An image denoising method for automatic optical inspection systems in the electronics industry

Abstract

The present invention provides an image denoising method for use in automatic optical inspection systems in the electronics industry, aiming to efficiently remove image noise and retain details. The method comprises the following steps: first, the collected image is constructed into multiple similarity group tensors by block segmentation and similarity grouping; second, the spatial and spectral features of the similarity group tensors are modeled and processed using low-rank tensor decomposition; then, the global low rank of the image is explored through the three-way logarithmic truncated tensor nuclear norm (3TLogTNN) to suppress noise and maintain global structural consistency; finally, an optimization objective function is constructed, and a non-local two-factor regularizer and a global regularizer are combined to achieve image denoising and detail retention. This method effectively combines non-local similarity with global low-rank characteristics, improves the image denoising effect and algorithm efficiency, is suitable for multispectral images and high-precision inspection scenarios, and can be widely used in AOI systems in the electronics industry.

Description

Technical Field

[0001] This invention relates to the fields of image processing and industrial intelligent manufacturing, and more particularly to an image denoising method for use in automated optical inspection (AOI) systems in the electronics industry. This method utilizes global and nonlocal low-rank tensor decomposition techniques to significantly improve the accuracy and efficiency of AOI systems by effectively suppressing noise while preserving image detail. Background Art

[0002] Automatic Optical Inspection (AOI) technology is a non-contact inspection method widely used in the electronics manufacturing industry, primarily for detecting defects in tiny components such as circuit boards and semiconductor chips. In the electronics manufacturing process, due to the small size and complex structure of components, traditional manual inspection methods have difficulty meeting the high precision and efficiency requirements. AOI systems use computer vision technology to capture images of the object under test using a camera. Then, through image processing and pattern recognition algorithms, they automatically analyze the images to identify defects. This technology can significantly improve production efficiency and quality, making it an indispensable quality control tool in the current electronics manufacturing industry.

[0003] Image processing is one of the core modules of an AOI system. This module typically includes image acquisition, image denoising, feature extraction, and image matching. The quality of image acquisition and denoising directly impacts the accuracy of subsequent inspection processes. Due to the high image quality requirements of AOI systems, the image acquisition process is often affected by external factors such as camera performance, varying lighting conditions, and electronic noise, resulting in varying degrees of noise in the image. The presence of noise can reduce image clarity, blur details, and even mask defect features, severely impacting the detection algorithm's ability to accurately identify defects. Therefore, image denoising technology has become a key step in the image processing phase of AOI systems.

[0004] Currently, common image denoising methods can be divided into two categories: global denoising and local denoising.

[0005] 1. Global denoising method

[0006] Global denoising methods typically analyze the overall structural characteristics of an image, such as its sparsity or low-rank properties, and utilize the redundant information present in the image to restore the noise-contaminated image. Global denoising methods are effective in removing large, evenly distributed noise. Common global denoising methods include Fourier transform, discrete wavelet transform, and low-rank matrix decomposition. These methods suppress noise by processing the overall structure of the image in the frequency or spatial domain. However, global denoising methods often have a weak ability to retain image details and are prone to blurred edges. Especially when processing images containing a large amount of details, some important information may be lost, thus affecting detection accuracy.

[0007] 2. Local denoising method

[0008] Local denoising methods are primarily based on neighborhood filtering algorithms, which reduce noise by calculating the average or weighted value of each pixel's neighboring area. Local denoising methods include mean filtering, median filtering, and Gaussian filtering. These methods are effective in preserving image edge details and can effectively remove high-frequency noise in local areas. However, local denoising methods are limited to processing noise within the pixel neighborhood and are less effective at removing large-area, low-frequency noise. Furthermore, local denoising methods have high computational complexity, resulting in extended runtimes when processing high-resolution images or large batches of images.

[0009] In the application of AOI systems in the electronics manufacturing industry, although traditional image denoising methods have improved image quality to a certain extent, they still have some obvious shortcomings, such as:

[0010] Non-local similarity is prevalent in natural images, meaning that similar blocks at different locations in the image are highly correlated. For example, structures such as textures and edges in one area of ​​an image often reappear in other areas. Non-local similarity can be used to enhance image denoising because it can integrate similar information from different locations to further suppress noise. However, traditional denoising methods are mostly based on the similarity of local neighborhoods and only process within a limited pixel range, failing to fully utilize the non-local information of the image. This results in limited image denoising effects, especially in detailed areas such as edges and textures.

[0011] Most commonly used image denoising methods in AOI systems treat images as two-dimensional matrices. For multi-channel RGB or multispectral images, each channel is typically processed separately, ignoring inter-channel correlations. While this approach can reduce noise simply and directly, it fails to fully utilize the complementary information between different image channels. In multispectral images, in particular, data correlations between channels are strong, and traditional methods are unable to effectively exploit these multi-channel connections, resulting in denoising results that fail to meet high-precision requirements.

[0012] In industrial production, AOI systems must process large volumes of images captured in real time. Therefore, denoising methods must not only deliver excellent results but also meet real-time requirements. However, while many advanced denoising algorithms improve image quality, they also increase computational complexity. These methods typically require significant computing resources and long processing times, making them difficult to meet the fast processing requirements of AOI systems.

[0013] In AOI applications, detail preservation is crucial, as tiny defects often reside only in subtle image structures. Excessive denoising can lead to loss of image detail, masking product defects. Both global and local denoising methods have limitations in processing image edges and details. For example, global denoising can easily cause edge blurring, while local denoising may not fully restore image details. Existing image denoising methods still need to improve the balance between detail preservation and noise suppression.

[0014] With the increasing application of multi-dimensional data such as multispectral and hyperspectral images in AOI systems, how to retain the multi-dimensional information of images while denoising has become a new challenge. Tensors are a tool for representing multi-dimensional data that can simultaneously express image information in multiple dimensions such as space, channels, and spectra. Compared with traditional matrix representations, tensor models have natural advantages in processing multi-dimensional image data. However, most current denoising methods are still based on two-dimensional matrices and fail to effectively utilize the multi-dimensional representation characteristics of tensors. As a result, the multi-dimensional structural information of the image is not fully preserved during the denoising process, thereby affecting the detection accuracy of the AOI system.

[0015] In summary, existing AOI image denoising methods have significant shortcomings in terms of exploiting non-local similarity, mining inter-channel correlations, real-time performance, detail preservation, and multi-dimensional information processing. Faced with the increasingly stringent quality inspection requirements of the electronics manufacturing industry, there is an urgent need for an improved denoising method that can effectively overcome these shortcomings in existing technologies, enhance the image denoising performance of AOI systems in practical applications, and provide clearer and more accurate image data for high-precision defect detection. Summary of the Invention

[0016] In order to solve the above problems in the prior art, the present invention proposes an image denoising method for an automatic optical inspection system in the electronics industry, which is characterized by comprising the following steps:

[0017] 1) Divide the collected images into blocks and group them by similarity, including the following sub-steps:

[0018] Dividing the acquired image into several sub-regions to form multiple three-dimensional image blocks;

[0019] Calculate the similarity between each 3D image block and other image blocks through a similarity measurement method;

[0020] According to the similarity, similar image blocks are grouped to form several similarity groups containing multiple image blocks;

[0021] Build a three-dimensional tensor for each similar group of image patches;

[0022] 2) Low-rank tensor modeling of similar blocks, including the following sub-steps:

[0023] The similarity group tensor G i Decomposed into spatial factor tensor B i and the spectral matrix A i , where i is the similarity group index currently being processed, indicating the i-th similarity group;

[0024] For the spatial factor tensor B i Processing to extract spatial features of the image;

[0025] For the spectral matrix A i Processing is performed to smooth spectral features;

[0026] Synthesize the processed spatial factor tensor B i and the spectral matrix A i , generate the denoised similarity tensor G i ;

[0027] 3) Perform global low-rank exploration on the denoised image, including the following sub-steps:

[0028] Compute multimodal low-rank properties of images;

[0029] Use global regularization methods to explore the global low-rank nature of images;

[0030] 4) Construct an image denoising model to achieve noise suppression and detail preservation by optimizing the objective function. The optimization objective function is:

[0031]

[0032] in:

[0033] Y: collected original image data;

[0034] The denoised image represents the noise-free image restored by the denoising model;

[0035] s: sparse noise term, used to compensate for the high-frequency noise remaining in the original image;

[0036] G i : The i-th similarity group tensor obtained by tensor decomposition;

[0037] B i : The spatial factor tensor of the i-th similarity group, used to extract spatial features;

[0038] A i : The spectral matrix of the i-th similarity group, used to extract spectral features;

[0039] Spectral matrix A i A smoothed version of , used to remove noise in the spectral dimension;

[0040] D3A i : spectral matrix A i The differential representation of is used to further suppress spectral noise;

[0041] ρ: regularization parameter, used to control the weight of the three-way logarithmic truncated tensor nuclear norm (3TLogTNN) regularization term;

[0042] μ: spectral smoothing regularization parameter, used to control the weight of the spectral difference regularization term;

[0043] γ: sparse regularization parameter, used to control the regularization weight of the sparse noise term S;

[0044] The square of the Frobenius norm, used to measure the reconstruction error and regularization term;

[0045] ‖·‖ 3DTLogTNN : Three-directional logarithmic truncated tensor nuclear norm, used to explore the global low rank of the image;

[0046] ‖·‖1: norm, used for sparse regularization;

[0047] n: The total number of similar groups, used to represent the number of all similar blocks that constitute the complete image.

[0048] The non-local two-factor regularizer includes the spatial difference image D1B i and D2B i constraints.

[0049] The spectrum smoothing includes the following steps: i Perform singular value decomposition.

[0050] The calculation of the three-way logarithmic truncated tensor nuclear norm (3TLogTNN) includes the TLogMNN calculation of the mode-k slice of the tensor, and its formula is as follows:

[0051]

[0052] in:

[0053] X: input tensor slice;

[0054] σ i : The i-th singular value of the input tensor slice X, arranged in descending order;

[0055] ∈: a small constant used to avoid taking logarithms of zero values, usually a very small positive number;

[0056] m: The number of retained singular values, which means performing logarithmic operations on the first m singular values.

[0057] The objective function of the image denoising model includes the following constraints:

[0058]

[0059] in:

[0060] G i : represents the low-rank tensor representation of the i-th similarity block obtained by tensor decomposition;

[0061] B i : The spatial factor tensor of the i-th similarity block, used to extract the spatial features of the image;

[0062] A i : The spectral matrix of the i-th similar block, used to extract the spectral features of the image;

[0063] ×3: represents tensor multiplication along the third mode (channel direction);

[0064] The complete image after denoising is represented by a low-rank tensor G of all n similar blocks i Reconstruction is obtained;

[0065] n: The total number of similar blocks, used to indicate the total number of similar groups formed by image blocks.

[0066] Beneficial effects:

[0067] By combining non-local similarity and global low-rank properties, an image denoising model based on low-rank tensor decomposition and regularization is proposed. This model effectively suppresses image noise while preserving the spatial details and spectral information of the image. Blocking and similarity group construction techniques are used to fully utilize the non-local similarity of the image to improve denoising accuracy. The three-way logarithmic truncated tensor kernel norm (3TLogTNN) is used to explore the global low-rank property of the image, improving the denoising effect and reducing detail loss. The optimization of the objective function and the combination of multiple regularizers further enhance the robustness and efficiency of the model, significantly improving the quality and reliability of image processing in AOI systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application, but do not constitute an improper limitation of the present invention. In the drawings:

[0069] Figure 1 : shows the overall flow chart of the image denoising method. DETAILED DESCRIPTION

[0070] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The exemplary embodiments and descriptions are only used to explain the present invention but are not intended to limit the present invention.

[0071] This embodiment provides an image denoising method suitable for automated optical inspection systems in the electronics industry. By performing block segmentation, grouping, low-rank tensor modeling, and global low-rank exploration on the acquired images, it achieves denoising effects while preserving image details. The specific implementation steps are as follows:

[0072] Step 1: Image acquisition, segmentation, and grouping

[0073] 1.1 Image Acquisition

[0074] In this embodiment, the camera in the AOI system is first used to capture an image of the target to be measured, and an RGB image or a multispectral image of the target is obtained. The captured image data is recorded as Y and contains multiple channels. For RGB images, three channels (red, green, and blue) are usually included, while multispectral images may contain more spectral channels. The data size of each channel of the image is M×N, that is, the number of pixels in the horizontal and vertical directions of the image. Since the AOI system has high requirements for image clarity and details, ambient light interference should be minimized during image acquisition, and the camera equipment should have sufficient resolution to ensure the accuracy and effectiveness of subsequent image denoising processing.

[0075] 1.2 Image Segmentation

[0076] In order to better denoise the image, in this embodiment, the collected RGB image or multispectral image is divided into blocks according to the preset patch size. Specifically, each image is divided into n three-dimensional image blocks, and the size of each image block is Where s represents the spatial size of the image block (i.e., the number of pixels contained in each block), and k represents the number of channels in the image. This block partitioning method divides the entire image into several small regions, ensuring that each block contains rich local information. This local information not only improves the resolution of image details but also provides a foundation for subsequent similarity analysis and denoising.

[0077] The advantage of block processing is that it can divide large images into smaller, more manageable blocks, allowing the algorithm to perform detailed analysis of the image's local features. Each block can be treated as a small, independent unit, facilitating block-by-block similarity calculations and tensor modeling in subsequent steps. This block-based approach also makes the denoising process more flexible, allowing optimization based on the noise characteristics of each block, thereby improving the overall denoising effect.

[0078] 1.3 Similarity Calculation and Grouping

[0079] After the block processing, the similarity between each image block and other image blocks is calculated so as to group similar image blocks. In this embodiment, the similarity is measured using the Euclidean distance as the standard. Specifically, an image block G is selected. i The Euclidean distance between the currently processed reference block and all other image blocks is calculated. Euclidean distance is a commonly used similarity metric that measures the degree of similarity between two image blocks. A smaller distance indicates a higher degree of similarity between the image blocks.

[0080] By i The Euclidean distance is used to sort the image blocks, and the other image blocks that are most similar to the current image block can be identified. The first b-1 most similar image blocks in the sorting result are selected and compared with the reference image block G i The image blocks are grouped into similarity groups, each containing b blocks, where i represents the index of the currently processed similarity group. The purpose of similarity grouping is to group blocks with similar structural or texture features together, making it easier to perform denoising based on the statistical properties of similar blocks.

[0081] This similarity grouping method fully utilizes the non-local self-similarity of images, allowing the information of multiple similar image blocks to be considered simultaneously during the denoising process, improving the denoising effect and the preservation of image details. Furthermore, this similarity group-based denoising method can effectively reduce the noise differences between image blocks and improve the stability of the processing.

[0082] 1.4 Tensor Construction

[0083] After the similarity grouping is completed, the image blocks of each similarity group are expanded along the third mode (i.e., the channel direction) to transform the three-dimensional structure of each image block into a two-dimensional matrix structure. Specifically, for each b image block in each similarity group, each image block is first expanded along the channel direction to obtain b matrices of size s×k. Then, these b matrices are stacked along the third mode to form a new tensor G of size k×s×b i , where i represents the index of the current similarity group.

[0084] Through this tensor construction process, each similarity group is transformed into a three-dimensional tensor Gi , where each dimension represents channel, space, and similarity respectively. This tensor structure can capture the characteristic information of image patches in multiple dimensions, allowing subsequent low-rank tensor modeling to simultaneously utilize the multi-channel information, spatial structure, and similarity characteristics of the image.

[0085] The advantage of tensor construction is that it preserves the multidimensional features of an image, providing a rich source of information for low-rank modeling. In particular, in multispectral image denoising, the tensor form can effectively exploit the correlations between different channels, improving denoising effectiveness. Furthermore, the resulting tensor structure is more suitable for low-rank tensor decomposition algorithms, resulting in higher accuracy and robustness in denoising, helping to preserve image detail and effectively suppress noise.

[0086] In summary, this embodiment lays the foundation for subsequent image denoising through steps such as image acquisition, segmentation, similarity calculation and grouping, and tensor construction. By leveraging non-local self-similarity and the multidimensional characteristics of tensors, this method can maintain the integrity of image details during the denoising process, thereby improving the detection accuracy of the AOI system.

[0087] Step 2: Low-rank tensor modeling of similar blocks

[0088] 2.1 Modular Three-Tensor Matrix Decomposition

[0089] In this step, in order to fully utilize the low-rank characteristics of similar blocks, the non-local similarity tensor G i Perform modular three-dimensional tensor matrix decomposition. Specifically, the tensor G i Decomposed into two parts: spatial factor tensor B i and the spectral matrix A i , thereby extracting the spatial and spectral information of the image. The decomposition process can be expressed as:

[0090] G i ≈B i ×3A i

[0091] The symbol ×3 represents the tensor multiplication operation along the third mode (i.e., the channel direction). Through this decomposition process, the spatial structure and spectral information of the image can be effectively separated, providing a basis for subsequent denoising processing. Spatial factor tensor B i Represents the spatial structural characteristics of the image block, and the spectral matrix A i This decomposition allows the spatial and spectral information of the image to be processed separately during the denoising process, thereby improving the denoising effect and the preservation of image details.

[0092] 2.2 Spatial Difference

[0093] In order to further enhance the spatial structural features of the image, in this step, the spatial factor tensor B i Perform differential operation. Specifically, B i Perform differential processing along the first mode (horizontal direction) and the second mode (vertical direction) to obtain the spatial difference image D1B i and D2B i Through the difference operation, the edge information of the image block can be effectively captured. Edge information is usually an important detail feature in the image, so retaining this information helps to enhance the quality of the denoised image and make the denoised image visually clearer.

[0094] Spatial difference image D1B i and D2B i This can highlight edges and structural changes in the image, which helps enhance the spatial consistency of the image during subsequent denoising. This difference operation not only effectively removes noise but also preserves the spatial details of the image, making the denoised image closer to the original image.

[0095] 2.3 Spectral smoothing

[0096] In this step, the spectrum matrix A i Perform singular value decomposition (SVD) and perform spectrum smoothing. Specifically, through singular value decomposition, the spectrum matrix A i Decomposed into a combination of singular values ​​and singular vectors. In order to achieve spectral smoothing, the singular values ​​are screened and only the main singular values ​​are retained, while the smaller singular values ​​are removed. The retained singular values ​​can effectively represent the main spectral information of the image, while the smaller singular values ​​usually contain noise components. Removing these noise components helps improve the denoising effect. In this way, the smoothed spectral matrix D3A is obtained. i .

[0097] This spectral smoothing process can remove noise while retaining the key part of the spectral information, thereby enhancing the consistency of the denoised image in the spectral dimension. i It can retain the spectral characteristics of the image during the denoising process, making the denoised image more realistic in visual effect and maintaining the essential characteristics of the multispectral image.

[0098] 2.4 Tensor Reconstruction

[0099] After completing the spatial difference and spectral smoothing processing, the processed data is reconstructed to obtain the denoised similarity tensor. Specifically, the spatial difference image D1B i and D2B iAnd the smoothed spectral matrix D3A i Combined together, the denoised non-local similarity tensor is generated. This reconstruction process is based on the inverse process of the above decomposition operation, integrating the differential spatial information and the processed spectral information, so that the advantages of each part are fully utilized to achieve the denoising effect.

[0100] Step 3: Exploring global low-rank properties of the image after non-local self-similarity processing

[0101] 3.1 Calculation of Three-Direction Logarithmic Truncated Tensor Nuclear Norm (3TLogTNN)

[0102] To further enhance image denoising, this example introduces a three-dimensional logarithmic truncated tensor kernel norm (3TLogTNN) method to exploit the global low-rank characteristics of the image. In natural images, global low-rank characteristics often manifest as redundancy and consistency across different modalities, which are crucial for denoising. By introducing 3TLogTNN, it is possible to maintain the overall structural information of the image while removing noise, thereby enhancing the denoising effect.

[0103] Specifically, we first perform TLogMNN calculation on the mode-k slice of the image tensor. The definition of TLogMNN is as follows:

[0104]

[0105] Among them, σ i represents the i-th singular value, ∈ is a small constant used to avoid taking the logarithm of zero values, usually a very small positive number to ensure computational stability, and m is the number of singular values. By performing logarithmic truncation on the singular values ​​of each modality, TLogMNN can effectively retain the main singular values ​​of the image, thereby suppressing noise.

[0106] After performing TLogMNN calculations on the three modalities of the image tensor, the resulting values ​​are summed up to define 3TLogTNN. The process can be described as:

[0107] 3TLogTNN=TLogMNN(X (1) )+TLogMNN(X (2) )+TLogMNN(X (3) )

[0108] Among them, X (k)represents a slice of the image tensor X at the kth mode (k = 1, 2, 3). This method, through truncation and logarithmic operations, effectively preserves image details while suppressing noise. By selectively retaining singular values ​​in each mode, 3TLogTNN ensures global image consistency during the denoising process, maintaining high visual clarity and detail integrity.

[0109] 3.2 Global Low-Rank Exploration

[0110] After completing the 3TLogTNN calculation, this step further explores the image's global low-rank properties to preserve its overall structural information. Global low-rank refers to the uniformity exhibited by an image across multiple modalities, a characteristic prevalent in natural images. By calculating the 3TLogTNN, we not only remove random noise from the image but also preserve the image's low-rank structure over a large area, ensuring that the denoising effect encompasses the image's global information.

[0111] The significance of global low-rank exploration lies in extending the denoising process to the entire image, rather than limiting it to local regions. This approach evenly removes noise across different regions, resulting in a denoised image that appears more natural in both detail and overall structure. Furthermore, the 3TLogTNN method avoids frequent tensor expansion operations, making global low-rank exploration computationally more efficient while minimizing loss of image detail.

[0112] By combining 3TLogTNN calculations with global low-rank exploration, it is possible to remove noise while preserving the global structural features of the image, achieving more efficient and accurate image denoising. This method is suitable for denoising complex data such as multispectral and hyperspectral images, providing higher-quality image data support for high-precision inspection in AOI systems.

[0113] Step 4: Build an image denoising model

[0114] In this step, the final image denoising model is constructed by combining the non-local two-factor regularizer and the three-directional logarithmic truncated tensor nuclear norm (3TLogTNN) regularizer. The optimization objective function of this model is designed as follows:

[0115]

[0116] The definitions of variables and parameters are as follows:

[0117] Y: collected original image data;

[0118] The denoised image represents the noise-free image finally restored by the model;

[0119] S: sparse noise term, used to compensate for the residual high-frequency noise components in the image, especially suitable for removing isolated outliers or noise spots;

[0120] G i =B i ×3A i : Reconstruction of the non-local similarity tensor, representing the representation of the image patch obtained by modulo three tensor decomposition, where B i is the spatial factor tensor, A i is the spectral matrix;

[0121] The reconstruction results G of all similar blocks i The combination results in a complete denoised image.

[0122] The various terms of the optimization objective function are used to control the image restoration quality and noise suppression effect during the denoising process. Specifically:

[0123] Represents the reconstruction error term between the denoised image and the original image. The purpose of this term is to ensure that the denoised image The combination of the sparse noise term S is as close to the original image Y as possible, thereby reducing the loss of image information.

[0124] The three-way logarithmic truncated tensor nuclear norm (3TLogTNN) regularization term is used to maintain the global low-rank characteristics of the image during the denoising process, thereby enhancing the smoothness and consistency of the overall image structure.

[0125] Used to constrain the spectral matrix A i and its smoothed version The difference between them is calculated to ensure that the smoothness and continuity of the spectral information are preserved during denoising.

[0126] For the spectral matrix A i The smoothing constraint term after the difference is performed to reduce the fluctuation in the spectral dimension and suppress the interference of noise on the spectral information.

[0127] γ‖S‖1: sparse regularization term, through The norm constrains the sparse noise term S, making it tend to be sparsely distributed, which is particularly suitable for suppressing isolated high-frequency noise.

[0128] The parameters ρ, μ, and γ in the optimization objective function are the weight coefficients of the regularization term, which are used to achieve a balance between noise suppression and detail preservation during the denoising process. By properly adjusting these parameters, the denoising model's sensitivity to detail and its degree of noise suppression can be controlled to achieve the best denoising effect. Through this optimization process, image data with excellent denoising effects and complete detail preservation is ultimately obtained, significantly improving the image processing quality and inspection accuracy of the AOI system.

[0129] The above description is only a preferred embodiment of the present invention. Therefore, any equivalent changes or modifications made according to the structure, characteristics and principles described in the scope of the patent application of the present invention are included in the scope of the patent application of the present invention.

Claims

1. An image denoising method for an automated optical inspection system in the electronics industry, characterized by: The following steps are involved: 1) Divide the collected images into blocks and group them by similarity, including the following sub-steps: Dividing the acquired image into several sub-regions to form multiple three-dimensional image blocks; Calculate the similarity between each 3D image block and other image blocks through a similarity measurement method; According to the similarity, similar image blocks are grouped to form several similarity groups containing multiple image blocks; Build a three-dimensional tensor for each similar group of image patches; 2) Low-rank tensor modeling of similar blocks, including the following sub-steps: The similarity group tensor G i ′ Decomposed into spatial factor tensor B i ′ and the spectral matrix A ′ i , where i is the similarity group index currently being processed, indicating the i-th similarity group; For the spatial factor tensor B i ′ Processing to extract spatial features of the image; For the spectral matrix A ′ i Processing is performed to smooth spectral features; Synthesize the processed spatial factor tensor B i and the spectral matrix A i , generate the denoised similarity tensor G i ; 3) Perform global low-rank exploration on the denoised image, including the following sub-steps: Compute multimodal low-rank properties of images; Use global regularization methods to explore the global low-rank nature of images; 4) Construct an image denoising model to achieve noise suppression and detail preservation by optimizing the objective function. The optimization objective function is: in: Y: collected original image data; Denoised image; S: sparse noise term; G i : The low-rank tensor of the i-th similarity block obtained by tensor decomposition; B i : spatial factor tensor of the i-th similarity group; A i : spectral matrix of the i-th similarity group; Spectral matrix A i A smoothed version of D3A i : spectral matrix A i The differential representation of ρ: regularization parameter; μ: spectral smoothing regularization parameter; γ: sparse regularization parameter; The square of the Frobenius norm; ‖·‖ 3DTLogTNN : three-directional logarithmic truncated tensor nuclear norm; ‖·‖1: norm; n: total number of similar groups; The calculation of the three-way logarithmic truncated tensor nuclear norm (3TLogTNN) includes the TLogMNN calculation of the mode-k slice of the tensor, and its formula is as follows: in: X: input tensor slice; σ i : the i-th singular value of the input tensor slice X; ∈: a very small positive number; m: the number of singular values ​​retained; The objective function of the image denoising model includes the following constraints: G i =B i ×3A i , in: G i : represents the low-rank tensor representation of the i-th similarity block obtained by tensor decomposition; B i : spatial factor tensor of the ith similarity block; A i : spectral matrix of the ith similarity block; ×3: represents tensor multiplication along the third mode; Denoised image; n: The total number of similar blocks.

2. The image denoising method for an automated optical inspection system in the electronics industry according to claim 1, wherein: The non-local two-factor regularizer includes the spatial difference image D1B i and D2B i constraints.

3. The image denoising method for an automated optical inspection system in the electronics industry according to claim 1, wherein: The spectrum smoothing includes the following steps: i Perform singular value decomposition.

Images