Laser processing substrate production process

Abstract

The present application relates to the field of carrier board production, and specifically to a laser processing carrier board production process. The process comprises: providing a substrate, and cleaning the substrate to obtain a cleaned substrate; coating a layer of photosensitive material on the surface of the cleaned substrate and drying the photosensitive material to obtain a substrate coated with the photosensitive material; using a laser device to expose the substrate coated with the photosensitive material and remove the unexposed portion to obtain an exposed and developed substrate; placing the exposed and developed substrate into an etching solution to remove the portion not protected by the photosensitive material to form a desired circuit pattern to obtain a carrier board; and performing an appearance inspection on the carrier board to obtain an appearance inspection result, wherein the appearance inspection result is used to indicate whether there is a defect in the appearance of the carrier board.

Description

Technical Field

[0001] The present application relates to the field of carrier board production, and specifically to a laser processing carrier board production process. Background Art

[0002] In modern manufacturing, especially in high-end electronics and semiconductor industries, substrates are key components that connect and support electronic components. Their quality and reliability directly affect the overall performance of the product. With the continuous advancement of technology, the design of substrates has become increasingly complex, and higher requirements have been placed on the production process and quality control of substrates.

[0003] Laser Technology, as an advanced processing method, has been widely used in the field of substrate production due to its high precision, high repeatability, low thermal impact and non-contact characteristics. Through the laser beam of the laser, the laser process can realize various process flows such as drilling, cutting, and scratching of substrate materials, greatly improving production efficiency and product quality. However, during the laser processing process, the substrate may still have appearance defects such as scratches, stains or errors in etching circuit patterns due to various reasons such as material properties, processing parameters, and environmental factors. If these defects are not discovered and handled in time, they will seriously affect the overall performance of the product.

[0004] The traditional quality inspection method of laser processed substrates mainly relies on manual visual inspection, which is easily affected by subjective factors such as fatigue and vision differences, resulting in unstable and error-prone inspection results, and facing the problems of low inspection efficiency and high labor costs.

[0005] Therefore, in order to overcome the limitations of traditional inspection methods and improve the accuracy and efficiency of substrate appearance inspection, an optimized laser processing substrate production process is expected. Summary of the invention

[0006] This application is made in view of the above problems. One object of this application is to provide a laser processing carrier board production process.

[0007] The embodiment of the present application provides a laser processing substrate production process, which includes:

[0008] Providing a substrate, and cleaning the substrate to obtain a cleaned substrate;

[0009] Coating a layer of photosensitive material on the surface of the cleaned substrate and drying the photosensitive material to obtain a substrate coated with the photosensitive material;

[0010] Using a laser device to expose the substrate coated with the photosensitive material and removing the unexposed portion to obtain an exposed and developed substrate;

[0011] placing the exposed and developed substrate into an etching solution to remove the portion not protected by the photosensitive material to form a desired circuit pattern to obtain a carrier;

[0012] Performing appearance inspection on the carrier board to obtain an appearance inspection result, wherein the appearance inspection result is used to indicate whether there is any defect in the appearance of the carrier board.

[0013] For example, according to the laser processing carrier board production process of the embodiment of the present application, the carrier board is subjected to appearance inspection to obtain the appearance inspection result, including:

[0014] Acquire an image of the appearance of the carrier board captured by a camera;

[0015] Extracting high-frequency components of the carrier appearance image to obtain a time-domain image of high-frequency components of the carrier appearance;

[0016] Performing image feature extraction on the carrier appearance high-frequency component time domain image to obtain a carrier appearance high-frequency component time domain feature image;

[0017] Performing feature enhancement processing on the time-domain feature map of the high-frequency component of the carrier appearance to obtain a time-domain significant feature map of the high-frequency component of the carrier appearance;

[0018] The appearance detection result is generated based on the time-domain significant feature map of the high-frequency component of the carrier appearance.

[0019] For example, according to the laser processing carrier production process of the embodiment of the present application, extracting the high-frequency component of the carrier appearance image to obtain the carrier appearance high-frequency component time domain diagram includes:

[0020] Performing a fast Fourier transform on the carrier appearance image to obtain a carrier appearance frequency domain image;

[0021] The high-frequency components in the carrier appearance frequency domain image are extracted and the high-frequency components are subjected to inverse Fourier transformation to obtain a carrier appearance high-frequency component time domain image.

[0022] For example, according to the laser processing carrier production process of the embodiment of the present application, the image feature extraction is performed on the carrier appearance high-frequency component time domain image to obtain the carrier appearance high-frequency component time domain feature image, including:

[0023] The time domain image of the high-frequency component of the carrier appearance is input into an image feature extractor based on a hole convolutional neural network model to obtain the time domain feature image of the high-frequency component of the carrier appearance.

[0024] For example, according to the laser processing carrier production process of the embodiment of the present application, the feature enhancement processing is performed on the carrier appearance high-frequency component time domain feature map to obtain the carrier appearance high-frequency component time domain significant feature map, including:

[0025] The temporal feature map of the high-frequency component of the carrier appearance is input into a gradient saliency mask feature enhancer to obtain the temporal saliency feature map of the high-frequency component of the carrier appearance.

[0026] For example, according to the laser processing carrier production process of the embodiment of the present application, the time domain feature map of the high-frequency component of the carrier appearance is input into the gradient saliency mask feature enhancer to obtain the time domain saliency feature map of the high-frequency component of the carrier appearance, including:

[0027] Calculating the multi-directional gradient value distribution of each position in the time-domain feature map of the high-frequency component of the carrier appearance, and determining the gradient amplitude value of each position in the time-domain feature map of the high-frequency component of the carrier appearance based on the multi-directional gradient value distribution of each position to obtain a gradient amplitude distribution map of the time-domain feature of the high-frequency component of the carrier appearance;

[0028] Calculating the local description operator of the gradient amplitude at each position in the time-domain characteristic gradient amplitude distribution map of the high-frequency component of the carrier appearance to obtain a local significant distribution map of the time-domain characteristic gradient amplitude of the high-frequency component of the carrier appearance;

[0029] Inputting the local significant distribution map of the time-domain characteristic gradient amplitude of the high-frequency component of the carrier appearance into a gated masker based on the GELU function to obtain a local significant gated mask map of the gradient amplitude;

[0030] The position point multiplication between the gradient amplitude local significant gated mask map and the carrier appearance high-frequency component time-domain feature map is calculated to obtain the carrier appearance high-frequency component time-domain significant feature map.

[0031] For example, according to the laser processing carrier production process of the embodiment of the present application, the local description operator of the gradient amplitude at each position in the time domain characteristic gradient amplitude distribution map of the carrier appearance high-frequency component is calculated to obtain the local significant distribution map of the time domain characteristic gradient amplitude of the carrier appearance high-frequency component, including:

[0032] The scale of the local neighborhood is determined, and the mean value of the difference between the gradient amplitude value at a predetermined position in the time domain characteristic gradient amplitude distribution diagram of the high-frequency component of the carrier appearance and the gradient amplitude values ​​at other positions in the local neighborhood is calculated to obtain a local description operator of the gradient amplitude corresponding to the predetermined position.

[0033] For example, according to the laser processing carrier board production process of the embodiment of the present application, the appearance detection result is generated based on the time domain significant feature map of the high-frequency component of the appearance of the carrier board, including:

[0034] The time-domain significant feature map of the high-frequency component of the carrier appearance is input into a classifier-based appearance detection generator to obtain the appearance detection result.

[0035] For example, according to the laser processing carrier production process of the embodiment of the present application, the time domain significant feature map of the high-frequency component of the carrier appearance is input into the classifier-based appearance detection generator to obtain the appearance detection result, including:

[0036] Expanding the time-domain significant feature map of the high-frequency component of the carrier appearance into a classification feature vector according to a row vector or a column vector;

[0037] Performing full-connection encoding on the classification feature vector using the fully-connected layer of the classifier-based appearance detection generator to obtain an encoded classification feature vector;

[0038] The encoded classification feature vector is input into the Softmax classification function of the classifier-based appearance detection generator to obtain the appearance detection result.

[0039] According to the laser processing carrier board production process of the embodiment of the present application, it uses computer vision technology based on deep learning to perform appearance inspection on the laser processing carrier board, extracts the high-frequency component feature representation of the carrier board appearance image, and performs a saliency analysis based on the feature gradient to capture the key feature area in the carrier board appearance, thereby intelligently identifying defects in the carrier board appearance. In this way, the accuracy and efficiency of carrier board inspection can be effectively improved, labor costs can be reduced, and the quality and reliability of the carrier board can be ensured. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application are briefly introduced below. Obviously, the drawings described below only relate to some embodiments of the present application, and are not intended to limit the present application.

[0041] Figure 1 A flow chart showing a laser processing carrier board production process in an embodiment of the present application is shown;

[0042] Figure 2 A flowchart of sub-step S550 of the laser processing carrier board production process in an embodiment of the present application is shown;

[0043] Figure 3 A flowchart of sub-step S554 of the laser processing carrier board production process in an embodiment of the present application is shown;

[0044] Figure 4 A schematic diagram of the structure of a laser processing carrier board production system in an embodiment of the present application is shown; and

[0045] Figure 5 The application scenario diagram of the laser processing carrier board production process in the embodiment of the present application is shown. DETAILED DESCRIPTION

[0046] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.

[0047] The terms used in this specification are those common terms currently widely used in the art in consideration of the functions of the present application, but these terms may vary according to the intentions of those skilled in the art, precedents, or new technologies in the art. In addition, specific terms may be selected, and in this case, their detailed meanings will be described in the detailed description of the present application. Therefore, the terms used in the specification should not be understood as simple names, but rather as a general description based on the meaning of the terms and the present application.

[0048] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0049] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. At the same time, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0050] In addition, the application architecture diagram in the embodiment of the present application is intended to more clearly illustrate the technical solution in the embodiment of the present application, and does not constitute a limitation on the technical solution provided in the embodiment of the present application. Of course, for other application architectures and business applications, the technical solution provided in the embodiment of the present application is also applicable to similar problems.

[0051] The following is a non-restrictive description of the laser processing carrier production process provided according to at least one embodiment of the present application through several examples or embodiments. As described below, different features in these specific examples or embodiments can be combined with each other without conflicting with each other to obtain new examples or embodiments, and these new examples or embodiments also fall within the scope of protection of the present application.

[0052] The present application provides a laser processing substrate production process, such as Figure 1As shown, it includes: S510, providing a substrate, and cleaning the substrate to obtain a cleaned substrate; S520, coating a layer of photosensitive material on the surface of the cleaned substrate and drying the photosensitive material to obtain a substrate coated with the photosensitive material; S530, using a laser device to expose the substrate coated with the photosensitive material and remove the unexposed portion to obtain an exposed and developed substrate; S540, placing the exposed and developed substrate into an etching solution to remove the portion not protected by the photosensitive material to form a desired circuit pattern to obtain a carrier; S550, performing an appearance inspection on the carrier to obtain an appearance inspection result, wherein the appearance inspection result is used to indicate whether there are defects in the appearance of the carrier.

[0053] It should be understood that the laser processing substrate production process is a method for manufacturing electronic circuit boards, which involves multiple steps, each of which has its specific purpose and function. Step S510 is used to provide the basic material for manufacturing circuit boards, usually a copper-clad plate. The cleaning process is to remove impurities, oil stains, etc. on the surface of the substrate to ensure that the subsequent photosensitive material can be evenly coated and adhered to the substrate. In step S520, the photosensitive material is coated to form a protective layer on the substrate that can be irradiated by light and chemically react. The drying is to remove the solvent in the photosensitive material so that it is fixed on the substrate and prepared for the next exposure. In step S530, laser exposure is to use laser to accurately irradiate the substrate so that the photosensitive material reacts chemically in a specific area to form a circuit pattern. The removal of the unexposed part is completed through the development process, so that only the circuit pattern that needs to be retained is left. Among them, laser exposure is a process of irradiating a specific material with a laser beam to achieve a change in the surface or internal structure of the material. Laser exposure is a key technology in the field of electronic manufacturing, especially in the production of printed circuit boards (PCBs). Laser light sources are highly monochromatic and coherent, and can produce very fine spots, which allows laser exposure to define the size and shape of circuit patterns very accurately. Compared with traditional photolithography, laser exposure can quickly adjust the exposure pattern by changing the path of the laser beam, thereby achieving the production of different designs without changing the mask, and laser exposure can produce very fine lines and patterns, which is very important for the manufacture of high-density circuit boards. Laser exposure is a non-contact processing process that does not cause physical pressure or damage to the material, which is crucial to maintaining the integrity of the substrate and the reliability of the circuit. Moreover, the laser exposure process can be fully automated, and the movement of the laser beam and the exposure parameters are controlled by a computer to achieve efficient and consistent production. In step S540, etching is to remove the portion of the copper layer that is not protected by the photosensitive material by chemical or electrochemical methods, thereby forming the desired circuit pattern on the substrate. This step is key to forming the functional layer of the circuit board. In step S550, the appearance inspection is to check whether the circuit board has defects, such as broken wires, short circuits, unclear patterns, etc. This step ensures the quality and reliability of the circuit board and is a quality control link in the production process. The entire process is highly automated and precisely controlled to ensure that the quality and performance of the circuit board meet the requirements of electronic equipment.

[0054] In view of the limitations of traditional appearance inspection methods, this application uses computer vision technology based on deep learning to perform appearance inspection on laser-processed substrates. By extracting the high-frequency component feature representation of the substrate appearance image and performing a saliency analysis based on feature gradients, the key feature areas in the substrate appearance are captured, thereby intelligently identifying defects in the substrate appearance. In this way, the accuracy and efficiency of substrate inspection can be effectively improved, labor costs can be reduced, and the quality and reliability of the substrate can be ensured.

[0055] Accordingly, if Figure 2 As shown, the appearance inspection of the carrier is performed to obtain the appearance inspection result, including: S551, obtaining a carrier appearance image of the carrier collected by a camera; S552, extracting the high-frequency component of the carrier appearance image to obtain a carrier appearance high-frequency component time domain map; S553, performing image feature extraction on the carrier appearance high-frequency component time domain map to obtain a carrier appearance high-frequency component time domain feature map; S554, performing feature enhancement processing on the carrier appearance high-frequency component time domain feature map to obtain a carrier appearance high-frequency component time domain significant feature map; S555, generating the appearance inspection result based on the carrier appearance high-frequency component time domain significant feature map.

[0056] Specifically, in the technical solution of the present application, first, an image of the carrier appearance of the carrier captured by a camera is obtained. It should be understood that the camera can directly capture the physical appearance of the carrier, including its shape, size, color, texture and any possible defects (such as scratches, cracks, stains, deformation, etc.), providing an intuitive information source for the appearance inspection of the carrier.

[0057] Next, considering that in the carrier appearance image, features such as texture, edge and details that are closely related to the appearance defects of the carrier usually correspond to the high-frequency components of the image. Therefore, in order to effectively extract the high-frequency component features of the carrier appearance image to reduce the interference of background information on defect detection, thereby improving the accuracy and robustness of defect detection, in the technical solution of the present application, firstly, a fast Fourier transform is performed on the carrier appearance image, and the carrier appearance image is converted from the spatial domain to the frequency domain, decomposed into components of different frequencies, and the details and texture information in the carrier appearance image are more intuitively displayed to obtain a carrier appearance frequency domain map.

[0058] Then, the high-frequency components in the carrier appearance frequency domain image are further extracted, and the high-frequency components are subjected to inverse Fourier transform to convert the high-frequency components from the frequency domain to the time domain, and restore the original image form, thereby obtaining a carrier appearance high-frequency component time domain image. Since the low-frequency components are filtered, the carrier appearance high-frequency component time domain image mainly contains the edge, texture and other detail features of the original image, thereby providing clearer and more significant visual information for subsequent defect detection. In an embodiment of the present application, a high-pass filter is used to extract high-frequency components from the carrier appearance frequency domain image.

[0059] Correspondingly, in step S552, the high-frequency components of the carrier appearance image are extracted to obtain a time domain map of the high-frequency components of the carrier appearance, including: performing a fast Fourier transform on the carrier appearance image to obtain a frequency domain map of the carrier appearance; extracting the high-frequency components in the frequency domain map of the carrier appearance and performing an inverse Fourier transform on the high-frequency components to obtain a time domain map of the high-frequency components of the carrier appearance.

[0060] Furthermore, in order to further extract deeper features of the carrier appearance, the present application adopts a dilated convolutional neural network (DCNN) model as an image feature extractor to perform feature extraction on the time domain map of the high-frequency component of the carrier appearance. By utilizing the characteristics of the dilated convolution, the receptive field is effectively expanded without increasing the number of parameters to capture a wider range of image context information while maintaining a high spatial resolution, thereby effectively mining features such as the edges, textures, and shape contours of the circuit pattern on the carrier to obtain the time domain feature map of the high-frequency component of the carrier appearance.

[0061] Correspondingly, in step S553, image feature extraction is performed on the time domain map of the high-frequency component of the carrier appearance to obtain the time domain feature map of the high-frequency component of the carrier appearance, including: inputting the time domain map of the high-frequency component of the carrier appearance into an image feature extractor based on a void convolutional neural network model to obtain the time domain feature map of the high-frequency component of the carrier appearance.

[0062] Secondly, in order to further highlight the key feature areas in the time-domain feature map of the high-frequency component of the carrier appearance, the present application introduces a gradient significance mask feature enhancer to perform feature significance analysis and feature enhancement processing on the time-domain feature map of the high-frequency component of the carrier appearance, so as to improve the contrast of the key feature areas. Specifically, the gradient significance mask feature enhancer first quantifies the intensity of local pixel changes at each position by calculating the gradient amplitude value at each position in the time-domain feature map of the high-frequency component of the carrier appearance. Then, the gradient amplitude local description operator is calculated by measuring the significance of the gradient amplitude value at each position relative to the gradient amplitude values ​​at other positions in the local neighborhood, and the feature importance of each position is evaluated. Then, the GELU function is used to perform nonlinear activation on the gradient amplitude local description operator at each position to generate a gated mask to guide the enhancement of key features in the time-domain feature map of the high-frequency component of the carrier appearance. Finally, based on the generated gated mask, the temporal feature map of the high-frequency component of the carrier appearance is weighted element by element, thereby effectively enhancing the key feature areas in the temporal feature map of the high-frequency component of the carrier appearance, while suppressing noise interference, so as to obtain the temporal significant feature map of the high-frequency component of the carrier appearance.

[0063] Correspondingly, in step S554, feature enhancement processing is performed on the time domain feature map of the high-frequency component of the carrier appearance to obtain a time domain significant feature map of the high-frequency component of the carrier appearance, including: inputting the time domain feature map of the high-frequency component of the carrier appearance into a gradient significance mask feature enhancer to obtain the time domain significant feature map of the high-frequency component of the carrier appearance.

[0064] Specifically, Figure 3 As shown, the time domain feature map of the high-frequency component of the carrier appearance is input into the gradient saliency mask feature enhancer to obtain the time domain saliency feature map of the high-frequency component of the carrier appearance, including: S5541, calculating the multi-directional gradient value distribution of each position in the time domain feature map of the high-frequency component of the carrier appearance, and determining the gradient amplitude value of each position in the time domain feature map of the high-frequency component of the carrier appearance based on the multi-directional gradient value distribution of each position to obtain the gradient amplitude distribution map of the time domain feature of the high-frequency component of the carrier appearance; S5542, calculating the time domain feature map of the high-frequency component of the carrier appearance A local description operator of the gradient amplitude at each position in the gradient amplitude distribution map is used to obtain a local significant distribution map of the gradient amplitude of the time domain feature of the high-frequency component of the carrier appearance; S5543, inputting the local significant distribution map of the gradient amplitude of the time domain feature of the high-frequency component of the carrier appearance into a gated masker based on the GELU function to obtain a gradient amplitude local significant gated mask map; S5544, calculating the position point multiplication between the gradient amplitude local significant gated mask map and the time domain feature map of the high-frequency component of the carrier appearance to obtain the time domain significant feature map of the high-frequency component of the carrier appearance.

[0065] Among them, in step S5542, the local description operator of the gradient amplitude of each position in the gradient amplitude distribution map of the time-domain characteristics of the high-frequency component of the carrier appearance is calculated to obtain the local significant distribution map of the gradient amplitude of the time-domain characteristics of the high-frequency component of the carrier appearance, including: determining the scale of the local neighborhood, calculating the mean of the difference between the gradient amplitude value of the predetermined position in the gradient amplitude distribution map of the time-domain characteristics of the high-frequency component of the carrier appearance and the gradient amplitude values ​​of other positions in the local neighborhood to obtain the gradient amplitude local description operator corresponding to the predetermined position.

[0066] In a specific example, the time-domain feature map of the high-frequency component of the carrier appearance is input into a gradient saliency mask feature enhancer to obtain the time-domain saliency feature map of the high-frequency component of the carrier appearance, including: processing the time-domain feature map of the high-frequency component of the carrier appearance with the following feature distribution gradient mask formula to obtain the time-domain saliency feature map of the high-frequency component of the carrier appearance, wherein the feature distribution gradient mask formula is:

[0067] in, , , , , and Respectively represent the grayscale values ​​of the corresponding positions of the time domain feature map of the high-frequency components of the carrier appearance, The time domain characteristic diagram of the high frequency component of the appearance of the carrier board is shown in The gradient value in the horizontal direction of the pixel point. The time domain characteristic diagram of the high frequency component of the appearance of the carrier board is shown in The gradient value in the ordinate direction of the pixel point. The time domain characteristic diagram of the high frequency component of the appearance of the carrier board is shown in The channel direction gradient value of the position pixel point, The first time domain characteristic diagram of the high frequency component of the appearance of the carrier board is represented by The gradient magnitude value at the position, The time domain characteristic gradient amplitude distribution diagram of the high frequency component of the carrier appearance is represented by The set of gradient magnitude values ​​in the local neighborhood centered at position, represents the number of gradient magnitude values ​​in the set of gradient magnitude values ​​within the local neighborhood, , and Respectively represent the offset in the horizontal axis direction, the vertical axis direction and the channel direction, The first time domain characteristic diagram of the high frequency component of the appearance of the carrier board is represented by The local description operator of the gradient magnitude of the position, represents the GELU function, The first part of the gradient magnitude local significant gated mask map is represented by The local description operator of the gradient magnitude of the position, represents the local significant gated mask map of the gradient magnitude, It means point multiplication by position. A time domain significant feature diagram of the high-frequency components of the carrier appearance is shown.

[0068] Next, the time-domain significant feature map of the high-frequency component of the carrier appearance is input into the classifier-based appearance detection generator to obtain the appearance detection result. Accordingly, in step S555, the appearance detection result is generated based on the time-domain significant feature map of the high-frequency component of the carrier appearance, including: inputting the time-domain significant feature map of the high-frequency component of the carrier appearance into the classifier-based appearance detection generator to obtain the appearance detection result. Specifically, inputting the time-domain significant feature map of the high-frequency component of the carrier appearance into the classifier-based appearance detection generator to obtain the appearance detection result includes: expanding the time-domain significant feature map of the high-frequency component of the carrier appearance into a classification feature vector according to a row vector or a column vector; using the fully connected layer of the classifier-based appearance detection generator to fully connect the classification feature vector to obtain an encoded classification feature vector; inputting the encoded classification feature vector into the Softmax classification function of the classifier-based appearance detection generator to obtain the appearance detection result.

[0069] It should be understood that the role of the classifier is to use the given categories and known training data to learn classification rules and classifiers, and then classify (or predict) unknown data. Logistic regression, SVM, etc. are often used to solve binary classification problems. For multi-class classification problems, logistic regression or SVM can also be used, but multiple binary classifications are required to form a multi-classification, but this is prone to errors and inefficient. Commonly used multi-classification methods include the Softmax classification function.

[0070] In a preferred example, the time-domain significant feature map of the high-frequency component of the carrier appearance is input into a classifier-based appearance detection generator to obtain the appearance detection result, including the steps of: optimizing the time-domain significant feature map of the high-frequency component of the carrier appearance to obtain an optimized time-domain significant feature map of the high-frequency component of the carrier appearance; inputting the optimized time-domain significant feature map of the high-frequency component of the carrier appearance into the classifier-based appearance detection generator to obtain the appearance detection result. The method of optimizing the time-domain significant feature map of the high-frequency component of the carrier appearance to obtain an optimized time-domain significant feature map of the high-frequency component of the carrier appearance comprises the steps of: clustering all eigenvalues ​​of the time-domain significant feature map of the high-frequency component of the carrier appearance based on the L2 distance between the eigenvalues, and arranging the clustering features into a time-domain significant clustering vector of the high-frequency component of the carrier appearance; determining a clustering ratio value of the number of eigenvalues ​​of the time-domain significant clustering vector of the high-frequency component of the carrier appearance and the number of eigenvalues ​​of the time-domain significant feature map of the high-frequency component of the carrier appearance; dividing the second norm of the time-domain significant clustering vector of the high-frequency component of the carrier appearance by the second norm of the time-domain significant feature vector of the high-frequency component of the carrier appearance obtained by expanding the time-domain significant feature map of the high-frequency component of the carrier appearance to obtain a time-domain significant conflict representation value of the high-frequency component of the carrier appearance; dividing the first power value of the first norm of the time-domain significant clustering vector of the high-frequency component of the carrier appearance with the clustering ratio value as an exponent by the first power value of the first norm of the time-domain significant clustering vector of the high-frequency component of the carrier appearance The norm is raised to the second power value with the clustering ratio value as the exponent to obtain the time-domain significant confrontation representation value of the carrier appearance high-frequency component; for each eigenvalue of the carrier appearance high-frequency component time-domain significant clustering vector, it is multiplied by the reciprocal of the difference between the carrier appearance high-frequency component time-domain significant conflict representation value and the carrier appearance high-frequency component time-domain significant confrontation representation value to obtain the optimized eigenvalue of the carrier appearance high-frequency component time-domain significant clustering vector; for each eigenvalue outside the cluster in the carrier appearance high-frequency component time-domain significant feature map, it is multiplied by the reciprocal of the sum of the carrier appearance high-frequency component time-domain significant conflict representation value and the carrier appearance high-frequency component time-domain significant confrontation representation value to obtain the optimized out-of-class eigenvalue of the carrier appearance high-frequency component time-domain significant feature map; the optimized eigenvalue of the carrier appearance high-frequency component time-domain significant clustering vector and the optimized out-of-class eigenvalue of the carrier appearance high-frequency component time-domain significant feature map are combined to form the optimized carrier appearance high-frequency component time-domain significant feature map.

[0071] In one example, optimizing the time-domain significant feature map of the high-frequency component of the carrier appearance to obtain an optimized time-domain significant feature map of the high-frequency component of the carrier appearance includes: optimizing the time-domain significant feature map of the high-frequency component of the carrier appearance using the following optimization formula to obtain the optimized time-domain significant feature map of the high-frequency component of the carrier appearance; wherein the optimization formula is expressed as: , in, is the time domain significant feature vector of the high-frequency component of the carrier appearance, is the number of eigenvalues ​​of the time-domain significant eigenvector of the high-frequency component of the carrier appearance, is the time-domain significant clustering vector of the high-frequency components of the carrier appearance, is the number of eigenvalues ​​of the time-domain significant clustering vector of the high-frequency components of the carrier appearance, represents the clustering feature set corresponding to the time-domain significant clustering vector of the high-frequency component of the carrier appearance, is the clustering ratio value, and Represents the binary norm and the uninorm of the vector respectively. Power, is the characteristic value of each position in the time domain significant characteristic map of the high-frequency component of the carrier appearance, It is the characteristic value of each position in the time domain significant feature map of the high-frequency component of the optimized carrier appearance.

[0072] That is, considering that the time domain feature map of the high-frequency component of the carrier appearance represents the image semantic features of the time domain map of the high-frequency component of the carrier appearance obtained by inverse Fourier transform of the high-frequency components in the frequency domain map of the carrier appearance, when further gradient saliency mask feature enhancement is performed, image semantic aggregation alignment conflicts will be caused due to the difference in reinforcement of the local image semantic feature distribution, thereby causing the loss of aggregation key information and affecting the expression effect of the time domain saliency feature map of the high-frequency component of the carrier appearance.

[0073] Therefore, in order to avoid the loss of key suffix semantic information relative to the original feature set as a whole due to aggregation conflict caused by the aggregation feature based on the temporal significant feature map of the high-frequency component of the carrier appearance, the adversarial decision of the absolute representation of the set of the high-frequency component of the carrier appearance and the one-norm of the temporal significant feature vector of the high-frequency component of the carrier appearance is performed by taking the clustering ratio of the number of eigenvalues ​​of the temporal significant clustering vector of the high-frequency component of the carrier appearance relative to the number of eigenvalues ​​of the temporal significant feature map of the high-frequency component of the carrier appearance as the decision function, and respectively comparing the clustering ratio of the number of eigenvalues ​​of the temporal significant clustering vector of the high-frequency component of the carrier appearance and the temporal significant feature vector of the high-frequency component of the carrier appearance with the temporal significant clustering vector of the high-frequency component of the carrier appearance. The clustering intrinsic conflict representation of the binary norm of the time-domain significant feature vector of the high-frequency component of the board appearance is expressed through positive and negative interactions to construct a solid alignment guardrail with the original feature set as a whole based on the aggregated features of the optimized time-domain significant feature map of the high-frequency component of the carrier appearance, thereby achieving the intentional mitigation of harmful information loss based on the aggregated risk transferability of the optimized time-domain significant feature map of the high-frequency component of the carrier appearance, improving the expression effect of the optimized time-domain significant feature map of the high-frequency component of the carrier appearance, thereby improving the accuracy of the appearance detection result obtained by inputting the time-domain significant feature map of the high-frequency component of the carrier appearance into the classifier-based appearance detection generator.

[0074] Based on the above embodiments, see Figure 4 The figure is a schematic diagram of the structure of a laser processing carrier production system 800 in an embodiment of the present application. The laser processing carrier production system 800 includes: a cleaning module 810, which is used to provide a substrate and clean the substrate to obtain a substrate after cleaning; a drying module 820, which is used to coat a layer of photosensitive material on the surface of the substrate after cleaning and dry the photosensitive material to obtain a substrate coated with photosensitive material; an exposure module 830, which is used to use a laser device to expose the substrate coated with photosensitive material and remove the unexposed part to obtain an exposed and developed substrate; an etching module 840, which is used to put the exposed and developed substrate into an etching solution to remove the part not protected by the photosensitive material to form a desired circuit pattern to obtain a carrier; an appearance inspection module 850, which is used to perform an appearance inspection on the carrier to obtain an appearance inspection result, and the appearance inspection result is used to indicate whether there is a defect in the appearance of the carrier.

[0075] Here, those skilled in the art will appreciate that the specific functions and operations of each module in the laser processing substrate production system 800 have been described in detail above. Figures 1 to 3 The laser processing carrier production process has been described in detail, and therefore, its repeated description will be omitted.

[0076] Figure 5 FIG. 1 is an application scenario diagram of the laser processing substrate production process according to an embodiment of the present application. Figure 5 As shown, in this application scenario, first, an image of the appearance of the carrier board captured by a camera is obtained (for example, Figure 5 D), and then, the carrier appearance image is input to a server (for example, Figure 5 In S) as shown in , the server can use the laser processing carrier board production algorithm to process the carrier board appearance image to obtain an appearance inspection result indicating whether there are defects in the carrier board appearance.

[0077] Based on the above embodiments, another exemplary embodiment of an electronic device is also provided in the embodiments of the present application. In some possible implementations, the electronic device in the embodiments of the present application may include a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the steps of the laser processing substrate production process in the above embodiments can be implemented when the processor executes the program.

[0078] The embodiment of the present application also provides a computer-readable storage medium, on which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, the laser processing carrier production process according to the embodiment of the present application described with reference to the above figures can be executed. The computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. The volatile memory may, for example, include a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may, for example, include a read-only memory (ROM), a hard disk, a flash memory, etc.

[0079] The embodiment of the present application also provides a computer program product or a computer program, which includes computer executable instructions, and the computer executable instructions are stored in a computer readable storage medium. The processor of the computer device reads the computer executable instructions from the computer readable storage medium, and the processor executes the computer executable instructions, so that the computer device performs the laser processing carrier production process according to the embodiment of the present application.

[0080] Those skilled in the art will appreciate that the contents disclosed in this application may be subject to various variations and improvements. For example, the various devices or components described above may be implemented by hardware, or by software, firmware, or a combination of some or all of the three.

[0081] In addition, although the present application makes various references to certain units in the system according to embodiments of the present application, any number of different units can be used and run on the client and / or server. The units are only illustrative, and different aspects of the system and method can use different units.

[0082] Those skilled in the art will appreciate that all or part of the steps in the above method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, such as a read-only memory, a disk or an optical disk. Optionally, all or part of the steps in the above embodiment can also be implemented using one or more integrated circuits. Accordingly, each module / unit in the above embodiment can be implemented in the form of hardware or in the form of software function modules. The present application is not limited to any particular form of combination of hardware and software.

[0083] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art to which this application belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined as such herein.

[0084] The above is an explanation of the present application and should not be considered as limiting thereof. Although several exemplary embodiments of the present application have been described, those skilled in the art will readily appreciate that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present application.

Claims

1. A laser processing carrier board production process, characterized in that: include: Providing a substrate, and cleaning the substrate to obtain a cleaned substrate; Coating a layer of photosensitive material on the surface of the cleaned substrate and drying the photosensitive material to obtain a substrate coated with the photosensitive material; Using a laser device to expose the substrate coated with the photosensitive material and removing the unexposed portion to obtain an exposed and developed substrate; placing the exposed and developed substrate into an etching solution to remove the portion not protected by the photosensitive material to form a desired circuit pattern to obtain a carrier; Performing appearance inspection on the carrier board to obtain an appearance inspection result, wherein the appearance inspection result is used to indicate whether there is a defect in the appearance of the carrier board; Performing appearance inspection on the carrier board to obtain an appearance inspection result includes: Acquire an image of the appearance of the carrier board captured by a camera; Extracting high-frequency components of the carrier appearance image to obtain a time-domain image of high-frequency components of the carrier appearance; Performing image feature extraction on the carrier appearance high-frequency component time domain image to obtain a carrier appearance high-frequency component time domain feature image; Performing feature enhancement processing on the time-domain feature map of the high-frequency component of the carrier appearance to obtain a time-domain significant feature map of the high-frequency component of the carrier appearance; Generating the appearance detection result based on the time-domain significant feature map of the high-frequency component of the carrier appearance; Performing feature enhancement processing on the time domain feature map of the high-frequency component of the carrier appearance to obtain a time domain significant feature map of the high-frequency component of the carrier appearance, including: Inputting the temporal feature map of the high-frequency component of the carrier appearance into a gradient saliency mask feature enhancer to obtain the temporal saliency feature map of the high-frequency component of the carrier appearance; Inputting the temporal feature map of the high-frequency component of the carrier appearance into a gradient saliency mask feature enhancer to obtain the temporal saliency feature map of the high-frequency component of the carrier appearance, comprising: Calculating the multi-directional gradient value distribution of each position in the time-domain feature map of the high-frequency component of the carrier appearance, and determining the gradient amplitude value of each position in the time-domain feature map of the high-frequency component of the carrier appearance based on the multi-directional gradient value distribution of each position to obtain a gradient amplitude distribution map of the time-domain feature of the high-frequency component of the carrier appearance; Calculating the local description operator of the gradient amplitude at each position in the time-domain characteristic gradient amplitude distribution map of the high-frequency component of the carrier appearance to obtain a local significant distribution map of the time-domain characteristic gradient amplitude of the high-frequency component of the carrier appearance; Inputting the local significant distribution map of the time-domain characteristic gradient amplitude of the high-frequency component of the carrier appearance into a gated masker based on the GELU function to obtain a local significant gated mask map of the gradient amplitude; The position point multiplication between the gradient amplitude local significant gated mask map and the carrier appearance high-frequency component time-domain feature map is calculated to obtain the carrier appearance high-frequency component time-domain significant feature map.

2. The laser processing substrate production process according to claim 1, characterized in that: Extracting the high-frequency component of the carrier appearance image to obtain a carrier appearance high-frequency component time domain image includes: Performing a fast Fourier transform on the carrier appearance image to obtain a carrier appearance frequency domain image; The high-frequency components in the carrier appearance frequency domain image are extracted and the high-frequency components are subjected to inverse Fourier transformation to obtain a carrier appearance high-frequency component time domain image.

3. The laser processing carrier board production process according to claim 2, characterized in that: Performing image feature extraction on the carrier appearance high-frequency component time domain image to obtain a carrier appearance high-frequency component time domain feature image, including: The time domain image of the high-frequency component of the carrier appearance is input into an image feature extractor based on a hole convolutional neural network model to obtain the time domain feature image of the high-frequency component of the carrier appearance.

4. The laser processing carrier production process according to claim 1, characterized in that: Calculating the local description operator of the gradient amplitude at each position in the time domain characteristic gradient amplitude distribution map of the high-frequency component of the carrier appearance to obtain the local significant distribution map of the time domain characteristic gradient amplitude of the high-frequency component of the carrier appearance, including: The scale of the local neighborhood is determined, and the mean value of the difference between the gradient amplitude value at a predetermined position in the time domain characteristic gradient amplitude distribution diagram of the high-frequency component of the carrier appearance and the gradient amplitude values ​​at other positions in the local neighborhood is calculated to obtain a local description operator of the gradient amplitude corresponding to the predetermined position.

5. The laser processing carrier production process according to claim 4, characterized in that: Generating the appearance detection result based on the time domain significant feature map of the high-frequency component of the appearance of the carrier board includes: The time-domain significant feature map of the high-frequency component of the carrier appearance is input into a classifier-based appearance detection generator to obtain the appearance detection result.

6. The laser processing carrier production process according to claim 5, characterized in that: Inputting the time-domain significant feature map of the high-frequency component of the carrier appearance into the classifier-based appearance detection generator to obtain the appearance detection result, including: Expanding the time-domain significant feature map of the high-frequency component of the carrier appearance into a classification feature vector according to a row vector or a column vector; Performing full-connection encoding on the classification feature vector using the fully-connected layer of the classifier-based appearance detection generator to obtain an encoded classification feature vector; The encoded classification feature vector is input into the Softmax classification function of the classifier-based appearance detection generator to obtain the appearance detection result.

Images