Visual Inspection-Based Anomaly Detection Method for Industrial Circuit Boards

Abstract

This invention relates to the field of industrial circuit board anomaly monitoring technology and discloses a vision-based industrial circuit board anomaly monitoring method, comprising the following steps: Step 1: Acquire image data and performance test data of all circuit boards, and classify them into image datasets and performance datasets; Step 2: Based on the image datasets, evaluate the appearance defects of each circuit board, generate corresponding bridging area ratios, color deviations, and pin deviations, and construct a three-dimensional defect scoring matrix by comprehensively utilizing image processing, color analysis, and geometric measurement technologies; Step 3: Based on the performance datasets, evaluate the functional defects of each circuit board, generate corresponding functional data sets, achieving high accuracy in multi-dimensional monitoring; Step 4: Based on the image datasets, bridging area ratios, color deviations, pin deviations, and functional data sets, classify the defect types of each circuit board into connection defects, scratch defects, and electrical defects, and output corresponding control commands, achieving high efficiency in intelligent repair.

Description

Technical Field

[0001] This invention relates to the field of industrial circuit board anomaly monitoring technology, specifically to a vision-based method for industrial circuit board anomaly monitoring. Background Technology

[0002] Industrial circuit boards (PCBs) are a crucial component of modern industrial production, and their performance and quality directly impact the stability and efficiency of the entire production process. During production, PCBs are prone to connection defects. When a PCB has poor soldering, it leads to unstable and intermittent circuit connections, affecting signal transmission and potentially causing intermittent equipment failures. Bridging defects can cause short circuits between different circuits, damaging components and potentially leading to safety accidents, endangering personnel and equipment on the production floor. Timely detection of PCB anomalies is crucial. It effectively prevents further deterioration of faults. Accurate detection and proper handling of defects as they appear significantly reduce maintenance costs and minimize production losses due to equipment downtime. Anomaly detection also allows for the early identification of potential safety hazards, preventing catastrophic accidents such as fires and explosions caused by PCB failures, thus creating a safe and reliable working environment for industrial production and effectively protecting personnel and property.

[0003] Currently, traditional methods for monitoring anomalies on industrial circuit boards typically rely on manual visual inspection. The results depend heavily on the subjective experience of the inspectors, which can easily lead to missed defects and misjudgments, resulting in a high rate of missed defects. Furthermore, manual visual inspection often struggles to accurately detect minor defects such as cold solder joints and bridging defects, while manual repair is inefficient. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a vision-based method for monitoring anomalies in industrial circuit boards. This method has advantages such as high accuracy in multi-dimensional monitoring and high efficiency in intelligent repair, solving the problems of high false negative rates and low efficiency in manual repair of traditional industrial circuit board anomaly monitoring methods.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a visual inspection-based method for monitoring anomalies in industrial circuit boards, comprising the following steps:

[0006] Step 1: Connect a high-definition camera and a robotic arm via the network to acquire image data and performance test data of all circuit boards, and classify them into image datasets and performance datasets;

[0007] Step 2: Based on the image dataset, evaluate the appearance defects of each circuit board and generate the corresponding bridging area percentage. Color deviation and pin deviation ;

[0008] Step 3: Based on the performance dataset, evaluate the functional defects of each circuit board and generate the corresponding functional data set. ;

[0009] Step 4: Based on the image dataset and bridging area ratio Color deviation Pin deviation and functional data groups The system categorizes the defects of each circuit board into connection defects, scratch defects, and electrical defects, and outputs corresponding control commands.

[0010] Preferably, in step one, the image dataset includes pad images, polarity identification images, component images, and copper foil conductor images for each circuit board.

[0011] Preferably, in step one, the performance dataset includes the measured voltage, measured current, and measured resistance of each circuit board.

[0012] Preferably, in step two, the bridging area accounts for... The calculation process is as follows:

[0013] Based on the image dataset, extract the first... Image data of each circuit board, randomly selecting adjacent connected pads and solder pads The image is then processed using image processing software to identify the first... In each circuit board, the pads and pads The bridging area between them, and the pads The area is marked as , solder pad The area is marked as , solder pad and solder pads The area of ​​the bridging region between them is marked as ;

[0014]

[0015] In the formula, Indicates adjacent pads and pads Total area, Indicates the first In each circuit board, the pads and pads The proportion of bridging area between .

[0016] Preferably, in step two, the color deviation The calculation process is as follows:

[0017] Based on the image dataset, the first The polarity identification image of each circuit board is marked as follows ;

[0018] Using image processing software, the first Polarity marking image of each circuit board Convert from RGB color space to HSV color space, and identify the first... Polarity marking image of each circuit board The chromaticity angle in the figure, and marked as ;

[0019]

[0020] In the formula, This represents the standard value used to measure the chromaticity angle. This represents the absolute difference between the standard value and the chromaticity angle, which is the first value. Color deviation of polarity markings on individual circuit boards .

[0021] Preferably, in step two, the pin deviation The calculation process is as follows:

[0022] Based on the image dataset, the first The component images of the circuit boards are labeled as follows ;

[0023] Using image processing software, identify the first... Image of a circuit board component The pin height and solder joint height are specified, and the pin height is marked as... Mark the solder joint height as ;

[0024]

[0025] In the formula, This represents the absolute difference between the pin height and the solder joint height, which is the [number]th [pin height]. Pin deviation of individual circuit board components .

[0026] Preferably, in step three, the functional data group The calculation process is as follows:

[0027] Based on the performance dataset, extract the first... The electrical performance test data of the first circuit board, and the first The measured voltage of each circuit board is marked as follows: , will the The measured current of each circuit board is marked as follows: , will the The measured resistance of each circuit board is marked as follows: ;

[0028]

[0029] In the formula, Indicates the rated voltage. This represents the difference between the measured voltage and the rated voltage. Indicates the rated current. This represents the difference between the measured current and the rated current. Indicates the rated resistance. This represents the difference between the measured resistance and the rated resistance. Indicates the first Functional data groups for each circuit board.

[0030] Preferably, in step four, the connection defect assessment process is as follows:

[0031] In the In a circuit board, if the bridging area between any two adjacent pads accounts for a certain percentage If the percentage is greater than 15%, it indicates that the circuit board has a bridging defect. The bridging area should be controlled for rework by the robotic arm, and excess solder should be removed using a desoldering pump.

[0032] If the first Color deviation of individual circuit boards >10° indicates a polarity deviation defect in the circuit board, and the polarity marking of the robotic arm should be controlled for rework.

[0033] If the first Individual circuit board pin deviation A value >0.1mm indicates a poor solder joint defect on the circuit board, and the robotic arm should be controlled to rework the solder joint.

[0034] Preferably, in step four, the scratch defect assessment process is as follows:

[0035] Based on the image dataset, the first The copper foil conductor image of the circuit board is marked as ;

[0036] Using image processing software, identify the first... Image of copper foil conductors on a circuit board In the diagram, the grayscale value of each pixel is used for the first... Texture features of the circuit board;

[0037] By using a database, standard texture features of the circuit board are obtained, and then the first... Image of copper foil wires on a circuit board Texture features in the image are compared with standard texture features;

[0038] If the first Image of copper foil wires on a circuit board If the texture features in the circuit board differ from the standard texture features, it indicates that the circuit board has scratches or defects, and the robotic arm should be controlled to rework the copper foil wires.

[0039] Preferably, in step four, the electrical defect assessment process is as follows:

[0040] In the functional data group If any value is greater than 0, it indicates that the circuit board has an electrical defect and a high risk of short circuit. The robotic arm should be controlled to repair the circuit board.

[0041] Compared with existing technologies, this invention provides a vision-based method for monitoring anomalies in industrial circuit boards, which has the following advantages:

[0042] 1. This invention uses a network connection to a high-definition camera and a robotic arm to acquire image data and performance test data of all circuit boards, and classifies them into image datasets and performance datasets. Based on the image datasets, it evaluates the appearance defects of each circuit board and generates the corresponding bridging area percentage. Color deviation and pin deviation By comprehensively utilizing image processing, color analysis, and geometric measurement technologies, a three-dimensional defect scoring matrix is ​​constructed. Based on the performance dataset, the functional defects of each circuit board are evaluated, and corresponding functional data sets are generated. It not only focuses on the appearance defects on the surface of the circuit board, but also analyzes the deviation of its electrical performance indicators in depth, so as to achieve a multi-dimensional comprehensive evaluation of the quality of the circuit board. This is more conducive to discovering potential quality problems and providing richer and more accurate information for subsequent quality control and improvement. The multi-dimensional monitoring has high accuracy.

[0043] 2. This invention utilizes image datasets and bridging area ratios. Color deviation Pin deviation and functional data groups The system categorizes defects in each circuit board into connection defects, scratch defects, and electrical defects, and outputs corresponding control commands. With the help of image processing software and preset algorithms, it realizes automated monitoring and rework processes, enabling continuous and intelligent operation of circuit board production, improving production efficiency, reducing manual labor intensity, adapting to the needs of modern industrial large-scale production, and achieving high intelligent repair efficiency. Attached Figure Description

[0044] Figure 1 This is a diagram illustrating the steps of the method of the present invention. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] Traditional industrial circuit board anomaly detection methods typically rely on manual visual inspection. The results depend heavily on the inspector's subjective experience, easily leading to missed or misjudged defects, resulting in a high rate of missed detections. Furthermore, manual visual inspection often struggles to accurately detect subtle solder joint defects and bridging defects, and manual repair is inefficient. Therefore, please refer to [the relevant documentation / reference needed]. Figure 1 This invention provides a vision-based method for monitoring anomalies in industrial circuit boards, comprising the following steps:

[0047] Step 1: Connect a high-definition camera and a robotic arm via the network to acquire image data and performance test data of all circuit boards, and classify them into image datasets and performance datasets;

[0048] The image dataset includes pad images, polarity identification images, component images, and copper trace images for each circuit board;

[0049] The performance dataset includes the measured voltage, measured current, and measured resistance for each board.

[0050] Specifically, by acquiring various image data of the circuit board through a high-definition camera and combining it with performance test data, the appearance and functional defects of the circuit board can be comprehensively and meticulously evaluated.

[0051] Step 2: Based on the image dataset, evaluate the appearance defects of each circuit board and generate the corresponding bridging area percentage. Color deviation and pin deviation ;

[0052] Bridging area ratio The calculation process is as follows:

[0053] Based on the image dataset, extract the first... Image data of each circuit board, randomly selecting adjacent connected pads and pads The image is then processed using image processing software to identify the first... In each circuit board, the pads and pads The bridging area between them, and the pads The area is marked as , solder pad The area is marked as , solder pad and pads The area of ​​the bridging region between them is marked as ;

[0054]

[0055] In the formula, Indicates adjacent pads and pads Total area, Indicates the first In each circuit board, the pads and pads The proportion of bridging area between ;

[0056] Color deviation The calculation process is as follows:

[0057] Based on the image dataset, the first The polarity identification image of each circuit board is marked as follows ;

[0058] Using image processing software, the first Polarity marking image of each circuit board Convert from RGB color space to HSV color space, and identify the first... Polarity marking image of each circuit board The chromaticity angle in the figure, and marked as ;

[0059]

[0060] In the formula, This represents the standard value used to measure the chromaticity angle. This represents the absolute difference between the standard value and the chromaticity angle, which is the first value. Color deviation of polarity markings on individual circuit boards ;

[0061] Pin deviation The calculation process is as follows:

[0062] Based on the image dataset, the first The component images of the circuit boards are labeled as follows ;

[0063] Using image processing software, identify the first... Image of a circuit board component The pin height and solder joint height are specified, and the pin height is marked as... Mark the solder joint height as ;

[0064]

[0065] In the formula, This represents the absolute difference between the pin height and the solder joint height, which is the [number]th [pin height]. Pin deviation of individual circuit board components ;

[0066] Specifically, by comprehensively utilizing technologies such as image processing, color analysis, and geometric measurement, a three-dimensional defect scoring matrix is ​​constructed, which can effectively identify common defects on circuit boards. Each defect type has a clearly defined judgment threshold to ensure the accuracy and consistency of the detection results.

[0067] Step 3: Based on the performance dataset, evaluate the functional defects of each circuit board and generate the corresponding functional data set. ;

[0068] Functional Data Group The calculation process is as follows:

[0069] Based on the performance dataset, extract the first... The electrical performance test data of the first circuit board, and the first The measured voltage of each circuit board is marked as follows: , will the The measured current of each circuit board is marked as follows: , will the The measured resistance of each circuit board is marked as follows: ;

[0070]

[0071] In the formula, Indicates the rated voltage. This represents the difference between the measured voltage and the rated voltage. Indicates the rated current. This represents the difference between the measured current and the rated current. Indicates the rated resistance. This represents the difference between the measured resistance and the rated resistance. Indicates the first Functional data groups for each circuit board;

[0072] Specifically, it not only focuses on the appearance defects on the surface of the circuit board, but also analyzes the deviation of its electrical performance indicators in depth, so as to achieve a multi-dimensional comprehensive evaluation of the quality of the circuit board. This is more conducive to discovering potential quality problems and providing richer and more accurate information for subsequent quality control and improvement. The multi-dimensional monitoring has high accuracy.

[0073] Step 4: Based on the image dataset and bridging area ratio Color deviation Pin deviation and functional data groups The system categorizes the defects of each circuit board into connection defects, scratch defects, and electrical defects, and outputs corresponding control commands.

[0074] The connection defect assessment process is as follows:

[0075] In the In a circuit board, if the bridging area between any two adjacent pads accounts for a certain percentage If the percentage is greater than 15%, it indicates that the circuit board has a bridging defect. The bridging area should be controlled for rework by the robotic arm, and excess solder should be removed using a desoldering pump.

[0076] If the first Color deviation of individual circuit boards >10° indicates a polarity deviation defect in the circuit board, and the polarity marking of the robotic arm should be controlled for rework.

[0077] If the first Individual circuit board pin deviation >0.1mm indicates a poor solder joint defect on the circuit board, and the robotic arm should be controlled to rework the solder joints;

[0078] The scratch defect assessment process is as follows:

[0079] Based on the image dataset, the first The copper foil conductor image of the circuit board is marked as ;

[0080] Using image processing software, identify the first... Image of copper foil conductors on a circuit board In the diagram, the grayscale value of each pixel is used for the first... Texture features of the circuit board;

[0081] By using a database, standard texture features of the circuit board are obtained, and then the first... Image of copper foil wires on a circuit board Texture features in the image are compared with standard texture features;

[0082] If the first Image of copper foil wires on a circuit board The difference between the texture features in the circuit and the standard texture features indicates that the circuit board has scratches and defects. The robotic arm should be controlled to rework the copper foil wires.

[0083] The electrical defect assessment process is as follows:

[0084] In the functional data group If any value is greater than 0, it indicates that the circuit board has an electrical defect and a high risk of short circuit. The robotic arm should be controlled to repair the circuit board.

[0085] Specifically, by using image processing software and preset algorithms, an automated monitoring and rework process is achieved, enabling continuous and intelligent operation of circuit board production, improving production efficiency, reducing manual labor intensity, adapting to the needs of modern industrial large-scale production, and achieving high intelligent repair efficiency.

[0086] Example 1

[0087] In this experiment, a circuit board with two solder pads was selected as the experimental subject. Testing revealed that the solder pads... The area is 4mm², and the pad is... The area is 6mm², and the bridging area between the two pads is 2mm². What is the percentage of the bridging area on this circuit board? The calculation process is as follows:

[0088]

[0089] In the formula, Indicates adjacent pads and pads Total area, This indicates the pads in the circuit board. and pads The proportion of bridging area between Based on the assessment, the solder pads and pads The proportion of bridging area between If the solder ratio is greater than 15%, it indicates that the circuit board has a bridging defect. The bridging area should be reworked by the robotic arm, and excess solder should be removed using a desoldering pump.

[0090] Example 2

[0091] In this experiment, a circuit board with a polarity indicator was selected as the experimental subject. Testing revealed that the chromaticity angle in the polarity indicator image of this circuit board was 105°, while the standard value for measuring the chromaticity angle was set to 120°. The chromaticity deviation of this circuit board was... The calculation process is as follows:

[0092]

[0093] In the formula, This represents the standard value used to measure the chromaticity angle. The absolute difference between the standard value and the chromaticity angle represents the chromaticity deviation of the circuit board's polarity marking. Upon inspection, the color deviation of the polarity markings on the circuit board was determined to be... A value greater than 10° indicates a polarity deviation defect in the circuit board, and the polarity marking of the robotic arm should be controlled for rework.

[0094] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value, it is acceptable.

[0095] The above formulas are all derived from software simulation using a large amount of data and are selected to be close to the actual values. The coefficients in the formulas are set by those skilled in the art according to the actual situation. The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any equivalent substitutions or changes made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the protection scope of the present invention.

Claims

1. A method for detecting anomalies in industrial circuit boards based on vision inspection, characterized in that, Includes the following steps: Step 1: Connect a high-definition camera and a robotic arm via the network to acquire image data and performance test data of all circuit boards, and classify them into image datasets and performance datasets; Step 2: Based on the image dataset, evaluate the appearance defects of each circuit board and generate the corresponding bridging area percentage. Color deviation and pin deviation ; Bridging area ratio The calculation process is as follows: Based on the image dataset, extract the first... Image data of each circuit board, randomly selecting adjacent connected pads and pads The image is then processed using image processing software to identify the first... In each circuit board, the pads and pads The bridging area between them, and the pads The area is marked as , solder pad The area is marked as , solder pad and pads The area of ​​the bridging region between them is marked as ; In the formula, Indicates adjacent pads and pads Total area, Indicates the first In each circuit board, the pads and pads The proportion of bridging area between ; Color deviation The calculation process is as follows: Based on the image dataset, the first The polarity identification image of each circuit board is marked as follows ; Using image processing software, the first Polarity marking image of each circuit board Convert from RGB color space to HSV color space, and identify the first... Polarity marking image of each circuit board The chromaticity angle in the figure, and marked as ; In the formula, This represents the standard value used to measure the chromaticity angle. This represents the absolute difference between the standard value and the chromaticity angle, which is the first value. Color deviation of polarity markings on individual circuit boards ; Pin deviation The calculation process is as follows: Based on the image dataset, the first The component images of the circuit boards are labeled as follows ; Using image processing software, identify the first... Image of a circuit board component The pin height and solder joint height are specified, and the pin height is marked as... Mark the solder joint height as ; In the formula, This represents the absolute difference between the pin height and the solder joint height, which is the [number]th [pin height]. Pin deviation of individual circuit board components ; Step 3: Based on the performance dataset, evaluate the functional defects of each circuit board and generate the corresponding functional data set. ; Step 4: Based on the image dataset and bridging area ratio Color deviation Pin deviation and functional data groups The system categorizes the defects of each circuit board into connection defects, scratch defects, and electrical defects, and outputs corresponding control commands.

2. The method for monitoring industrial circuit board anomalies based on vision inspection according to claim 1, characterized in that: In step one, the image dataset includes pad images, polarity identification images, component images, and copper foil conductor images for each circuit board.

3. The method for monitoring industrial circuit board anomalies based on vision inspection according to claim 2, characterized in that: In step one, the performance dataset includes the measured voltage, measured current, and measured resistance of each circuit board.

4. The industrial circuit board anomaly monitoring method based on vision inspection according to claim 3, characterized in that: In step three, the functional data group The calculation process is as follows: Based on the performance dataset, extract the first... The electrical performance test data of the first circuit board, and the first The measured voltage of each circuit board is marked as follows: , will the The measured current of each circuit board is marked as follows: , will the The measured resistance of each circuit board is marked as follows: ; In the formula, Indicates the rated voltage. This represents the difference between the measured voltage and the rated voltage. Indicates the rated current. This represents the difference between the measured current and the rated current. Indicates the rated resistance. This represents the difference between the measured resistance and the rated resistance. Indicates the first Functional data groups for each circuit board.

5. The method for detecting anomalies in industrial circuit boards based on vision inspection according to claim 4, characterized in that: In step four, the connection defect assessment process is as follows: In the In a circuit board, if the bridging area between any two adjacent pads accounts for a certain percentage If the percentage is greater than 15%, it indicates that the circuit board has a bridging defect. The bridging area should be controlled for rework by the robotic arm, and excess solder should be removed using a desoldering pump. If the first Color deviation of individual circuit boards >10° indicates a polarity deviation defect in the circuit board, and the polarity marking of the robotic arm should be controlled for rework. If the first Individual circuit board pin deviation A value >0.1mm indicates a poor solder joint defect on the circuit board, and the robotic arm should be controlled to rework the solder joint.

6. The method for detecting anomalies in industrial circuit boards based on vision inspection according to claim 5, characterized in that: In step four, the scratch defect assessment process is as follows: Based on the image dataset, the first The copper foil conductor image of the circuit board is marked as... ; Using image processing software, identify the first... Image of copper foil conductors on a circuit board In the diagram, the grayscale value of each pixel is used for the first... Texture features of a circuit board; By using a database, standard texture features of the circuit board are obtained, and then the first... Image of copper foil wires on a circuit board Texture features in the model are compared with standard texture features; If the first Image of copper foil wires on a circuit board If the texture features in the circuit board differ from the standard texture features, it indicates that the circuit board has scratches or defects, and the robotic arm should be controlled to rework the copper foil wires.

7. The method for detecting anomalies in industrial circuit boards based on vision inspection according to claim 6, characterized in that: In step four, the electrical defect assessment process is as follows: In the functional data group If any value is greater than 0, it indicates that the circuit board has an electrical defect and a high risk of short circuit. The robotic arm should be controlled to repair the circuit board.

Images