Spi and aoi detection online rejudgment method and system

Abstract

The application discloses an SPI and AOI detection online rejudgment method and system, relates to the surface mount technology field of electronic manufacturing, and discloses the SPI and AOI detection online rejudgment method and system, which realizes intelligent correlation analysis of solder paste printing defects by performing coordinate alignment and image superposition processing on SPI height data and AOI optical data, and can improve rejudgment accuracy and efficiency of the solder paste printing defects.

Description

Technical Field

[0001] This application relates to the field of surface mount technology in electronic manufacturing, and in particular to a method and system for online verification of SPI and AOI inspections. Background Technology

[0002] In the electronics manufacturing process, solder paste printing is one of the key steps in surface mount technology (SMT). The quality of solder paste printing directly affects the subsequent component placement and soldering results. Traditional quality inspection uses SPI (Solder Paste Inspection) equipment and AOI (Automated Optical Inspection) equipment for separate inspections. The SPI equipment uses 3D scanning technology to obtain parameters such as the height, volume, and shape of the solder paste, while the AOI equipment uses optical imaging to detect defects on the surface of the circuit board after solder paste printing. This separate inspection method has obvious limitations: First, the independent operation of the two devices makes it difficult to correlate the inspection data in real time. When AOI detects a defect, the operator needs to manually retrieve the SPI data for comparison and analysis, which is inefficient and prone to errors. Second, because SPI data and AOI data use different coordinate systems and representation methods, it is difficult to establish an accurate correspondence when comparing manually, which can easily lead to misjudgments. Third, when AOI detects solder paste-related defects, it cannot quickly distinguish whether the problem is caused by the solder paste printing itself or by subsequent processes, increasing the complexity of fault diagnosis. Furthermore, existing technologies lack effective algorithms for intelligent correlation analysis between SPI height data and AOI optical data, resulting in insufficient accuracy and reliability in defect re-judgment. These issues severely restrict the efficiency and precision of quality control in electronic manufacturing processes, especially in the production of high-density circuit boards, where even minor solder paste defects can lead to serious soldering quality problems.

[0003] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main purpose of this application is to provide a method and system for online review of SPI and AOI inspections, which aims to improve the accuracy and efficiency of solder paste printing defect review.

[0005] To achieve the above objectives, this application proposes a method for online verification of SPI and AOI detection, the method comprising:

[0006] Obtain the SPI height data of the solder paste printed circuit board using an SPI device;

[0007] Extract component pad height distribution data for a predetermined region from the SPI height data, and convert the component pad height distribution data into first grayscale image data for the corresponding region;

[0008] The AOI optical image data of the same predetermined area of ​​the solder paste printed circuit board is obtained by an AOI device, and the AOI optical image data is converted into second grayscale image data.

[0009] Align the first grayscale image data with the second grayscale image data in coordinates to obtain the registered first image data and second image data, and then perform image overlay processing on the registered first image data and second image data to obtain the overlay image data.

[0010] When the AOI device determines that there is a defect in the predetermined area based on the second grayscale image data, it analyzes the correlation between the solder paste height distribution and optical features in the superimposed image data to obtain the re-judgment analysis results.

[0011] In one embodiment, the step of extracting component pad height distribution data of a predetermined region from the SPI height data includes:

[0012] Based on a preset component positioning algorithm, the height data of each component pad area is identified and segmented from the SPI height data;

[0013] Calculate the height value of the pixels within the pad area of ​​each component to obtain the component pad height distribution data.

[0014] In one embodiment, the preset component positioning algorithm includes:

[0015] Based on the Gerber file or CAD data of the solder paste printed circuit board, obtain the expected location and boundary data of the component pads;

[0016] Match the expected position and boundary data with the actual spatial coordinates of the SPI height data;

[0017] Based on the matching results, the area range of each component pad in the SPI height data is defined.

[0018] In one embodiment, the step of converting the component pad height distribution data into first grayscale image data of the corresponding area includes:

[0019] Define a mapping function from SPI height value to grayscale value, wherein the mapping function maps the minimum height value to the minimum grayscale value and the maximum height value to the maximum grayscale value;

[0020] Based on the mapping function, each height value in the component pad height distribution data is converted point by point into the corresponding grayscale value to generate the first grayscale image data.

[0021] In one embodiment, the mapping function is:

[0022] ;

[0023] in, Grayscale value This is the solder paste height value. The minimum height value in the component pad height distribution data. The maximum height value in the component pad height distribution data described in the step.

[0024] In one embodiment, the step of aligning the first grayscale image data and the second grayscale image data with coordinates to obtain registered first image data and second image data includes:

[0025] The coordinate data of the same predetermined feature points are extracted from the first grayscale image data and the second grayscale image data, respectively.

[0026] Based on the extracted coordinate data of two sets of feature points, calculate the coordinate transformation matrix data;

[0027] Using the coordinate transformation matrix data, the first grayscale image data is subjected to affine transformation processing to obtain the registered first image data, and the second grayscale image data is directly used as the registered second image data.

[0028] In one embodiment, the step of performing image overlay processing on the registered first image data and second image data to obtain overlay image data includes:

[0029] The registered first image data and second image data are subjected to transparency fusion processing, wherein the first image data is given a first transparency, the second image data is given a second transparency, and the sum of the first transparency and the second transparency is 1;

[0030] Based on the first transparency and the second transparency, the gray values ​​of corresponding pixels in the first image data and the second image data are weighted and summed to generate the gray values ​​of corresponding pixels in the superimposed image data.

[0031] In one embodiment, the step of analyzing the correlation between solder paste height distribution and optical features in the overlaid image data to obtain a re-judgment analysis result when the AOI device determines that a defect exists in the predetermined area based on the second grayscale image data includes:

[0032] In the overlaid image data, locate the defect area determined by the AOI device;

[0033] Extract the solder paste height distribution data and its spatial variation gradient data corresponding to the defect area in the overlay image;

[0034] Extract the AOI optical feature data and its abnormal pattern data corresponding to the defect region in the overlay image;

[0035] Based on a preset association rule base, the correlation between the solder paste height distribution and gradient data and the AOI optical features and abnormal pattern data is analyzed to determine whether solder paste printing abnormalities are the main cause of AOI defects and to obtain the re-judgment analysis results. The re-judgment analysis results include confirming AOI defect data, excluding AOI misjudged data, and suggesting re-inspection of SPI printing parameter data. The preset association rule base is used to define the correspondence between solder paste height abnormality patterns and AOI optical defect patterns.

[0036] Furthermore, to achieve the above objectives, this application also proposes an SPI and AOI detection online verification system, which includes: a memory, a processor, and an SPI and AOI detection online verification program stored in the memory and executable on the processor. The SPI and AOI detection online verification program is configured to implement the steps of the SPI and AOI detection online verification method.

[0037] The SPI and AOI detection online re-judgment method and system proposed in this application realizes intelligent correlation analysis of solder paste printing defects by aligning the coordinates of SPI height data and AOI optical data and performing image overlay processing, which can improve the accuracy and efficiency of solder paste printing defect re-judgment. Attached Figure Description

[0038] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0039] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 A flowchart illustrating an embodiment of the SPI and AOI detection online re-judgment method of this application;

[0041] Figure 2 This is a schematic diagram of a structural embodiment of the SPI and AOI detection online verification system of this application.

[0042] Explanation of icon numbers:

[0043] 10. Memory; 20. Processor.

[0044] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0045] The technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of this application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0046] It should be understood that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0047] In existing technologies, solder paste printing quality inspection in electronic manufacturing processes primarily relies on independent inspection by SPI (Solution-of-Place) and AOI (Automated Optical Inspection) equipment. The SPI equipment measures solder paste height parameters, while the AOI equipment inspects for optical surface defects. Because the data from these two types of equipment is not effectively correlated, when AOI detects a defect, it cannot directly determine whether the defect is caused by abnormal solder paste printing. This necessitates manual data verification, resulting in low inspection efficiency and a high false positive rate. For example, on a mobile phone motherboard production line, AOI may detect component misalignment in the pad area, but it cannot quickly determine whether the issue is caused by solder paste collapse or a problem in the mounting process.

[0048] To address these issues, researchers discovered that traditional detection methods suffer from data silos, with the SPI height data lacking an effective correlation dimension with the AOI optical image. Analysis revealed that abnormal solder paste height distribution directly affects the optical reflection characteristics of solder joints. Based on this understanding, a method was proposed to convert SPI height data into grayscale images, perform spatial registration and fusion analysis with the AOI optical image, and establish a correlation model between solder paste physical properties and optical performance. This allows for automatic tracing of solder paste printing quality factors when an AOI alarm is triggered.

[0049] Based on this, embodiments of this application provide a method for online re-judgment of SPI and AOI detection, referring to... Figure 1 The method includes steps S100 to S500, wherein:

[0050] Step S100: Obtain the SPI height data of the solder paste printed circuit board through the SPI device;

[0051] Step S200: Extract component pad height distribution data of a predetermined area from the SPI height data, and convert the component pad height distribution data into first grayscale image data of the corresponding area;

[0052] Step S300: Obtain AOI optical image data of the same predetermined area of ​​the solder paste printed circuit board through AOI equipment, and convert the AOI optical image data into second grayscale image data;

[0053] Step S400: Align the first grayscale image data with the second grayscale image data in coordinates to obtain the registered first image data and second image data, and perform image overlay processing on the registered first image data and second image data to obtain overlay image data;

[0054] Step S500: When the AOI device determines that there is a defect in the predetermined area based on the second grayscale image data, it analyzes the correlation between the solder paste height distribution and optical features in the superimposed image data to obtain the re-judgment analysis result.

[0055] In this embodiment, SPI height data refers to the physical measurement data of the solder paste printing layer in the vertical direction obtained through three-dimensional optical scanning. Specifically, it can be implemented using laser triangulation or structured light projection, and is used to quantitatively characterize the geometric shape of the solder paste printing. Component pad height distribution data refers to the set of height parameters of specific component soldering areas segmented from the overall SPI data. Specifically, it can be extracted using CAD coordinate matching or image segmentation algorithms, and is used to focus on key detection areas. Grayscale image conversion refers to mapping continuous height values ​​to a grayscale range of 0-255, which can be achieved using a linear normalization algorithm, making the data from different detection devices comparable. Coordinate alignment refers to establishing the spatial transformation relationship between the coordinate systems of different detection devices, which can be achieved using a feature point matching algorithm, ensuring the spatial consistency of multi-source data. Overlay image processing refers to fusing and displaying the two registered images, which can be achieved using a transparency mixing algorithm, intuitively presenting the correspondence between physical characteristics and optical performance. Correlation analysis refers to establishing the correspondence between solder paste morphological anomalies and optical defects using pattern recognition technology, which can be achieved using convolutional neural networks to extract feature associations, and is used to determine the root cause of defects.

[0056] In this embodiment, in a mobile phone motherboard manufacturing scenario, the SPI device first scans and obtains three-dimensional height data containing thousands of solder joints. For a preset chip area where component offset is detected, the height values ​​of the corresponding number of solder pads for that chip are extracted from the SPI data and converted into a grayscale image through linear mapping. Simultaneously, the AOI device captures an optical image of this area and converts it into a grayscale image. By identifying the feature points at the edge of the solder pads for coordinate registration, the SPI height grayscale image and the AOI optical grayscale image are superimposed and fused with transparency. When the AOI determines that there is an abnormal solder joint shape, the SPI height distribution at that location is observed to exhibit edge collapse characteristics in the superimposed image, and the collapsed area highly overlaps with the reflective abnormal area detected by the AOI, thus determining that the defect is caused by poor solder paste printing.

[0057] In this embodiment, spatial synchronization and visual correlation of two types of inspection data are achieved through data format standardization and automatic registration technology. Existing technologies lack a dimension for correlation analysis between the physical properties and optical performance of solder paste. This solution, however, establishes a quantifiable defect tracing mechanism through image overlay and pattern recognition, enabling automatic correlation analysis between solder paste printing quality inspection data and optical defect inspection data. This allows for rapid determination of whether a defect is caused by solder paste printing abnormalities when an AOI alarm is triggered, effectively reducing the workload of manual re-inspection, avoiding misjudgments caused by data silos, and improving the quality control efficiency of the electronic manufacturing process.

[0058] In one feasible implementation, the step of extracting component pad height distribution data of a predetermined area from the SPI height data includes: identifying and segmenting the height data of each component pad area from the SPI height data based on a preset component positioning algorithm; and calculating the height value of the pixel points within each component pad area to obtain component pad height distribution data.

[0059] In this embodiment, the component positioning algorithm refers to obtaining the expected position and boundary information of component pads from the solder paste printed circuit board design file, and matching it with the actual spatial coordinates of the SPI height data to delineate the area range of each component pad in the SPI height data. This algorithm, by combining design data and measured coordinates, can accurately identify pad areas, avoiding positioning errors that may occur when manually setting areas. Height distribution data refers to a quantized dataset generated by statistically analyzing the height value of each pixel within the pad area. Specifically, this can be achieved by traversing all pixels within the area and recording their height values. This data provides the basis for subsequent grayscale image conversion, enabling the solder paste height information to be presented in a visual form.

[0060] In this embodiment, when extracting component pad height distribution data, a preset component positioning algorithm is first invoked. This algorithm generates the expected coordinate range of the pad area based on predefined pad position information in Gerber files or CAD data. Next, the expected coordinates are spatially matched with the actual height data collected by the SPI device. The precise position of the pad in the SPI data is determined through coordinate transformation or feature point alignment. After positioning, the height values ​​of all pixels within each pad area are calculated, for example, using a region pixel coordinate traversal method, extracting height values ​​one by one and generating a dataset containing spatial distribution features. This process is fully automated, requiring no manual intervention in region division, ensuring the consistency and repeatability of data extraction.

[0061] In this embodiment, by combining the matching mechanism of design data and measured coordinates, it can dynamically adapt to the actual positional shifts of different batches of circuit boards. For example, when the pads undergo slight displacement due to fluctuations in the printing process, the area boundaries can still be accurately delineated through the coordinate transformation matrix. Furthermore, existing technologies often use regional average values ​​for pad height data extraction, while this solution collects height values ​​pixel by pixel, preserving the three-dimensional morphological details of solder paste printing. This allows for the capture of subtle defects such as local depressions or excess paste, achieving automated and accurate extraction of pad area height data. This solves the problems of low efficiency and susceptibility to subjective factors in manually setting detection areas. Dynamic matching of design data and measured coordinates ensures stable identification of pad areas in different production batches, such as accurate positioning even when the circuit board experiences positional shifts due to thermal expansion. Simultaneously, pixel-by-pixel height value calculation provides a high-resolution data foundation for subsequent grayscale image generation, ensuring the complete preservation of solder paste height distribution characteristics and laying the data foundation for height-optical feature correlation analysis during online review.

[0062] In one feasible implementation, the preset component positioning algorithm includes: obtaining the expected position and boundary data of component pads based on the Gerber file or CAD data of the solder paste printed circuit board; matching the expected position and boundary data with the actual spatial coordinates of the SPI height data; and defining the area range of each component pad in the SPI height data according to the matching result.

[0063] In this embodiment, the Gerber file refers to a standardized file format containing graphic information of each layer of the circuit board. Specifically, it can be a photoplot file output by circuit board design software, used to define the theoretical position and shape of the pads. CAD data refers to component layout data generated by computer-aided design software. Specifically, it can be a coordinate file or a component position table, used to provide the design coordinates of the pads on the circuit board. Expected position and boundary data refers to the ideal coordinates and contour range of the pads predetermined according to the design file. Specifically, it can be polygon vertex coordinates or region boundary parameters, used to establish a correspondence with the actual data detected by SPI. Actual spatial coordinates refer to the physical position information recorded when the SPI device collects height data. Specifically, it can be a set of three-dimensional coordinate points in the device coordinate system, used to spatially align the detection data with the design data. Matching refers to mapping the design data and detection data to the same coordinate system through coordinate transformation algorithms. Specifically, it can be a rigid transformation or an affine transformation, used to eliminate positional deviations between devices.

[0064] In this embodiment, after acquiring Gerber files or CAD data, the design positions and boundary information of the pads are extracted to form theoretical pad area coordinates. Subsequently, the actual coordinates in the height data collected by the SPI device are spatially matched with the theoretical coordinates. A coordinate transformation matrix is ​​used to adjust the positional deviation of the actual data, aligning the two in space. After matching, the corresponding area is delineated in the adjusted actual data based on the theoretical pad boundaries, thereby accurately segmenting the height data of each pad. This process, through automated matching of design and inspection data, avoids positioning errors that may occur when manually delineating areas, ensuring that subsequent height analysis is based on accurate pad areas.

[0065] In this embodiment, high-precision automatic matching is achieved by directly calling the theoretical position data in the design file and combining it with a coordinate transformation algorithm. This effectively eliminates coordinate deviations between devices, ensures the accuracy of pad area division, and reduces the need for manual intervention. Thus, this application achieves automated and accurate positioning of the pad area in SPI detection data, solving the problem of inaccurate area division caused by manual operation or coordinate deviations in traditional methods. This provides a reliable data foundation for subsequent solder paste height analysis and improves detection efficiency and consistency.

[0066] In one feasible implementation, the step of converting the component pad height distribution data into first grayscale image data of the corresponding area includes: setting a mapping function from SPI height values ​​to grayscale values, wherein the mapping function maps the minimum height value to the minimum grayscale value and the maximum height value to the maximum grayscale value; based on the set mapping function, converting each height value in the component pad height distribution data point by point into the corresponding grayscale value to generate the first grayscale image data.

[0067] In this embodiment, the mapping function refers to establishing a linear conversion rule between solder paste height values ​​and image grayscale values. Specifically, a normalization algorithm can be used to linearly map the height value range to a standard grayscale value range, for example, mapping the 0-100 micrometer height range to the 0-255 grayscale value range. This function is used to eliminate the difference in data dimensions between different detection devices, making SPI height data and AOI optical data comparable. Point-by-point conversion refers to independently calculating the pixel height data within each pad area. Specifically, an algorithm that traverses the pixel matrix can be used to perform mapping calculations on each height value, ensuring that the height distribution characteristics are completely preserved in the grayscale image.

[0068] In this embodiment, after acquiring the component pad height distribution data, the maximum and minimum height values ​​within the region are first determined. The height value of each pixel is then converted to a corresponding grayscale value using a normalization formula. For example, when the minimum height is 20 micrometers and the maximum height is 80 micrometers, a pixel with a height of 50 micrometers will be converted to a grayscale value of (50-20) / (80-20)*255=127. The first grayscale image data generated in this process has the same spatial resolution and grayscale range as the AOI optical image, providing a data foundation for subsequent image registration and overlay.

[0069] In this embodiment, by standardizing grayscale conversion, the SPI height data is converted into visual information of the same dimension as the AOI image, which solves the technical obstacle of multi-source heterogeneous data fusion. At the same time, the key features of the three-dimensional morphology of solder paste are preserved, realizing the efficient conversion of SPI detection data to standard image format. This enables the solder paste height distribution features to be compared with AOI optical features at the pixel level, providing a unified data processing basis for subsequent defect re-judgment and effectively improving the accuracy of correlation analysis between SPI and AOI detection results.

[0070] In one feasible implementation, the mapping function is:

[0071] ;

[0072] in, Grayscale value This is the solder paste height value. The minimum height value in the component pad height distribution data. The maximum height value in the component pad height distribution data described in the step.

[0073] In this embodiment, H_min refers to the minimum height of all pixels within the component pad area. This can be calculated by traversing the height data matrix to determine the minimum value, and is used to determine the lower limit of grayscale mapping. H_max refers to the maximum height of all pixels within the component pad area. This can be calculated by traversing the height data matrix to determine the upper limit of grayscale mapping. The Gray value refers to the conversion of the actual height value into a standard grayscale value using a linear normalization formula. This can be implemented using floating-point operations, ensuring that differences in height create a discernible contrast in the grayscale image.

[0074] In this embodiment, after acquiring the component pad height distribution data, the actual height extremes H_min and H_max within the region are dynamically calculated to establish a linear mapping relationship between height values ​​and grayscale values ​​ranging from 0 to 255. Each solder paste height point H is converted into a corresponding Gray value according to a formula, generating a first grayscale image with the same resolution as the AOI optical image. This mapping method can adapt to the height fluctuation range of different circuit boards. For example, when the overall solder paste height of a batch of circuit boards is too high, H_max will be automatically adjusted to the actual maximum value of that batch, avoiding image overexposure or insufficient contrast caused by a preset fixed threshold.

[0075] In this embodiment, by dynamically calculating the actual height extreme value of each detection area, the grayscale image can accurately reflect the relative height distribution characteristics within that area, eliminating image distortion caused by preset parameter deviations. Thus, this application achieves precise alignment of SPI height data and AOI optical images in grayscale space, enabling a direct correspondence between the three-dimensional morphology of the solder paste and its two-dimensional optical features in the superimposed image. This mapping method effectively improves the identifiability of height anomaly regions in the image. For example, when local collapse or bridging defects occur, the corresponding grayscale abrupt change area can form a spatial association with the optical anomaly region detected by AOI, providing a reliable basis for re-judgment. Simultaneously, dynamic normalization processing avoids manual adjustment of mapping parameters, ensuring the consistency of detection data across different batches.

[0076] In one feasible implementation, the step of aligning the first grayscale image data and the second grayscale image data with coordinates to obtain registered first image data and second image data includes: extracting coordinate data of the same predetermined feature points from the first grayscale image data and the second grayscale image data respectively; calculating coordinate transformation matrix data based on the extracted two sets of feature point coordinate data; and performing affine transformation processing on the first grayscale image data using the coordinate transformation matrix data to obtain registered first image data, wherein the second grayscale image data is directly used as registered second image data.

[0077] In this embodiment, predetermined feature points refer to positioning points with significant geometric features in the image. Specifically, these can be implemented using pad corner points, printed mark points, or points with specific shapes and contours, and are used to establish the spatial correspondence between two images. Coordinate transformation matrix data refers to mathematical parameters describing the translation, rotation, and scaling relationships between two images. Specifically, it can be calculated using the least squares method or singular value decomposition algorithm, and is used to unify images acquired by different devices to the same coordinate system. Affine transformation processing refers to the operation of performing linear geometric transformations on the image, specifically implemented as a product of translation, rotation, and scaling matrices, and is used to eliminate image misalignment caused by equipment installation deviations or circuit board position offsets.

[0078] In this embodiment, after the SPI and AOI devices generate grayscale images respectively, multiple feature points with the same physical location are first selected in both images, such as the four corners of a solder pad or the center point of a printed positioning mark. By comparing the coordinate differences between the two sets of feature points, a transformation matrix that can map the SPI image to the AOI image coordinate system is calculated. The SPI image is then subjected to an affine transformation according to this matrix to make it perfectly aligned with the AOI image in space. Since the AOI image itself already possesses accurate physical coordinate information, it can be used as a registration reference without additional transformation.

[0079] In this embodiment, by automatically extracting and transforming image feature points, spatial deviations between different detection devices can be accurately eliminated, ensuring that height data and optical images are accurately superimposed at the pixel level. This solves the problem of misjudgment caused by inaccurate spatial alignment of multi-source detection data, provides an accurate registration basis for subsequent image superposition and defect correlation analysis, reduces the probability of false alarms or missed detections caused by coordinate deviations, and improves the reliability of the online review system.

[0080] In one feasible implementation, the step of performing image overlay processing on the registered first image data and second image data to obtain overlay image data includes: performing transparency fusion processing on the registered first image data and second image data, wherein the first image data is given a first transparency, the second image data is given a second transparency, and the sum of the first transparency and the second transparency is 1; based on the first transparency and the second transparency, performing a weighted summation calculation on the gray values ​​of corresponding pixels in the first image data and the second image data, and fusing to generate the gray values ​​of corresponding pixels in the overlay image data.

[0081] In this embodiment, transparency blending refers to superimposing two registered images according to a set ratio. This can be achieved using an alpha blending algorithm, for example, setting the transparency of both the first and second image data to 0.5, so that both images appear semi-transparent when superimposed. The weighted summation calculation refers to linearly combining the pixel grayscale values ​​at the same location in the two images according to the set transparency ratio. This can be achieved using a grayscale value weighting formula, for example, the superimposed pixel grayscale value equals the first image grayscale value multiplied by the first transparency plus the second image grayscale value multiplied by the second transparency.

[0082] In this embodiment, after coordinate alignment, the registered first image data represents solder paste height distribution information, and the second image data represents AOI optical feature information. By setting different transparency ratios, such as setting the first transparency to 0.6 and the second transparency to 0.4, the two images can simultaneously preserve the details of height distribution and optical features when superimposed. The final grayscale value of each pixel is generated through weighted calculation. For example, when the grayscale value of a pixel in the first image is 200 and the grayscale value of the corresponding pixel in the second image is 100, the superimposed grayscale value is 200×0.6+100×0.4=160. This fusion method allows operators to intuitively observe the correspondence between areas of abnormal solder paste height and areas of optical defects.

[0083] In this embodiment, the display weights of the two types of data are dynamically adjusted by the transparency parameter, so that the superimposed image can retain the quantitative information of solder paste height distribution and highlight the optical anomaly features of AOI detection. This provides a clearer visual basis for defect correlation analysis and solves the misjudgment problem caused by isolated analysis of SPI and AOI data in traditional detection. The adjustable transparency fusion mechanism realizes the collaborative visualization of the two types of detection data, which enables the rapid identification of the causal relationship between solder paste printing defects and optical detection results during the re-judgment process, thereby improving the accuracy of defect judgment and detection efficiency.

[0084] In one feasible implementation, when the AOI device determines that a defect exists in the predetermined area based on the second grayscale image data, the step of analyzing the correlation between solder paste height distribution and optical features in the overlay image data to obtain a re-judgment analysis result includes: locating the defect area determined by the AOI device in the overlay image data; extracting the solder paste height distribution data and its spatial variation gradient data corresponding to the defect area in the overlay image; extracting the AOI optical feature data and its abnormal pattern data corresponding to the defect area in the overlay image; and analyzing the correlation between the solder paste height distribution and gradient data and the AOI optical features and abnormal pattern data based on a preset association rule base to determine whether the solder paste printing abnormality is the main cause of the AOI defect, and obtaining a re-judgment analysis result. The re-judgment analysis result includes confirming AOI defect data, excluding AOI misjudged data, and suggesting re-inspection of SPI printing parameter data. The preset association rule base is used to define the correspondence between solder paste height abnormality patterns and AOI optical defect patterns.

[0085] In this embodiment, the association rule base refers to a database of predefined correspondences between solder paste height anomaly patterns and AOI optical defect patterns. Specifically, it can be implemented using a set of rules built based on historical inspection data, used to establish the causal relationship between solder paste printing parameters and optical defects. Spatial variation gradient data refers to the rate of change of solder paste height within the defect region, specifically extracted by calculating the height difference between adjacent pixels or using gradient operators, used to characterize the uniformity of solder paste distribution. Anomaly pattern data refers to the optical feature classification results of defect manifestations in AOI images. For example, it can be used to perform pattern recognition on image texture and contrast through convolutional neural networks to distinguish different defect types.

[0086] In this embodiment, when the AOI detects a defect, the system first selects the coordinate range of the defect area in the overlay image and simultaneously acquires the three-dimensional distribution of the SPI height data of that area and its horizontal / vertical gradient distribution. Simultaneously, it extracts the grayscale distribution, edge sharpness, or texture features of the area from the AOI image and identifies its abnormal pattern category using a classification model. Subsequently, the maximum value, standard deviation, and gradient abrupt change point locations of the solder paste height distribution are matched with the offset, bridging, or missing defect patterns detected by the AOI against an association rule base. For example, when a local sudden drop in solder paste height is detected and the gradient exhibits a ring-shaped diffusion, if the corresponding AOI image displays component offset characteristics, it is determined to be a placement offset caused by insufficient solder paste; if the AOI only displays color abnormalities without positional offset, it is determined to be an optical false detection.

[0087] In this embodiment, by establishing an association rule base, solder paste printing quality parameters are cross-validated with optical defect patterns. For example, when a uniform solder paste height is detected but component defects are present, printing problems can be directly ruled out, and it is recommended to check the mounting process, thereby avoiding misjudgments caused by isolated data in traditional methods. Through the above technical solution, this application can accurately distinguish solder paste printing abnormalities from other process defects, reducing the misjudgment rate of AOI inspection. For example, when the solder paste height distribution meets the standard but AOI shows an abnormality, it can be quickly identified as an optical pseudo-defect; when height distribution abnormalities and optical defects coexist, printing process problems can be determined and parameter adjustments can be triggered, thereby improving the accuracy of defect analysis and production line debugging efficiency.

[0088] In this embodiment, the SPI and AOI detection online re-judgment method achieves intelligent correlation analysis of solder paste printing defects by aligning the coordinates of SPI height data and AOI optical data and performing image overlay processing, which can improve the accuracy and efficiency of solder paste printing defect re-judgment.

[0089] This application also provides an online verification system for SPI and AOI detection, see reference. Figure 2 The SPI and AOI detection online verification system includes: a memory 10, a processor 20, and an SPI and AOI detection online verification program stored in the memory 10 and executable on the processor 20. The SPI and AOI detection online verification program is configured to implement the steps of the SPI and AOI detection online verification method.

[0090] The SPI and AOI online inspection review system provided in this application, employing the SPI and AOI online inspection review method described in the above embodiments, can improve the accuracy and efficiency of solder paste printing defect review. Compared with the prior art, the beneficial effects of the SPI and AOI online inspection review system provided in this application are the same as those of the SPI and AOI online inspection review method provided in the above embodiments, and other technical features of the SPI and AOI online inspection review system are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0091] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for online verification and judgment of SPI and AOI detection, characterized in that, The method includes: Obtain the SPI height data of the solder paste printed circuit board using an SPI device; Extract component pad height distribution data for a predetermined region from the SPI height data, and convert the component pad height distribution data into first grayscale image data for the corresponding region; The AOI optical image data of the same predetermined area of ​​the solder paste printed circuit board is obtained by an AOI device, and the AOI optical image data is converted into second grayscale image data. Align the first grayscale image data with the second grayscale image data in coordinates to obtain the registered first image data and second image data, and then perform image overlay processing on the registered first image data and second image data to obtain the overlay image data. When the AOI device determines that there is a defect in the predetermined area based on the second grayscale image data, it analyzes the correlation between the solder paste height distribution and optical features in the superimposed image data to obtain the re-judgment analysis results; The step of converting the component pad height distribution data into first grayscale image data of the corresponding region includes: Define a mapping function from SPI height value to grayscale value, wherein the mapping function maps the minimum height value to the minimum grayscale value and the maximum height value to the maximum grayscale value; Based on the mapping function, each height value in the component pad height distribution data is converted into a corresponding grayscale value point by point to generate the first grayscale image data; The mapping function is: ； in, Grayscale value This is the solder paste height value. The minimum height value in the component pad height distribution data. The maximum height value in the component pad height distribution data described in the step; The step of aligning the first grayscale image data and the second grayscale image data with coordinates to obtain the registered first image data and second image data includes: The coordinate data of the same predetermined feature points are extracted from the first grayscale image data and the second grayscale image data, respectively. Based on the extracted coordinate data of two sets of feature points, calculate the coordinate transformation matrix data; Using the coordinate transformation matrix data, the first grayscale image data is subjected to affine transformation processing to obtain the registered first image data, and the second grayscale image data is directly used as the registered second image data.

2. The SPI and AOI detection online re-judgment method as described in claim 1, characterized in that, The step of extracting component pad height distribution data for a predetermined region from the SPI height data includes: Based on a preset component positioning algorithm, the height data of each component pad area is identified and segmented from the SPI height data; Calculate the height value of the pixels within the pad area of ​​each component to obtain the component pad height distribution data.

3. The SPI and AOI detection online re-judgment method as described in claim 2, characterized in that, The preset component positioning algorithm includes: Based on the Gerber file or CAD data of the solder paste printed circuit board, obtain the expected location and boundary data of the component pads; Match the expected position and boundary data with the actual spatial coordinates of the SPI height data; Based on the matching results, the area range of each component pad in the SPI height data is defined.

4. The SPI and AOI detection online re-judgment method as described in claim 1, characterized in that, The step of performing image overlay processing on the registered first image data and second image data to obtain overlay image data includes: The registered first image data and second image data are subjected to transparency fusion processing, wherein the first image data is given a first transparency, the second image data is given a second transparency, and the sum of the first transparency and the second transparency is 1; Based on the first transparency and the second transparency, the gray values ​​of corresponding pixels in the first image data and the second image data are weighted and summed to generate the gray values ​​of corresponding pixels in the superimposed image data.

5. The SPI and AOI detection online re-judgment method as described in claim 1, characterized in that, When the AOI device determines that there is a defect in the predetermined area based on the second grayscale image data, the step of analyzing the correlation between the solder paste height distribution and optical features in the overlaid image data to obtain the re-judgment analysis result includes: In the overlaid image data, locate the defect area determined by the AOI device; Extract the solder paste height distribution data and its spatial variation gradient data corresponding to the defect area in the overlay image; Extract the AOI optical feature data and its abnormal pattern data corresponding to the defect region in the overlay image; Based on a preset association rule base, the correlation between the solder paste height distribution and gradient data and the AOI optical features and abnormal pattern data is analyzed to determine whether solder paste printing abnormalities are the main cause of AOI defects and to obtain the re-judgment analysis results. The re-judgment analysis results include confirming AOI defect data, excluding AOI misjudged data, and suggesting re-inspection of SPI printing parameter data. The preset association rule base is used to define the correspondence between solder paste height abnormality patterns and AOI optical defect patterns.

6. A system for online verification and judgment of SPI and AOI detection, characterized in that, The SPI and AOI detection online verification system includes: a memory, a processor, and an SPI and AOI detection online verification program stored in the memory and executable on the processor. The SPI and AOI detection online verification program is configured to implement the steps of the SPI and AOI detection online verification method as described in any one of claims 1 to 5.

Images