain; A low-pass filter is applied to filter the frequency domain image; Perform an inverse Fourier transform on the filtered frequency domain image to obtain the spatial domain image after removing high-frequency noise.

6. The method for adaptive image detection of internal defects in composite materials according to claim 1, characterized in that, The adaptive Canny algorithm described in step S03 automatically finds the high threshold of strong boundary images and the low threshold of weak boundary images based on the gradient distribution of the image. The high threshold is used to filter pixels with extremely strong gradients and determine points with a high probability of being true edges, while the low threshold is used to delete weak boundaries and define the lower limit of the gradient at the boundary.

7. The method for adaptive image detection of internal defects in composite materials according to claim 1, characterized in that, The identification of minute defect target structures in the image described in step S03 specifically refers to: Pixels with gradient magnitudes greater than a high threshold are identified as strong edge points, while pixels with gradient magnitudes less than a low threshold are suppressed. For pixels whose gradient magnitude is between the high and low thresholds, they are retained if they are connected to strong edge points, otherwise they are suppressed.

8. An apparatus for adaptive image detection of internal defects in composite materials, characterized in that, The device implements the method as described in any one of claims 1 to 7, comprising: Noise Analysis Module: Used to acquire polarization images of composite materials using a polarization measurement device, and to perform noise analysis to identify noise in the images: random scattering noise and high-frequency speckle noise; Denoising module: Used to suppress random scattering noise by utilizing polarization characteristic differences through the IPPs method, and to remove high-frequency speckle noise in the frequency domain using Fourier transform; Defect recognition module: Used to enhance image edge details using the adaptive Canny algorithm and identify minute defect targets in the image.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 7.

Images