PCB production line defect detection analysis method and system based on recognition model

Abstract

The invention discloses a PCB production line defect detection analysis method and system based on an identification model, particularly relates to the technical field of PCB quality detection, and is used for solving the problems of low PCB defect detection precision and high omission ratio in the prior art. The method comprises the following steps: scanning the whole surface of a PCB to generate a thickness distribution diagram, identifying a thickness non-uniform region by adopting a threshold segmentation algorithm, and marking a key detection region; potential defects irrelevant to process fluctuation are identified through historical production data analysis; and adjusting the polarization angle of the light source to analyze the reflectivity difference, and identifying the hidden defect area in combination with the microstructure of the material. The union set of the potential defect area and the hidden defect area is marked as the suspected incomplete area, and the influence of the suspected incomplete area on the performance of the PCB is analyzed according to the quality standard, so that the detection precision and comprehensiveness are improved, and the high-quality control of the PCB is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of PCB quality detection, and more specifically, to a PCB production line defect detection and analysis method and system based on a recognition model. Background Art

[0002] With the rapid development of electronic equipment, the production quality of PCB boards has become an important link to ensure product performance. At present, defect detection in PCB board production lines mostly uses traditional optical detection and simple thickness measurement methods. These methods have great limitations when dealing with subtle defects in mass production, especially when detecting hidden defects caused by production process fluctuations and changes in material properties. It is difficult to ensure the comprehensiveness and accuracy of detection.

[0003] In the existing technology, PCB defect detection is mostly focused on surface feature recognition and thickness difference detection. However, these methods can usually only detect obvious surface defects. For small defects caused by uneven thickness and hidden defects caused by the optical reflection characteristics of the material structure, the detection effect is poor, which will lead to a high false detection rate in the production process, affecting the detection efficiency and the quality of the finished product.

[0004] In order to solve the above problems, a technical solution is now provided. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a PCB production line defect detection and analysis method and system based on a recognition model to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions: A PCB production line defect detection and analysis method based on a recognition model comprises the following steps: Scan the entire surface of the PCB board to be inspected, obtain the thickness distribution data of the board, and generate the corresponding thickness distribution map; The threshold segmentation algorithm is used to identify the area with uneven thickness and determine whether the thickness change exceeds the preset threshold. If so, the corresponding area is marked as the key detection area; Analyze the deformation trend of key inspection areas through historical production data to identify potential defects in key inspection areas that are not related to process fluctuations; By adjusting the polarization angle of the light source to analyze the reflectivity difference in the key inspection area, combined with the material microstructure to evaluate the change in reflective characteristics, hidden defects in the key inspection area can be identified; Based on the potential defects unrelated to process fluctuations and hidden defects in the key inspection area, the suspected defective areas in the key inspection area are preliminarily detected; Further analysis is performed on suspected defective areas based on preset quality standards to determine the extent of their impact on PCB performance.

[0007] In a preferred embodiment, the entire surface of the PCB plate to be inspected is scanned to obtain the thickness distribution data of the plate and generate a corresponding thickness distribution diagram, specifically: Start the scanning device to ensure that it can cover the entire surface of the PCB to be tested, and set the scanning resolution and sampling frequency; Use scanning equipment to scan the entire surface of the PCB board to obtain the original thickness data at each location; De-noising and normalizing the collected raw thickness data; Based on the processed thickness data, the corresponding thickness distribution map is generated to intuitively display the thickness change trend of each area.

[0008] In a preferred embodiment, a threshold segmentation algorithm is used to identify the uneven thickness area, and it is determined whether the thickness change exceeds a preset threshold. If so, the corresponding area is marked as a key detection area, specifically: Perform data analysis on the generated thickness distribution map, extract the thickness value of each area from the map, and convert it into a digital matrix; Select a threshold segmentation algorithm suitable for thickness detection, and set a thickness detection threshold suitable for the PCB board; The thickness data is processed using a threshold segmentation algorithm to identify areas with significant thickness changes; the area where the value in the thickness distribution is significantly higher or lower than the normal range is analyzed using the threshold segmentation algorithm to determine the area with uneven thickness; For the identified uneven thickness area, determine whether its thickness change exceeds its corresponding preset threshold; if the thickness change exceeds its corresponding preset threshold, mark the area as a key detection area.

[0009] In a preferred embodiment, the deformation trend of the key inspection area is analyzed by historical production data to identify potential defects in the key inspection area that are not related to process fluctuations, specifically: Extract historical production thickness data of key inspection areas; Calculate the dynamic deformation index of the key inspection area based on historical production thickness data, and quantify the deformation amplitude and trend based on the dynamic deformation index; A multi-dimensional recursive neural network model was established, and the dynamic deformation index was input to perform time series modeling of the deformation trend; A multi-dimensional recursive neural network model is used to predict the deformation trend of key inspection areas; Abnormal deformation that is not related to process fluctuations can be identified based on the predicted deformation trend, and potential defect areas of deformation can be identified.

[0010] In a preferred embodiment, the dynamic deformation index of the key detection area is calculated based on the historical production thickness data, and the deformation amplitude and trend are quantified based on the dynamic deformation index, specifically: The dynamic deformation index is defined as the thickness change rate in time and space, and is calculated as follows: ;in, Indicates The dynamic deformation index of the batch, For the Batch No. Thickness measurements, Indicates The average thickness of the batch, represents the number of locations where thickness measurements are made in each production batch, is the batch sequence in time, is the measurement position within each batch; Calculate the increment: ;in, It is an increment.

[0011] In a preferred embodiment, the reflectivity difference of the key inspection area is analyzed by adjusting the polarization angle of the light source, and the change of the reflective characteristics is evaluated in combination with the microstructure of the material to identify the hidden defects in the key inspection area, specifically: Adjust the polarization angle of the light source according to the material characteristics of the key detection area; Collect multi-angle reflection data of key detection areas at different polarization angles; Analyze multi-angle reflection data in combination with material microstructural properties to identify abnormal changes in reflection characteristics; Based on abnormal changes in reflection characteristics, hidden defect areas within the key inspection area are identified.

[0012] In a preferred embodiment, based on potential defects unrelated to process fluctuations in the key inspection area and hidden defects in the key inspection area, the suspected defective area in the key inspection area is preliminarily detected, specifically: In the key inspection area, the area corresponding to the union of the deformation potential defect area and the hidden defect area is marked as a suspected defect area.

[0013] In a preferred embodiment, the suspected defective area is further analyzed according to a preset quality standard to determine the degree of its impact on the PCB performance, specifically: Conduct preliminary screening of the marked suspected defect areas to determine the detection parameters that meet the preset quality standards; For the suspected defective areas after screening, performance parameters are extracted, including physical characteristics and structural characteristics; Based on the extracted performance parameters and the preset quality standards, the impact of the suspected defective area is quantitatively analyzed; Compare the assessment results with the quality standards to determine whether the impact of the suspected defective area exceeds the allowable range and make a qualification determination; Mark the unqualified suspected defective areas and record the location information, and generate a test report.

[0014] On the other hand, the present invention provides a PCB production line defect detection and analysis system based on a recognition model, including a plate surface scanning module, a thickness distribution recognition module, a deformation trend analysis module, a polarization reflection analysis module, a suspected defect marking module, and a quality standard evaluation module; Plate surface scanning module: scans the entire surface of the PCB plate to be tested, obtains the thickness distribution data of the plate, and generates the corresponding thickness distribution map; Thickness distribution identification module: uses threshold segmentation algorithm to identify uneven thickness areas and determine whether the thickness change exceeds the preset threshold. If so, the corresponding area is marked as a key detection area; Deformation trend analysis module: Analyzes the deformation trend of key inspection areas through historical production data, and identifies potential defects in key inspection areas that are not related to process fluctuations; Polarized reflection analysis module: By adjusting the polarization angle of the light source, the reflectivity difference in the key inspection area is analyzed, and the change of the reflection characteristics is evaluated in combination with the microstructure of the material to identify hidden defects in the key inspection area; Suspected defect marking module: Based on potential defects unrelated to process fluctuations and hidden defects in key inspection areas, it preliminarily detects suspected defect areas in key inspection areas; Quality standard assessment module: Further analyze the suspected defective areas according to the preset quality standards to determine the degree of impact on PCB performance.

[0015] The technical effects and advantages of the PCB production line defect detection and analysis method and system based on the recognition model of the present invention are as follows: 1. The present invention proposes a PCB production line defect detection and analysis method based on a recognition model, which can effectively improve the accuracy and comprehensiveness of PCB board detection. In this method, the thickness data of the entire surface of the board is scanned to generate a thickness distribution map, and a threshold segmentation algorithm is used to identify areas with uneven thickness, and accurately mark key detection areas. This approach ensures the effective identification of key areas, so that the detection range is not only limited to surface features, but also covers potential defective areas caused by thickness changes. In addition, the present invention analyzes production history data to identify potential defects in key detection areas that are not related to process fluctuations, further reducing the false detection rate caused by process fluctuations, and providing data support for the accurate calibration of defective areas.

[0016] 2. The present invention also combines the light source polarization angle adjustment technology to identify hidden defects in key inspection areas through multi-angle reflection analysis. By analyzing the reflection characteristics in the key inspection area, combined with the material microstructure characteristics, hidden defect areas that are difficult to capture in conventional inspections are identified. Ultimately, this method combines potential defect areas that are not related to process fluctuations with hidden defect areas, preliminarily marks suspected defective areas, and further evaluates their degree of impact based on preset quality standards. This detection method significantly improves the sensitivity and accuracy of the detection system, reduces the possibility of missed detection and false detection, improves the overall quality control capability of the PCB production line, and provides reliable guarantees for high-quality PCB production. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic diagram of a PCB production line defect detection and analysis method based on a recognition model according to the present invention; Figure 2 The present invention is a structural schematic diagram of a PCB production line defect detection and analysis system based on a recognition model. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Example 1 Figure 1 The present invention provides a PCB production line defect detection and analysis method based on a recognition model, which includes the following steps: Scan the entire surface of the PCB board to be inspected, obtain the thickness distribution data of the board, and generate the corresponding thickness distribution map.

[0020] The threshold segmentation algorithm is used to identify the area with uneven thickness and determine whether the thickness change exceeds the preset threshold. If so, the corresponding area is marked as the key detection area.

[0021] Analyze the deformation trends of key inspection areas through historical production data, and identify potential defects in key inspection areas that are not related to process fluctuations.

[0022] By adjusting the polarization angle of the light source to analyze the reflectivity differences in the key inspection area, and combining the material microstructure to evaluate the changes in the reflective characteristics, hidden defects in the key inspection area can be identified.

[0023] Based on the potential defects in the key inspection area that are not related to process fluctuations and the hidden defects in the key inspection area, the suspected defective areas in the key inspection area are preliminarily detected.

[0024] Further analysis is performed on suspected defective areas based on preset quality standards to determine the extent of their impact on PCB performance.

[0025] Scan the entire surface of the PCB board to be tested, obtain the thickness distribution data of the board, and generate the corresponding thickness distribution map, specifically: Start the scanning device to ensure that it can cover the entire surface of the PCB to be tested, and set the scanning resolution and sampling frequency: First, start the scanning device (such as an optical scanner or laser rangefinder) and set the working parameters of the scanning device, including scanning resolution, sampling frequency and scanning angle. The resolution should be set according to the detailed requirements of the PCB board. Usually, high-precision scanning at the micron level can be selected to ensure the accuracy of the thickness data. The sampling frequency needs to be reasonably configured according to the size and requirements of the PCB board to be tested to ensure that the full surface is covered and the data density is high enough. The equipment should be able to ensure that the scanning covers the entire area of ​​the board to be tested without missing any corners through automatic adjustment function or manual calibration.

[0026] The design of the scanning equipment should ensure that the scanning range can be automatically adjusted during the scanning process to ensure that the entire PCB surface is within the coverage area of ​​the scanner. For larger PCB boards, the scanning equipment can adopt a block scanning method to ensure the integrity of the scanning range and the continuity of the data. During the scanning process, the equipment needs to record the coordinate information of each scanning area and adjust it in real time to cover the entire board surface.

[0027] Use a scanning device to scan the entire surface of the PCB board to obtain the original thickness data at each location: The device scans the surface of the PCB board step by step according to the previously set scanning range and parameters. Each time it scans, the device will capture the height data of the PCB board surface through a high-precision sensor and calculate the thickness value at each position. The thickness data is composed of multiple scanning points, each scanning point represents the thickness of a certain position on the PCB surface, and all the collected thickness data will be combined into a complete thickness data set.

[0028] Depending on the different types of PCB materials and their surface conditions, the device may need to adjust the scanning accuracy in time during the scanning process. For example, for thinner or smoother PCB materials, a higher resolution scanning mode can be used; for thicker or uneven PCB materials, the scanning accuracy can be appropriately reduced to improve the stability of data acquisition.

[0029] De-noising and normalization of the collected raw thickness data: Preprocess the raw thickness data, including removing the noise that may be introduced during the scanning process. For example, because the equipment may be affected by environmental factors (such as temperature, light, etc.), the collected thickness data may have certain fluctuations or abnormal points. At this time, algorithms such as median filtering and Kalman filtering can be used to smooth the data to remove the influence of noise.

[0030] Normalize the denoised thickness data to map the data values ​​to a uniform range, such as normalizing the thickness values ​​to between 0 and 1. The normalization operation helps eliminate thickness data deviations caused by differences in equipment accuracy or environmental changes, making subsequent analysis more consistent.

[0031] Based on the processed thickness data, the corresponding thickness distribution diagram is generated to intuitively display the thickness change trend of each area: Based on the pre-processed thickness data, a thickness distribution map is generated using visualization tools or algorithms (such as heat map drawing, contour map, etc.). This map shows the thickness difference in each area and can help inspectors intuitively observe the thickness changes between PCB surface areas. For example, thicker areas can be represented by red and thinner areas by blue.

[0032] The thickness distribution map not only intuitively displays the thickness value of each area, but also reflects the trend of thickness change through color changes. These trends are crucial for subsequent defect analysis and can help identify potential problem areas that may arise due to uneven thickness, such as reduced mechanical properties due to local insufficient thickness.

[0033] The threshold segmentation algorithm is used to identify the area with uneven thickness and determine whether the thickness change exceeds the preset threshold. If so, the corresponding area is marked as the key detection area. Specifically: The generated thickness distribution map is analyzed, the thickness value of each area is extracted from the map, and converted into a digital matrix to provide standardized thickness data input for subsequent threshold segmentation processing.

[0034] Select a threshold segmentation algorithm suitable for thickness detection (such as histogram-based threshold segmentation or region growing algorithm), and set a thickness detection threshold suitable for the PCB material.

[0035] Among them, the thickness detection threshold is determined based on experiments or historical data to ensure the scientificity and applicability of the threshold.

[0036] The thickness data is processed using a threshold segmentation algorithm to identify areas with significant thickness changes. The threshold segmentation algorithm automatically analyzes areas in the thickness distribution where the values ​​are significantly higher or lower than the normal range, thereby determining areas with uneven thickness.

[0037] For the identified uneven thickness area, determine whether its thickness change exceeds its corresponding preset threshold. If the thickness change exceeds its corresponding preset threshold, the area is marked as a key detection area to provide a clear focus area for subsequent detailed detection.

[0038] Among them, the preset threshold corresponding to the thickness change is determined based on the production standards and historical data of the PCB board, taking into account the normal process fluctuation range and the allowable error of the board thickness, and obtaining the most suitable value for detection accuracy through experimental calibration.

[0039] Analyze the deformation trend of key inspection areas through historical production data, and identify potential defects in key inspection areas that are not related to process fluctuations, specifically: Extract historical production thickness data of key inspection areas: The historical thickness data related to the key inspection areas are extracted from the production database to ensure that the analyzed data can accurately reflect the long-term trend of deformation. The data should include the thickness measurements of the key inspection areas in each production batch in history to ensure the integrity and continuity of the data in time. In addition, to ensure the accuracy and consistency of the data, a data cleaning step is used to remove outliers or missing data.

[0040] The dynamic deformation index of the key inspection area is calculated based on the historical production thickness data, and the deformation amplitude and trend are quantified based on the dynamic deformation index: Assume that each thickness data point is ,in is the batch sequence in time, is the measurement position within each batch. The dynamic deformation index is defined as the thickness change rate in time and space, and the calculation formula is as follows: ;in, Indicates The dynamic deformation index of the batch, For the Batch No. Thickness measurements, Indicates The average thickness of the batch, Indicates the number of locations where thickness measurements are made in each production batch.

[0041] The dynamic deformation index value obtained through this calculation can reflect the deformation amplitude and trend of the key inspection area of ​​each batch.

[0042] In order to further accurately reflect the amplitude of dynamic deformation, the increment is calculated: ;in, It is an increment, which indicates the increment of the dynamic deformation index between the current batch and the previous batch, reflecting the deformation change rate of the key detection area.

[0043] like If there is a significant increase or decrease, it indicates that there may be a large deformation change at that time point, which requires further attention.

[0044] A multi-dimensional recursive neural network model is established, and the dynamic deformation index is input to perform time series modeling of the deformation trend: The Long Short-Term Memory (LSTM) network is selected as the basic unit of the multidimensional recurrent neural network to ensure its long-term memory capability. and The number of nodes in each layer is determined according to the complexity of the data. The network structure includes an input layer, a hidden layer, and an output layer, and the hidden layer is composed of multiple layers of LSTM units to enhance the ability to capture long-term data trends.

[0045] The history and Input into a multidimensional recurrent neural network to build a time series model. The input data for each time step includes and , so the input data of the network is in the form of , which is used to predict the dynamic deformation trend of future time steps.

[0046] The multidimensional recurrent neural network is trained using the back propagation algorithm to minimize the loss function. The loss function uses the mean squared error (MSE), which is calculated as follows: ;in, Represents mean square error (MSE), which is used to quantify the average error between the predicted value and the actual value; Represents the total number of time batches, that is, the number of batches contained in the entire time series; It is The predicted dynamic deformation index values ​​of each batch are calculated by a multi-dimensional recursive neural network model and used to estimate the deformation state of the key detection area at that time step; It is The actual dynamic deformation index value of each batch, that is, the real deformation index value obtained from the historical data, represents the actual deformation of the key detection area at this time step.

[0047] The multi-dimensional recursive neural network model is used to predict the deformation trend of the key detection area: After completing the training of the multidimensional recurrent neural network model, the current and As input, the future deformation trend is predicted to obtain the predicted dynamic deformation index. The sequence of predicted dynamic deformation index can characterize the future deformation changes of the key detection area and help predict possible abnormal deformation.

[0048] The predicted dynamic deformation index includes not only the thickness variation of a specific batch in the future, but also the acceleration or deceleration trend of deformation. The model can be used to identify possible deformation patterns in the future, providing a basis for subsequent defect detection.

[0049] Identify abnormal deformation that is not related to process fluctuations based on the predicted deformation trend and identify potential defect areas of deformation: The future deformation trend sequence predicted by the multidimensional recursive neural network model is analyzed to analyze its change trend and incremental pattern, and to identify abnormal deformation that is not related to process fluctuations.

[0050] If the predicted dynamic deformation index or its increment at a certain time point in the key inspection area exceeds the normal process fluctuation range, the area is identified as a potential deformation defect area.

[0051] By adjusting the polarization angle of the light source to analyze the reflectivity difference in the key inspection area, combined with the material microstructure to evaluate the change in reflective characteristics, hidden defects in the key inspection area can be identified, specifically: Adjust the polarization angle of the light source according to the material characteristics of the key detection area to ensure the reflection capture effect within the optimal angle range: The material characteristics of the key inspection area are analyzed to determine its surface microstructure and reflection characteristics. The surface structures of different materials (such as copper, nickel, gold, etc.) will produce different polarization responses to reflected light, so adjusting the polarization angle of the light source to the optimal range can ensure accurate capture of the reflected light. In specific implementations, a light source polarization adjustment device (such as a polarizing filter or a variable angle polarizer) is used to adjust the incident angle of the light source.

[0052] Polarization angle setting: According to the material characteristics, the polarization angle range is preliminarily set. Assume that the optimal incident angle range is arrive , gradually adjust the polarization angle within this range, and measure the reflection effect at each angle one by one.

[0053] Optimal angle determination: gradually narrow the angle range based on the reflection capture effect, and finally determine the optimal polarization angle , so that the intensity and accuracy of the reflected signal are optimal at this angle. The setting of the angle takes into account the microscopic characteristics and multi-layer structure of the material to ensure the acquisition of effective reflectivity data.

[0054] Collect multi-angle reflection data of key detection areas at different polarization angles: After determining the optimal polarization angle range, multi-angle reflection data collection is performed on the key detection area by scanning angle by angle. At each angle, the reflected light intensity at that angle is recorded to generate a reflection data sequence at different polarization angles so that the changes in reflection characteristics can be compared in subsequent analysis.

[0055] Angle-by-angle data collection: Set the polarization angle to multiple different angles, such as the first angle, the second angle, etc., collect the reflected light intensity of the key detection area at each angle, and record the corresponding reflectivity data respectively.

[0056] Data recording format: The collected reflected light intensity data are organized into a matrix format, one part of which represents different polarization angles and the other part represents the reflected light intensity at different sampling points, ensuring that the reflection data at each angle can be used for subsequent analysis.

[0057] Analyze multi-angle reflection data in combination with material microstructural properties to identify abnormal changes in reflection characteristics: After the multi-angle reflection data is collected, the data is analyzed in combination with the microstructural characteristics of the material to identify anomalies in the reflection characteristics. During the analysis, it is necessary to consider the surface roughness of the material, the layered structure, and the differences in the reflection response of different microscopic particles at different polarization angles.

[0058] Reflectivity difference calculation: The reflectivity at different angles is calculated to identify changes in reflectivity characteristics. Specifically, the reflectivity data at different angles is compared to determine whether the difference in reflectivity between the two angles is significant. If the reflectivity data shows large changes between certain angles, it may point to abnormalities in the surface or sub-surface structure of the material.

[0059] Set the reflectivity difference threshold. If the reflectivity difference between the two angles exceeds the reflectivity difference threshold, it is determined that there is an abnormal reflective characteristic. The reflectivity difference threshold is set based on experimental data or historical data to ensure the reliability of the test results.

[0060] Identify hidden defect areas within key inspection areas based on abnormal changes in reflection characteristics: On the basis of identifying the reflectivity anomaly, the hidden defects in the key inspection area are further judged according to the abnormal changes in the reflection characteristics. Hidden defects usually manifest as abnormal reflection characteristics at certain angles, or inconsistent reflection changes in the microstructure of certain specific materials.

[0061] The area with abnormal reflection characteristics is marked as a hidden defect area, and the polarization angle and reflection intensity change characteristics of the area are recorded to generate a defect distribution map in the key inspection area to provide data reference for subsequent steps.

[0062] Based on the potential defects unrelated to process fluctuations in the key inspection area and the hidden defects in the key inspection area, the suspected defective areas in the key inspection area are preliminarily detected, specifically: In the key inspection area, the area corresponding to the union of the deformation potential defect area and the hidden defect area is marked as a suspected defect area, and the coordinate information and area range of the area are generated to facilitate subsequent analysis.

[0063] Specifically, the deformation potential defect area and the hidden defect area mark defects caused by different factors. The former is an unstable area identified based on dynamic deformation analysis, while the latter is an abnormal area on the surface or sub-surface identified based on reflection characteristics. By combining these two defect areas, all possible defect areas can be more comprehensively covered to ensure the integrity of the detection.

[0064] Further analysis is performed on the suspected defective areas according to the preset quality standards to determine the degree of impact on the PCB performance, specifically: Perform a preliminary screening of the marked suspected defect areas to determine the detection parameters that meet the preset quality standards for detailed analysis: Screening is based on preset quality standards, which usually include the area, depth, location, and defect type of the region. The screening process identifies the basic parameters of each suspected defect area (such as area size, depth range, etc.) and retains the areas that meet the inspection standards for subsequent detailed analysis. This can reduce unnecessary analysis and focus on suspected defect areas that have a real impact on PCB performance.

[0065] Screening basis: Based on the preset quality standards for area, depth, and location, determine which suspected defective areas may have a significant impact on performance and require further analysis.

[0066] Parameter determination: Determine the detection parameters after screening, such as measuring the area size and average depth of the defect area to ensure that the quality standards are met.

[0067] For the suspected defective areas after screening, extract the performance parameters: Performance parameters include physical characteristics and structural characteristics.

[0068] Physical feature acquisition: Use imaging equipment to obtain detailed information of suspected defective areas, such as defective contours, edge shapes, etc., to intuitively display the physical features of the defect.

[0069] Structural feature extraction: Use 3D scanning equipment or structural analysis equipment to obtain structural feature parameters such as depth, width, density, etc. of the defective area to provide detailed feature data for subsequent analysis.

[0070] The collected physical and structural characteristic parameters are recorded in the database, and the parameter values ​​of each suspected defective area are normalized for subsequent data analysis.

[0071] Based on the extracted performance parameters and the preset quality standards, the impact of the suspected defective area is quantitatively analyzed: The preset quality standards may include minimum requirements for structural integrity, thickness uniformity, edge strength, etc. By comparing performance parameters and quality standards, the impact of the defect area can be quantified. This analysis can use computational models or simulation technology to evaluate the performance of suspected defect areas under extreme working conditions one by one to ensure that the analysis results are accurate and reliable.

[0072] Quality standard parameter comparison: Compare the extracted performance parameters with the quality standards, for example, compare whether the depth of the defective area exceeds the standard value to determine the degree of impact on the structural integrity.

[0073] Establishment of quantitative analysis model: Establish a quantitative analysis model, take physical and structural characteristic parameters as input variables, calculate the impact index of each suspected defect area, and ensure the scientificity and objectivity of the quantitative analysis.

[0074] Impact degree calculation: Use the calculation model to calculate the impact index of each suspected defect area, and use a numerical value to represent its potential impact on the overall performance of the PCB.

[0075] Compare the assessment results with the quality standards to determine whether the impact of the suspected defective area exceeds the allowable range and make a qualification judgment: If the impact index of a suspected defective area exceeds the preset allowable range, the area is considered to be a defect that has a negative impact on the performance, and the PCB board is judged to be unqualified.

[0076] Qualification criteria: According to the allowable range of the quality standard, set the maximum value of the impact index; if the impact index of a suspected defective area exceeds this value, the area is deemed unqualified.

[0077] Qualification determination: Compare the impact index of each suspected defective area one by one to determine whether it is within the allowable range. If the impact index of all areas is within the standard range, the PCB board is determined to be qualified; if any area exceeds the range, it is determined to be unqualified.

[0078] Mark the unqualified suspected defective areas and record their location information, and generate a final test report for subsequent processing.

[0079] Defective area marking: suspected defective areas that are judged as unqualified are clearly marked so that they can be highlighted in the final test report.

[0080] Location information recording: Record the location information of the unqualified area, including the specific location, shape, size and other detailed parameters of the defect, to provide accurate data for subsequent processing.

[0081] Report generation: Summarize the information of all unqualified areas and generate a final test report, which contains detailed data and qualification results of each unqualified area so that it can be provided to relevant departments for processing and feedback.

[0082] Example 2 The difference between Example 2 of the present invention and Example 1 is that this example introduces a PCB production line defect detection and analysis system based on a recognition model.

[0083] Figure 2 A structural schematic diagram of a PCB production line defect detection and analysis system based on a recognition model is given in the present invention. A PCB production line defect detection and analysis system based on a recognition model includes a plate surface scanning module, a thickness distribution recognition module, a deformation trend analysis module, a polarization reflection analysis module, a suspected defect marking module and a quality standard evaluation module.

[0084] Plate surface scanning module: Scan the entire surface of the PCB plate to be tested, obtain the thickness distribution data of the plate, and generate the corresponding thickness distribution map.

[0085] Thickness distribution identification module: uses the threshold segmentation algorithm to identify areas with uneven thickness and determine whether the thickness change exceeds the preset threshold. If so, the corresponding area is marked as a key detection area.

[0086] Deformation trend analysis module: Analyze the deformation trend of key inspection areas through historical production data to identify potential defects in key inspection areas that are not related to process fluctuations.

[0087] Polarized reflection analysis module: By adjusting the polarization angle of the light source, the reflectivity difference in the key inspection area is analyzed, and the change of the reflection characteristics is evaluated in combination with the material microstructure to identify hidden defects in the key inspection area.

[0088] Suspected defect marking module: Based on the potential defects in the key inspection area that are not related to process fluctuations and the hidden defects in the key inspection area, the suspected defect areas in the key inspection area are preliminarily detected.

[0089] Quality standard assessment module: Further analyze the suspected defective areas according to the preset quality standards to determine the degree of impact on PCB performance.

[0090] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0091] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or may be transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0092] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0093] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0094] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0095] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0096] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0097] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0098] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0099] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A PCB production line defect detection and analysis method based on a recognition model, characterized in that: The steps include: Scan the entire surface of the PCB board to be inspected, obtain the thickness distribution data of the board, and generate the corresponding thickness distribution map; The threshold segmentation algorithm is used to identify the area with uneven thickness and determine whether the thickness change exceeds the preset threshold. If so, the corresponding area is marked as the key detection area; Analyze the deformation trend of key inspection areas through historical production data to identify potential defects in key inspection areas that are not related to process fluctuations; By adjusting the polarization angle of the light source to analyze the reflectivity difference in the key inspection area, combined with the material microstructure to evaluate the change in reflective characteristics, hidden defects in the key inspection area can be identified; Based on the potential defects unrelated to process fluctuations and hidden defects in the key inspection area, the suspected defective areas in the key inspection area are preliminarily detected; Further analysis is performed on suspected defective areas based on preset quality standards to determine the extent of their impact on PCB performance.

2. The PCB production line defect detection and analysis method based on the recognition model according to claim 1 is characterized in that: Scan the entire surface of the PCB board to be tested, obtain the thickness distribution data of the board, and generate the corresponding thickness distribution map, specifically: Start the scanning device to ensure that it can cover the entire surface of the PCB to be tested, and set the scanning resolution and sampling frequency; Use scanning equipment to scan the entire surface of the PCB board to obtain the original thickness data of each position; De-noising and normalizing the collected raw thickness data; Based on the processed thickness data, the corresponding thickness distribution map is generated to intuitively display the thickness change trend of each area.

3. The PCB production line defect detection and analysis method based on recognition model according to claim 2 is characterized in that: The threshold segmentation algorithm is used to identify the area with uneven thickness and determine whether the thickness change exceeds the preset threshold. If so, the corresponding area is marked as the key detection area. Specifically: Perform data analysis on the generated thickness distribution map, extract the thickness value of each area from the map, and convert it into a digital matrix; Select a threshold segmentation algorithm suitable for thickness detection, and set a thickness detection threshold suitable for the PCB board; The thickness data is processed using a threshold segmentation algorithm to identify areas with significant thickness changes; the area where the value in the thickness distribution is significantly higher or lower than the normal range is analyzed using the threshold segmentation algorithm to determine the area with uneven thickness; For the identified uneven thickness area, determine whether its thickness change exceeds its corresponding preset threshold; if the thickness change exceeds its corresponding preset threshold, mark the area as a key detection area.

4. The PCB production line defect detection and analysis method based on recognition model according to claim 3 is characterized in that: Analyze the deformation trend of key inspection areas through historical production data, and identify potential defects in key inspection areas that are not related to process fluctuations, specifically: Extract historical production thickness data of key inspection areas; Calculate the dynamic deformation index of the key inspection area based on historical production thickness data, and quantify the deformation amplitude and trend based on the dynamic deformation index; A multi-dimensional recursive neural network model was established, and the dynamic deformation index was input to perform time series modeling of the deformation trend; A multi-dimensional recursive neural network model is used to predict the deformation trend of key inspection areas; Abnormal deformation that is not related to process fluctuations can be identified based on the predicted deformation trend, and potential defect areas of deformation can be identified.

5. The PCB production line defect detection and analysis method based on recognition model according to claim 4 is characterized in that: The dynamic deformation index of the key inspection area is calculated based on the historical production thickness data, and the deformation amplitude and trend are quantified based on the dynamic deformation index, specifically: The dynamic deformation index is defined as the thickness change rate in time and space, and is calculated as follows: ;in, Indicates The dynamic deformation index of the batch, For the Batch No. Thickness measurements, Indicates The average thickness of the batch, represents the number of locations where thickness measurements are made in each production batch, is the batch sequence in time, is the measurement position within each batch; Calculate the increment: ;in, It is an increment.

6. The PCB production line defect detection and analysis method based on recognition model according to claim 5 is characterized in that: By adjusting the polarization angle of the light source to analyze the reflectivity difference in the key inspection area, combined with the material microstructure to evaluate the change in reflective characteristics, hidden defects in the key inspection area can be identified, specifically: Adjust the polarization angle of the light source according to the material characteristics of the key detection area; Collect multi-angle reflection data of key detection areas at different polarization angles; Analyze multi-angle reflection data in combination with material microstructural properties to identify abnormal changes in reflection characteristics; Based on abnormal changes in reflection characteristics, hidden defect areas within the key inspection area are identified.

7. The PCB production line defect detection and analysis method based on recognition model according to claim 6 is characterized in that: Based on the potential defects unrelated to process fluctuations in the key inspection area and the hidden defects in the key inspection area, the suspected defective areas in the key inspection area are preliminarily detected, specifically: In the key inspection area, the area corresponding to the union of the deformation potential defect area and the hidden defect area is marked as a suspected defect area.

8. The PCB production line defect detection and analysis method based on recognition model according to claim 7 is characterized in that: Further analysis is performed on the suspected defective areas according to the preset quality standards to determine the degree of impact on the PCB performance, specifically: Conduct preliminary screening of the marked suspected defect areas to determine the detection parameters that meet the preset quality standards; For the suspected defective areas after screening, performance parameters are extracted, including physical characteristics and structural characteristics; Based on the extracted performance parameters and the preset quality standards, the impact of the suspected defective area is quantitatively analyzed; Compare the assessment results with the quality standards to determine whether the impact of the suspected defective area exceeds the allowable range and make a qualification determination; Mark the unqualified suspected defective areas and record the location information, and generate a test report.

9. A PCB production line defect detection and analysis system based on a recognition model, used to implement a PCB production line defect detection and analysis method based on a recognition model as claimed in any one of claims 1 to 8, characterized in that: It includes plate surface scanning module, thickness distribution recognition module, deformation trend analysis module, polarization reflection analysis module, suspected defect marking module and quality standard assessment module; Plate surface scanning module: scans the entire surface of the PCB plate to be tested, obtains the thickness distribution data of the plate, and generates the corresponding thickness distribution map; Thickness distribution identification module: uses the threshold segmentation algorithm to identify the area with uneven thickness and judge whether the thickness change exceeds the preset threshold. If so, the corresponding area is marked as the key detection area; Deformation trend analysis module: Analyzes the deformation trend of key inspection areas through historical production data, and identifies potential defects in key inspection areas that are not related to process fluctuations; Polarized reflection analysis module: By adjusting the polarization angle of the light source, the reflectivity difference in the key inspection area is analyzed, and the change of the reflection characteristics is evaluated in combination with the microstructure of the material to identify hidden defects in the key inspection area; Suspected defect marking module: Based on potential defects unrelated to process fluctuations and hidden defects in key inspection areas, it preliminarily detects suspected defect areas in key inspection areas; Quality standard assessment module: Further analyze the suspected defective areas according to the preset quality standards to determine the degree of impact on PCB performance.

Images