A Printed Circuit Board Quality Assessment Method Based on Data Analysis

Abstract

The present invention relates to a method for evaluating the quality of a printed circuit board based on data analysis, and relates to the technical field of data processing. The method comprises: acquiring original image data of a printed circuit board; preprocessing the original image data to obtain a preprocessed image; determining corresponding edge detection parameters according to the preprocessed image; performing edge detection on the preprocessed image according to the edge detection parameters to extract key features of the printed circuit board; comparing the key features of the printed circuit board with a preset defect pattern to determine a defect area; generating a quality evaluation result of the printed circuit board according to the defect area; the present invention can detect the defect area, thereby avoiding missed detection or false detection that may occur in traditional manual detection.

Description

Technical Field

[0001] The invention relates to the technical field of data processing, and in particular to a printed circuit board quality assessment method based on data analysis. Background Art

[0002] Printed circuit boards (PCBs) play a vital role in various electronic devices. Therefore, the quality assessment of printed circuit boards has become a key link in ensuring the performance and reliability of electronic devices. Some traditional printed circuit board quality assessment methods still rely on manual visual inspection or use simple image processing technology, but these methods have many defects in data processing.

[0003] For example, the application of existing image processing technology in printed circuit board quality assessment also has limitations. Some simple image processing technologies may not accurately identify complex defect patterns, resulting in inaccurate assessment results. Summary of the invention

[0004] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a printed circuit board quality assessment method based on data analysis, which can detect defective areas, thereby avoiding missed detection or false detection that may occur in traditional manual detection.

[0005] In order to solve the above technical problems, the basic concept of the technical solution adopted by the present invention is:

[0006] In a first aspect, a printed circuit board quality assessment method based on data analysis comprises:

[0007] Acquire original image data of a printed circuit board;

[0008] Preprocessing the original image data to obtain a preprocessed image;

[0009] According to the preprocessed image, the corresponding edge detection parameters are determined, including: creating an initial population, each individual in the initial population represents a set of edge detection parameters; performing edge detection on the preprocessed image according to each individual in the population to obtain a detection result; determining the corresponding fitness function according to the detection result; according to the fitness function, determining that the corresponding individual enters the next generation, randomly determining two individuals in the population according to a set crossover rate, and exchanging some genes in the two individuals to generate a new individual, wherein the gene represents the edge detection parameter; randomly changing the gene of any individual according to a set mutation rate, repeating the steps until a preset number of iterations is reached to obtain the final edge detection parameters;

[0010] Performing edge detection on the preprocessed image according to edge detection parameters to extract key features of the printed circuit board;

[0011] Compare the key features of the printed circuit board with the preset defect pattern to identify the defect area;

[0012] Based on the defect areas, quality assessment results of printed circuit boards are generated.

[0013] Furthermore, the formula of the fitness function is:

[0014] ;

[0015] in, Represents the fitness value of an individual; , and is the weight coefficient; Indicates the number of real edge pixels successfully identified; Indicates the number of non-edge pixels that are incorrectly identified as edge pixels; Indicates the number of real edge pixels that cannot be identified; Indicates the number of detected edge pixels; Represents the direction angle of the i-th edge pixel; Indicates the index of edge pixels.

[0016] Further, edge detection is performed on the preprocessed image according to the edge detection parameters to extract key features of the printed circuit board, including:

[0017] According to the edge detection parameters;

[0018] pass ;

[0019] Calculate the gradient of the preprocessed image in the horizontal direction;

[0020] and through ;

[0021] Calculate the gradient of the preprocessed image in the vertical direction;

[0022] in, The coordinates in the image are The gray value of the pixel; , , and is a weight coefficient between 0 and 1;

[0023] According to the gradient of the preprocessed image in horizontal and vertical directions;

[0024] pass Calculate the gradient magnitude ;

[0025] in, Represents the gradient weight coefficient in the horizontal direction; Represents the gradient weight coefficient in the vertical direction;

[0026] Compare the gradient magnitude of each pixel in the preprocessed image with a preset threshold. If the gradient magnitude is ≥ the threshold, the corresponding pixel is an edge point.

[0027] According to the edge points, the edge image is determined.

[0028] Furthermore, the key features of the printed circuit board are compared with the preset defect pattern to determine the defect area, including:

[0029] Generate descriptors for key features of the printed circuit board, where the descriptors represent key point information in the preprocessed image;

[0030] Using the FLANN matcher, the feature descriptor of the printed circuit board image is matched with the feature descriptor of each defect image in the defect pattern library to obtain a matching result;

[0031] For each matching result, the similarity between feature descriptors is calculated, and the corresponding feature point pairs are screened according to the set similarity threshold;

[0032] For each preset defect pattern, count the number of feature points that are successfully matched with the printed circuit board image;

[0033] According to the number of matched feature points, the similarity between the printed circuit board image area and each defect pattern is evaluated to obtain an evaluation result;

[0034] Based on the evaluation results, potential defect areas are identified.

[0035] Furthermore, according to the defective area, a quality assessment result of the printed circuit board is generated, including:

[0036] Set a weight value for each defect type;

[0037] Calculate a composite score for each PCB based on the number and severity of defects and their weight values;

[0038] Based on the comprehensive score, a quality assessment result of the printed circuit board is generated.

[0039] Furthermore, the calculation formula for the comprehensive score is:

[0040]

[0041] in, is the number of defects of type i; is the weight value of the i-th defect; is the severity value of the i-th defect; is the total number of defect types, and i represents the index of the defect type.

[0042] In a second aspect, a printed circuit board quality assessment system based on data analysis comprises:

[0043] The acquisition module is used to acquire the original image data of the printed circuit board; preprocess the original image data to obtain a preprocessed image; and determine the corresponding edge detection parameters according to the preprocessed image;

[0044] The processing module is used to perform edge detection on the preprocessed image according to the edge detection parameters to extract the key features of the printed circuit board; compare the key features of the printed circuit board with the preset defect pattern to determine the defect area; and generate the quality assessment result of the printed circuit board according to the defect area.

[0045] After adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art:

[0046] Through refined image processing and data analysis processes, various defects on printed circuit boards can be accurately identified, defective areas can be located, and automated image processing and feature extraction processes can reduce the need for manual participation and speed up detection. Through real-time feedback of printed circuit board quality assessment results, production parameters or process flows can be adjusted to reduce the generation of defective products. Accurate quality assessment can reduce the amount of rework and scrap, thereby saving production costs and adapting to different types of printed circuit boards. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0048] Figure 1 It is a schematic diagram of a printed circuit board quality assessment method based on data analysis of the present invention.

[0049] Figure 2 The present invention is a schematic diagram of a printed circuit board quality assessment system based on data analysis. DETAILED DESCRIPTION

[0050] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present application.

[0051] The following embodiment of the present application takes the printed circuit board quality assessment method based on data analysis as an example to explain the scheme of the present application in detail, but this embodiment cannot limit the protection scope of the present application.

[0052] like Figure 1 As shown, the present invention provides a printed circuit board quality assessment method based on data analysis, comprising:

[0053] Step 11, obtaining original image data of the printed circuit board;

[0054] Step 12, preprocessing the original image data to obtain a preprocessed image;

[0055] Step 13, determining the corresponding edge detection parameters according to the preprocessed image, including: creating an initial population, each individual in the initial population represents a set of edge detection parameters; performing edge detection on the preprocessed image according to each individual in the population to obtain a detection result; determining the corresponding fitness function according to the detection result; determining the corresponding individual to enter the next generation according to the fitness function, randomly determining two individuals in the population according to a set crossover rate, and exchanging some genes in the two individuals to generate a new individual, wherein the gene represents the edge detection parameter; randomly changing the gene of any individual according to a set mutation rate, repeating the steps until a preset number of iterations is reached to obtain the final edge detection parameters;

[0056] Step 14, performing edge detection on the preprocessed image according to the edge detection parameters to extract key features of the printed circuit board;

[0057] Step 15, comparing the key features of the printed circuit board with the preset defect pattern to determine the defect area;

[0058] Step 16: generating a quality assessment result of the printed circuit board according to the defective area.

[0059] In the embodiment of the present invention, through preprocessing, the noise in the image can be removed and the image contrast can be enhanced, thereby improving the image quality; the edge detection parameter setting can ensure the accurate identification of the edge of the printed circuit board; through edge detection, the key features of the printed circuit board can be accurately extracted to provide data support for defect detection and quality assessment; by comparing with the preset defect pattern, the defect area on the printed circuit board can be accurately identified to improve the efficiency and accuracy of quality inspection; the quality assessment results based on the defect area can provide objective and comprehensive quality feedback for the production enterprise, timely discover problems in the production process and make improvements, thereby improving product quality and customer satisfaction. The optimal edge detection parameters are automatically found through the fitness function to avoid the tediousness and uncertainty of manual parameter adjustment; by optimizing the parameters, the edge features of the printed circuit board can be accurately extracted to improve the accuracy of defect detection; the fitness function has a strong global search capability, can adapt to different printed circuit board images and edge detection requirements, and improves production efficiency through an automated parameter optimization process.

[0060] In the specific implementation process of the present invention, an initial population is randomly generated, and each individual in the population is a set of edge detection parameters, wherein the parameters include threshold, filter type, kernel size, and edge detection parameters are used to process the preprocessed image; by using Canny edge detection, the edge of the preprocessed image is extracted for the parameters of each individual; the effect of edge detection is evaluated by calculating the fitness function, and the fitness value is calculated, and then according to the fitness value, according to the set crossover rate of 70%, according to the set mutation rate such as 1%, two individuals in the population are randomly selected, and parts of them, namely edge detection parameters, are exchanged, and the edge detection parameters of any individual can be randomly changed; the above steps are repeated until a preset number of iterations is reached, and in each iteration, excellent individuals are selected according to the fitness function, and new individuals are generated through crossover and mutation. After multiple iterations, a set of optimal edge detection parameters will be obtained.

[0061] In the above step 134, the formula of the fitness function is:

[0062] ;

[0063] in, Represents the fitness value of an individual; , and is the weight coefficient; Indicates the number of real edge pixels successfully identified; Indicates the number of non-edge pixels that are incorrectly identified as edge pixels; Indicates the number of real edge pixels that cannot be identified; Indicates the number of detected edge pixels; Represents the direction angle of the i-th edge pixel; Indicates the index of edge pixels.

[0064] In an embodiment of the present invention, the fitness function comprehensively considers recognition, false detection, missed detection and edge direction, and provides a comprehensive evaluation standard for edge detection parameter optimization; by adjusting the weight coefficient, the relative importance of each indicator can be adjusted according to actual needs, thereby meeting the specific requirements of edge detection performance; the fitness function can effectively search for the optimal edge detection parameter combination, and by optimizing the fitness function, the accuracy of edge detection can be significantly improved, the false detection rate and missed detection rate can be reduced, thereby improving the overall performance of printed circuit board quality assessment.

[0065] In a preferred embodiment of the present invention, the above step 14 may include:

[0066] Step 141, according to edge detection parameters;

[0067] pass ;

[0068] Calculate the gradient of the preprocessed image in the horizontal direction;

[0069] and through ;

[0070] Calculate the gradient of the preprocessed image in the vertical direction;

[0071] in, The coordinates in the image are The gray value of the pixel; , , and is a weight coefficient between 0 and 1;

[0072] Step 142, according to the gradients of the preprocessed image in the horizontal and vertical directions;

[0073] pass Calculate the gradient magnitude ;

[0074] in, Represents the gradient weight coefficient in the horizontal direction; Represents the gradient weight coefficient in the vertical direction;

[0075] Step 143, comparing the gradient amplitude of each pixel in the preprocessed image with a preset threshold, if the gradient amplitude is greater than or equal to the threshold, the corresponding pixel is an edge point;

[0076] Step 144: determine an edge image based on the edge points.

[0077] In an embodiment of the present invention, by calculating the gradients in the horizontal and vertical directions, the edge position in the image can be accurately located; by adjusting the weight coefficient and the threshold, the sensitivity and accuracy of edge detection can be flexibly controlled; the gradient-based edge detection method usually has high computational efficiency and is suitable for real-time or large-scale image processing tasks; the generated edge image can clearly display the edge information in the original image, which is convenient for observing and analyzing the structure and features in the image.

[0078] In the specific implementation process of the present invention, for each pixel in the preprocessed image, the Sobel operator is used to calculate the gradient of the point in the horizontal direction and the vertical direction, respectively, wherein the horizontal gradient reflects the grayscale change rate of the image in the horizontal direction, and the vertical gradient reflects the grayscale change rate in the vertical direction; the calculation formula involves the weighted difference of the neighboring pixels, wherein the weight coefficient G x and G y Used to adjust the importance of gradients in different directions. The weight coefficient is a value between 0 and 1 and can be adjusted according to specific needs. For each pixel, the Pythagorean theorem is used to calculate the gradient amplitude of the point in combination with its horizontal gradient and vertical gradient, where the amplitude represents the intensity and direction of the grayscale change at the point. The horizontal gradient weight coefficient and vertical gradient weight coefficient in the calculation formula are and The contribution of gradients in different directions to the amplitude calculation can be adjusted; a preset threshold is set, which can be adjusted according to the specific content of the image and the sensitivity requirements of edge detection, and the gradient amplitude of each pixel is compared with the preset threshold. If the gradient amplitude is greater than or equal to the threshold, the pixel is marked as an edge point; based on the edge point, a new edge image with a high grayscale value is generated, and the new edge image can clearly show the edge information in the original image.

[0079] In a preferred embodiment of the present invention, the above step 15 includes:

[0080] Step 151, generating a descriptor for a key feature of the printed circuit board, the descriptor representing key point information in the preprocessed image;

[0081] Step 152, using a FLANN matcher, matching the feature descriptor of the printed circuit board image with the feature descriptor of each defect image in the defect pattern library to obtain a matching result;

[0082] Step 153, for each matching result, calculate the similarity between feature descriptors, and select the corresponding feature point pairs according to the set similarity threshold;

[0083] Step 154, for each preset defect pattern, counting the number of feature points that are successfully matched with the printed circuit board image;

[0084] Step 155, evaluating the similarity between the printed circuit board image area and each defect pattern according to the number of matched feature points to obtain an evaluation result;

[0085] Step 156, determining potential defect areas based on the evaluation results.

[0086] In the embodiment of the present invention, the efficiency and accuracy of defect detection are improved by automating the feature extraction and matching process. The feature descriptor and the matching algorithm have strong robustness to image transformation and can accurately identify defects in complex environments. By adjusting the similarity threshold and the scoring function, different types of defects and detection accuracy requirements can be adapted. The defect pattern library is updated and expanded to adapt to new defect types and forms, thereby improving the detection capability of the system. Through automated detection, the cost and time of manual detection can be reduced, and production efficiency and product quality can be improved.

[0087] In the specific implementation process of the present invention, the SIFT algorithm is used to extract key points from the preprocessed printed circuit board image, and for each key point, a corresponding feature descriptor is generated; the feature descriptor of the printed circuit board image is matched with the feature descriptor of each defect image in the pre-established defect pattern library by using the FLANN matcher, and for each matching result, the similarity between the feature descriptors is calculated by using the Euclidean distance method, and a similarity threshold is set, and only feature point pairs with a similarity higher than the threshold are retained; a counter is initialized for each preset defect pattern to record the number of feature points successfully matched with the printed circuit board image; all matched feature point pairs are traversed, and the value of the corresponding counter is increased according to the defect pattern to which they belong; for each defect pattern, the ratio of the number of successfully matched feature points to the total number of feature points is calculated; according to the matching ratio, the similarity between the printed circuit board image area and each defect pattern is evaluated, and the higher the ratio, the greater the similarity; according to actual needs, a similarity threshold is set to determine potential defect areas; all evaluated similarity results are traversed, and areas with similarities exceeding the set threshold are screened out, and these areas are considered to be potential defect areas.

[0088] In a preferred embodiment of the present invention, the above step 16 includes:

[0089] Step 161, setting a weight value for each defect type;

[0090] Step 162, calculating a comprehensive score for each printed circuit board based on the number and severity of defects and weight values;

[0091] Step 163: Generate a quality assessment result of the printed circuit board according to the comprehensive score.

[0092] In an embodiment of the present invention, the quality of each printed circuit board can be evaluated through a comprehensive scoring system; based on the comprehensive score, circuit boards with poor quality can be identified to provide a basis for priority sorting for subsequent repairs; by regularly analyzing the comprehensive scoring data, quality problems in the production process can be discovered, and the production process and process parameters can be adjusted in time to improve product quality; the quality assessment results can provide real-time feedback on the product quality status; and detailed quality assessment reports can increase the transparency of the production process and improve product quality and production reliability.

[0093] In the specific implementation process of the present invention, the influence of different types of defects on the performance and reliability of the printed circuit board is analyzed, and a weight value is assigned to each defect type according to the degree of influence. The setting of the weight value can be determined by a historical data analysis method and saved in a configurable weight table; all detected defects on the printed circuit board are traversed, and for each defect, the corresponding weight value is obtained from the weight table according to its type and severity. According to the number of defects and the weight value, a comprehensive score is calculated for each printed circuit board using a scoring algorithm, and a threshold is set for dividing different quality levels. The calculated comprehensive score is compared with these thresholds to determine the quality level of the printed circuit board, and a detailed quality assessment report is generated.

[0094] In a preferred embodiment of the present invention, the calculation formula of the comprehensive score in the above step 163 is:

[0095]

[0096] in, is the number of defects of type i; is the weight value of the i-th defect; is the severity value of the i-th defect; is the total number of defect types, and i represents the index of the defect type.

[0097] In the embodiment of the present invention, the quality status of the printed circuit board can be quantified more accurately through specific numerical calculations; the comprehensive scoring formula takes into account the number, weight and severity of defects at the same time, and can comprehensively reflect the quality problems of the circuit board. The weight value and severity value in the formula can be adjusted according to actual conditions, making the evaluation more flexible and adaptable to different quality requirements; the comprehensive scoring record can be used as a basis for quality traceability, and the cause can be quickly located and measures can be taken when problems occur.

[0098] like Figure 2 As shown, a printed circuit board quality assessment system 20 based on data analysis includes:

[0099] The acquisition module 21 is used to acquire the original image data of the printed circuit board; preprocess the original image data to obtain a preprocessed image; and determine corresponding edge detection parameters according to the preprocessed image.

[0100] The processing module 22 is used to perform edge detection on the preprocessed image according to the edge detection parameters to extract the key features of the printed circuit board; compare the key features of the printed circuit board with the preset defect pattern to determine the defect area; and generate the quality assessment result of the printed circuit board according to the defect area.

[0101] It should be noted that the system is a system corresponding to the above method, and all implementation methods in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0102] The embodiment of the present invention further provides a computing device, comprising: a processor, a memory storing a computer program, wherein when the computer program is executed by the processor, the method described above is executed. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0103] The embodiment of the present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0104] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A printed circuit board quality assessment method based on data analysis, characterized in that: The method comprises the following steps: Acquire original image data of a printed circuit board; Preprocessing the original image data to obtain a preprocessed image; According to the preprocessed image, the corresponding edge detection parameters are determined, including: creating an initial population, each individual in the initial population represents a set of edge detection parameters; performing edge detection on the preprocessed image according to each individual in the population to obtain a detection result; determining the corresponding fitness function according to the detection result; according to the fitness function, determining that the corresponding individual enters the next generation, randomly determining two individuals in the population according to a set crossover rate, and exchanging some genes in the two individuals to generate a new individual, wherein the gene represents the edge detection parameter; randomly changing the gene of any individual according to a set mutation rate, repeating the crossover and mutation steps until a preset number of iterations is reached to obtain the final edge detection parameters; Performing edge detection on the preprocessed image according to edge detection parameters to extract key features of the printed circuit board; Compare the key features of the printed circuit board with the preset defect pattern to identify the defect area; Generate quality assessment results of printed circuit boards based on defect areas; The formula of the fitness function is: ； in, Represents the fitness value of an individual; , and is the weight coefficient; Indicates the number of real edge pixels successfully identified; Indicates the number of non-edge pixels that are incorrectly identified as edge pixels; Indicates the number of real edge pixels that cannot be identified; Indicates the number of detected edge pixels; Represents the direction angle of the i-th edge pixel; Indicates the index of edge pixels; Based on the edge detection parameters, edge detection is performed on the preprocessed image to extract key features of the printed circuit board, including: According to the edge detection parameters, Calculate the gradient of the preprocessed image in the horizontal direction , and through Calculate the gradient of the preprocessed image in the vertical direction ,in, The coordinates in the image are The gray value of the pixel; , , and is a weight coefficient between 0 and 1; According to the gradient of the preprocessed image in the horizontal and vertical directions, Calculate the gradient magnitude ,in, Represents the gradient weight coefficient in the horizontal direction; Represents the gradient weight coefficient in the vertical direction; Compare the gradient magnitude of each pixel in the preprocessed image with a preset threshold. If the gradient magnitude is ≥ the threshold, the corresponding pixel is an edge point. According to the edge points, an edge image is determined; Compare key features of the PCB to pre-defined defect patterns to identify defect areas, including: Generate descriptors for key features of the printed circuit board, where the descriptors represent key point information in the preprocessed image; use FLANN A matcher matches the feature descriptor of the printed circuit board image with the feature descriptor of each defect image in the defect pattern library to obtain a matching result; For each matching result, the similarity between feature descriptors is calculated, and the corresponding feature point pairs are screened according to the set similarity threshold; For each preset defect pattern, count the number of feature points that are successfully matched with the printed circuit board image; According to the number of matched feature points, the similarity between the printed circuit board image area and each defect pattern is evaluated to obtain an evaluation result; Based on the assessment results, potential defect areas are identified; Generates quality assessment results of printed circuit boards based on defect areas, including: Set a weight value for each defect type; Calculate a composite score for each PCB based on the number and severity of defects and their weight values; Generate a quality assessment result of the printed circuit board according to the comprehensive score; Overall Rating The calculation formula is: ； in, It is The number of defects; It is The weight of the defect; It is The severity value of the defect; is the total number of defect types, An index representing the defect type.

2. A printed circuit board quality assessment system based on data analysis, the system implementing the method according to claim 1, characterized in that: include: An acquisition module, used for acquiring original image data of a printed circuit board; Preprocessing the original image data to obtain a preprocessed image; According to the preprocessed image, determine the corresponding edge detection parameters; A processing module, used for performing edge detection on the pre-processed image according to edge detection parameters to extract key features of the printed circuit board; Compare the key features of the printed circuit board with the preset defect pattern to identify the defect area; Based on the defect areas, quality assessment results of printed circuit boards are generated.

3. A computing device, characterized in that: include: one or more processors; A storage device, used for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method as claimed in claim 1.

4. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, which implements the method according to claim 1 when executed by a processor.

Images