Automatic PCB (Printed Circuit Board) detection system based on machine vision

Abstract

The invention discloses an automatic PCB (Printed Circuit Board) detection system based on machine vision, which belongs to the technical field of electronic manufacturing and comprises an image information acquisition module used for acquiring image information of a normal welding spot of a PCB and marking the image information as reference image information; the reference image information comprises shape features, position features, texture features, color features and connectivity features of the welding spots; the image processing module responds to the reference image information and is used for preprocessing the reference image information, and preprocessing comprises graying, binaryzation and filtering denoising of the reference image information; and the image acquisition module is used for acquiring a welding spot image on the PCB to be detected and comparing the welding spot image with the reference image information. Through multi-feature analysis, dynamic reference value adjustment and misjudgment correction mechanisms, the quality of the welding spot can be comprehensively evaluated, misjudgment caused by single feature detection is avoided, and the accuracy and reliability of a detection result are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic manufacturing, and in particular to a PCB automatic detection system based on machine vision. Background Art

[0002] In modern electronics manufacturing, the soldering quality of printed circuit boards (PCBs) is directly related to the performance, reliability, and safety of electronic products. As electronic products continue to move toward miniaturization and higher density, traditional manual inspection methods are no longer able to meet the high-precision and high-efficiency requirements for PCB soldering quality testing.

[0003] Regarding this research, application publication number CN119804490A provides a printed circuit board defect detection method and system based on image recognition. This technical solution involves preprocessing the original PCB image to obtain a preprocessed PCB image; fusing scale features on the preprocessed PCB image to obtain a multi-scale feature map; learning background texture features from the multi-scale feature map to obtain a background suppression feature map; and performing background texture suppression based on the background suppression feature map to obtain a significant defect map. This technical solution solves the problem of complex background texture interference and improves the accuracy and robustness of PCB defect detection.

[0004] Another application document, with application number CN202011361716.6, provides a machine vision-based automatic PCB board defect detection system. This technical solution includes a camera module, a loading and unloading module, a fill light module, and an image processing module. The steps are as follows: the PCB board to be inspected enters the inspection area through the loading and unloading module, and is then positioned by the positioning sensor in the inspection area. When the positioning sensor locates the PCB board to be inspected and reaches the corresponding position, the fill light module is turned on to fill light on the PCB board, and at the same time, the camera module takes pictures of the PCB board. This technical solution can determine the location of unqualified areas by independently detecting images of multiple local areas, which is conducive to data statistics to determine whether it is a problem with the production line and facilitates maintenance of the production line.

[0005] Solder joints on PCBs play a crucial role in connecting electronic components and circuits. Poor soldering quality, such as cold solder joints (incomplete fusion due to insufficient soldering temperature), cold solder joints (appearing to be connected on the surface but not actually connected internally), or excessive soldering (potentially causing short circuits or poor appearance), can lead to circuit failure, signal transmission interruptions, and even failure of the entire electronic product. Therefore, accurate testing of PCB solder joints is crucial for ensuring the quality of electronic products.

[0006] However, existing technologies for identifying defects in solder joints on PCBs are still limited and are highly sensitive to small changes in solder joints. This makes it highly sensitive to small acceptable changes in solder joints during automatic inspection of PCBs based on machine vision, which can easily lead to false positives and reduce detection efficiency and reliability. Summary of the Invention

[0007] In view of the above problems existing in the existing electronic manufacturing technology field, the present invention is proposed.

[0008] Therefore, one of the objectives of the present invention is to provide a machine vision-based automatic PCB board inspection system, which can comprehensively evaluate the quality of solder joints through multi-feature analysis, dynamic reference value adjustment and misjudgment correction mechanism, avoid misjudgment caused by single feature detection, and ensure the accuracy and reliability of the inspection results.

[0009] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0010] The present invention provides a PCB automatic detection system based on machine vision, comprising:

[0011] An image information acquisition module is used to acquire image information of a normal solder joint of a PCB board and mark the image information as reference image information; the reference image information includes shape features, position features, texture features, color features, and connectivity features of the solder joint;

[0012] An image processing module, responsive to the reference image information, configured to preprocess the reference image information, the preprocessing comprising grayscale conversion, binarization, and filtering and denoising the reference image information;

[0013] An image acquisition module is used to acquire an image of a solder joint on a PCB to be inspected and compare the image of the solder joint with the reference image information;

[0014] An image fusion processing module, which responds to the result of comparing the solder joint image with the reference image information and is used to process the solder joint image according to the comparison result; the image fusion processing module includes a defect recognition unit, a distinction unit, and a judgment unit;

[0015] The defect recognition unit is used to identify defects in the solder joint image based on the comparison result, and the defects include but are not limited to shape defects, position defects, texture defects, color defects and connectivity defects;

[0016] The distinguishing unit is used to distinguish the shape feature, position feature, texture feature, color feature and connectivity feature respectively;

[0017] The judgment unit responds to the distinction of the shape feature, position feature, texture feature, color feature and connectivity feature by the distinction unit, and is used to judge the solder joint image collected from the PCB board in a future time period. If the collected solder joint image is different from any one of the shape feature, position feature, texture feature, color feature and connectivity feature, the system determines that the PCB board corresponding to the solder joint image is an abnormal PCB board and issues an alarm; otherwise, no judgment is made.

[0018] As a preferred solution of the present invention, wherein: in the image information acquisition module, the shape features include the solder joint contour, solder joint size and solder joint shape symmetry;

[0019] The position features include the coordinates of the soldering point and the relative position of the soldering point and surrounding components;

[0020] The texture characteristics include the surface texture of the solder joint and the uniformity of the texture;

[0021] The color characteristics include solder joint color and color variation;

[0022] The connectivity features include the connection between the solder joints and the circuits;

[0023] The solder joint size includes the diameter, height and width of the solder joint.

[0024] As a preferred embodiment of the present invention, the distinguishing unit distinguishes the shape features, including distinguishing the shape features into circular, elliptical, and irregular shapes according to the solder joint contour; if the solder joint corresponding to the collected solder joint image is an irregular shape, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board;

[0025] Distinguishing the positional features, including distinguishing the positional features based on the relative positions of the solder joint and surrounding components, wherein the distinguishing method includes distinguishing the positional features based on the relative positional relationship between the solder joint and the adjacent components; if the relative positions of the solder joint corresponding to the captured solder joint image and the adjacent components deviate, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board;

[0026] Differentiating the texture features, including differentiating the texture features into textured and non-textured based on the surface texture of the solder joints, wherein if both textured and non-textured solder joints appear on the same PCB board, the system determines that the PCB board is an abnormal PCB board;

[0027] Differentiating the color features, including differentiating the color features based on the color of the solder joints, wherein the color of the solder joints is the same color as the soldering material; if solder joints of different colors appear on the same PCB board, the system determines that the PCB board is an abnormal PCB board;

[0028] The connectivity features are distinguished, including distinguishing the connectivity features based on the connection status of the solder joints and the circuits.

[0029] As a preferred solution of the present invention, the solder joint is distinguished based on the relative positional relationship between the solder joint and adjacent components, and the distinguishing method includes distinguishing the components adjacent to the solder joint into a first component, a second component, and a third component, calculating the distance between the solder joint and the first component, the second component, and the third component, and marking the distance as a reference distance; among the distinguished components, if the distance between the solder joint and any one of the components is greater than and / or less than the reference distance, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board; otherwise, no determination is made.

[0030] As a preferred solution of the present invention, the connectivity features are distinguished based on the connection between the solder joint and the circuit, and the distinguishing method includes counting the number of circuits connected to the solder joint, and dividing the circuits into θ1, θ2, ..., θ n , where θ n Indicates the nth line that has been distinguished. The connection status between the line and the solder joint is collected in each distinguished line. If any line is disconnected from the solder joint, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board; otherwise, no determination is made.

[0031] As a preferred embodiment of the present invention, if the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board, the system obtains the circuit corresponding to the disconnection of the solder joint from the distinguished circuits, obtains a length of at least 1 cm of the circuit based on the edge of the solder joint, obtains the width of the circuit from the length, and equally divides the width into an upper layer width, a middle layer width, and a lower layer width. If any one of the width layers is connected to the solder joint, the system cancels the determination that the PCB board corresponding to the solder joint is an abnormal PCB board; otherwise, the original determination is maintained.

[0032] As a preferred solution of the present invention, if the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board, the diameter of the solder joint is obtained, a center point about the solder joint is selected from the diameter, and the distance from the center point to the edge of the solder joint is calculated. The calculation method includes:

[0033] Dividing the edge of the weld into at least 6 edge points according to the diameter;

[0034] Calculating the distance between the center point and each edge point;

[0035] If the distances to each edge point are the same, the system cancels the determination that the PCB board corresponding to the solder joint is an abnormal PCB board; otherwise, the original determination is maintained.

[0036] As a preferred solution of the present invention, the spacing between the solder joint and the first component, the second component and the third component is re-acquired based on the diameter, and the reference spacing is re-assigned based on the re-acquired spacing, and the re-assigned reference spacing is marked as the second reference spacing. If the spacing between the solder joint and the first component, the second component and the third component acquired in a future time period is equal to the reference spacing and / or the second reference spacing, the system determines that the PCB board corresponding to the solder joint is a normal PCB board; otherwise, no determination is made.

[0037] As a preferred solution of the present invention, the diameter is marked as a reference diameter, and the diameter corresponding to the solder joint is obtained based on the reference spacing, and this corresponding diameter is marked as a second reference diameter. If the diameter of the solder joint obtained in a future time period is consistent with the reference diameter and / or the second reference diameter, the system determines that the PCB board corresponding to the solder joint is a normal PCB board; otherwise, no determination is made.

[0038] 1. The present invention comprehensively analyzes the shape, position, texture, color, and connectivity characteristics of solder joints to comprehensively evaluate their quality, avoiding misjudgments caused by single feature detection, thereby improving the accuracy of defect identification.

[0039] 2. By distinguishing and comparing each feature in detail, it can accurately identify abnormal solder joints, such as those with irregular shapes, position deviations, uneven textures, abnormal colors, or connectivity issues. Furthermore, through comprehensive comparison and verification of multiple features, it reduces misjudgments caused by abnormal single features. For example, even if a feature of a solder joint (such as color) changes slightly, if other features are normal, the system can avoid misjudgment through comprehensive judgment.

[0040] 3. During the detection process, the system can dynamically adjust the reference value (such as reference spacing, reference diameter, etc.) according to actual conditions, further improving the accuracy and adaptability of the detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort. Among them:

[0042] Figure 1 Schematic diagram of the modular structure of a machine vision-based PCB automatic detection system according to an embodiment of the present invention;

[0043] Figure 2 A schematic diagram of the process structure of an embodiment of the present invention;

[0044] Numbers in the figure: 110 - image information acquisition module; 120 - image processing module; 130 - image acquisition module; 140 - image fusion processing module; 1401 - defect recognition unit; 1402 - distinction unit; 1403 - judgment unit. DETAILED DESCRIPTION

[0045] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of 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 described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of the present invention.

[0046] Because existing technologies still have limitations in identifying defects in solder joints on PCB boards and are highly sensitive to slight changes in solder joints, automatic inspection of PCB boards based on machine vision is highly sensitive to slight acceptable changes in solder joints, which can easily lead to false alarms and reduce detection efficiency and reliability.

[0047] Based on this, the present invention proposes a PCB board automatic inspection system based on machine vision. Through multi-feature analysis, dynamic reference value adjustment and misjudgment correction mechanism, it can comprehensively evaluate the quality of solder joints, avoid misjudgment caused by single feature detection, and ensure the accuracy and reliability of the inspection results.

[0048] The present invention will be further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0049] Reference Figures 1 to 2 , is an embodiment of the present invention, which provides a PCB board automatic detection system based on machine vision, including:

[0050] The image information acquisition module 110 is used to acquire image information of a normal solder joint of a PCB board and mark the image information as reference image information; the reference image information includes shape features, position features, texture features, color features, and connectivity features of the solder joint;

[0051] In this embodiment, a normal PCB board is a PCB board with normal welding quality;

[0052] An image processing module 120 is responsive to the reference image information and is used to pre-process the reference image information. The pre-processing includes gray-scaling, binarization, and filtering to remove noise from the reference image information.

[0053] In this embodiment, the contrast and clarity of the image are enhanced;

[0054] The image acquisition module 130 is used to acquire an image of a solder joint on the PCB to be inspected and compare the solder joint image with reference image information;

[0055] In this embodiment, a high-resolution industrial camera and an appropriate light source are used to capture images of the PCB. The camera is mounted on an adjustable bracket, capable of covering the entire surface of the PCB. The light source uses a ring-shaped LED light source to provide uniform illumination to reduce the impact of shadows and reflections on image quality.

[0056] The image fusion processing module 140 is responsive to the result of comparing the solder joint image with the reference image information and is used to process the solder joint image according to the comparison result. The image fusion processing module 140 includes a defect recognition unit 1401, a distinction unit 1402, and a judgment unit 1403.

[0057] The defect recognition unit 1401 is used to identify defects in the solder joint image based on the comparison result, and the defects include but are not limited to shape defects, position defects, texture defects, color defects and connectivity defects;

[0058] The distinguishing unit 1402 is used to distinguish shape features, position features, texture features, color features and connectivity features respectively;

[0059] The judgment unit 1403 is configured to judge the solder joint images collected from the PCB board in the future in response to the distinction of the shape feature, position feature, texture feature, color feature, and connectivity feature by the distinction unit. If the collected solder joint image is different from any one of the shape feature, position feature, texture feature, color feature, and connectivity feature, the system determines that the PCB board corresponding to the solder joint image is an abnormal PCB board and issues an alarm; otherwise, no judgment is made.

[0060] By acquiring multiple image information of solder joints (shape, position, texture, color, and connectivity features), the system can comprehensively evaluate the quality of solder joints and avoid misjudgment caused by single feature detection.

[0061] In the image information acquisition module, the shape features include solder joint outline, solder joint size, and solder joint shape symmetry;

[0062] Position features include the coordinates of the solder joint and the relative position of the solder joint and surrounding components;

[0063] Texture characteristics include the surface texture of the solder joint and the uniformity of the texture;

[0064] Color features include solder joint color and color variation;

[0065] Connectivity characteristics include the connection between the solder joint and the circuit; the connection condition includes that the connection between the solder joint and the circuit should be good, without open circuit or short circuit;

[0066] Among them, the solder joint size includes the diameter, height and width of the solder joint;

[0067] In this embodiment, a multi-dimensional quality assessment can be performed. Through comprehensive analysis of multiple features, the quality of the solder joint can be more comprehensively assessed, avoiding misjudgment caused by abnormality of a single feature.

[0068] In the differentiation unit, shape features are differentiated, including being classified into circular, elliptical, and irregular shapes based on the solder joint contours. If the solder joint corresponding to the captured solder joint image is irregular in shape, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board.

[0069] Distinguishing positional features, including distinguishing positional features based on the relative positions of solder joints and surrounding components, including distinguishing based on the relative positional relationships between solder joints and adjacent components; if the relative positions of the solder joints corresponding to the captured solder joint image and adjacent components deviate, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board;

[0070] Differentiate texture features, including classifying texture features into textured and non-textured based on the surface texture of solder joints. If both textured and non-textured solder joints appear on the same PCB, the system will determine that the PCB is abnormal.

[0071] Differentiate based on color features, including solder joint color, which should be the same color as the soldering material. If solder joints of different colors appear on the same PCB, the system will determine that the PCB is abnormal.

[0072] Differentiating connectivity features, including differentiating connectivity features based on the connection between solder joints and circuits;

[0073] In this embodiment, specific distinction methods are described for shape features, position features, texture features, color features, and connectivity features, enabling the system to more accurately identify abnormal situations;

[0074] Based on the above, the system distinguishes components based on the relative positional relationship between the solder joint and adjacent components. The distinguishing method includes distinguishing the components adjacent to the solder joint into a first component, a second component, and a third component, calculating the distances between the solder joint and the first component, the second component, and the third component, and marking the distances as reference distances. If the distance between the solder joint and any of the distinguished components is greater than and / or less than the reference distance, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board; otherwise, no determination is made.

[0075] In this embodiment, accurate position analysis and multi-component comparison can be performed, wherein, by calculating the distance between the solder joint and the adjacent components and comparing it with the reference distance, the position deviation of the solder joint can be accurately detected;

[0076] By considering the relative position relationship between the solder joint and multiple components, the accuracy and reliability of position detection are improved;

[0077] Furthermore, the connectivity features are differentiated based on the connection between the solder joints and the circuits. The differentiation method includes counting the number of circuits connected to the solder joints and dividing the circuits into θ1, θ2, ..., θ n , where θ n Indicates the nth line that has been distinguished. The connection status between the line and the solder joints in each distinguished line is collected. If any line is disconnected from the solder joint, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board; otherwise, no determination is made;

[0078] In this embodiment, by counting the number of lines connected to the solder joint and detecting the connection status of each line, the disconnection status between the line and the solder joint can be accurately identified;

[0079] Specifically, in this embodiment, if the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board, the system obtains the circuit corresponding to the disconnected solder joint from the distinguished circuits, obtains a length of at least 1 cm of the circuit based on the edge of the solder joint, obtains the width of the circuit from the length, and evenly divides the width into upper layer width, middle layer width, and lower layer width. If any of the width layers is connected to the solder joint, the system cancels the determination that the PCB board corresponding to the solder joint is an abnormal PCB board; otherwise, the original determination is maintained.

[0080] In this embodiment, by obtaining the width of the circuit and dividing it into upper, middle and lower layer widths, the connection between the circuit and the solder joint can be analyzed in more detail;

[0081] Furthermore, this embodiment also functions as a misjudgment correction mechanism. If a certain layer width of the circuit is found to be connected to a solder joint, the system can cancel the abnormal judgment to avoid misjudgment.

[0082] Obtaining the width of the circuit involves improving the resolution, focal length, and other parameters of the industrial camera to obtain image information of this length and width, ensuring that the camera can clearly capture the details of the circuit and solder joints, and ultimately determine whether the circuit and solder joints are disconnected;

[0083] In reality, even on the same PCB board, the width of the traces connected to the solder joints may not be the same, because the trace width is affected by many factors, including design requirements. The traces on the PCB board have different design requirements based on their different functions, such as the current they carry and the signal type.

[0084] For example, power lines usually require wider lines to carry larger currents and reduce line impedance and voltage drop.

[0085] Signal lines: such as digital signal lines or high-frequency signal lines, may require narrower lines to reduce signal interference and improve transmission efficiency;

[0086] In addition, the width of lines in different areas of the PCB board may also vary. For example, the line close to the power input terminal may be wider, while the line close to the signal processing chip may be narrower.

[0087] The manufacturing process is also a factor. During the PCB manufacturing process, the accuracy of the line width is limited by the manufacturing process. Even on the same PCB board, the line width at different locations may be slightly different due to manufacturing errors.

[0088] At the same time, during the PCB etching process, the width of the circuit may be affected by factors such as the uniformity of the etching solution and the etching time, resulting in a deviation between the actual width and the designed value;

[0089] Therefore, this embodiment makes further determination by obtaining whether layers of different widths are connected to solder joints, which has practical significance.

[0090] It should be emphasized in this embodiment that if the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board, the diameter of the solder joint is obtained, the center point of the solder joint is selected from the diameter, and the distance from the center point to the edge of the solder joint is calculated. The calculation method includes:

[0091] Divide the edge of the weld into at least 6 edge points according to the diameter;

[0092] Calculate the distance between the center point and each edge point;

[0093] If the distances to all edge points are the same, the system will cancel the judgment that the PCB board corresponding to the solder point is an abnormal PCB board; otherwise, the original judgment will be maintained;

[0094] In this embodiment, by calculating the diameter of the solder joint and the distance from the center point to the edge, the shape characteristics of the solder joint can be evaluated more accurately, reducing misjudgments caused by abnormal shape characteristics;

[0095] Furthermore, the spacings between the solder joint and the first component, the second component, and the third component are re-acquired based on the diameter, and the reference spacing is re-assigned based on the re-acquired spacings. The re-assigned reference spacings are marked as second reference spacings. If the spacings between the solder joint and the first component, the second component, and the third component acquired in a future time period are equal to the reference spacings and / or the second reference spacings, the system determines that the PCB board corresponding to the solder joint is a normal PCB board; otherwise, no determination is made.

[0096] In this embodiment, the reference spacing is dynamically adjusted according to the actual diameter of the solder joint, thereby improving the adaptability and flexibility of the system. Moreover, by introducing a second reference spacing, the system can more accurately determine whether the relative position of the solder joint and the component is normal.

[0097] Based on the above, the diameter is marked as the reference diameter, and the diameter corresponding to the solder joint is obtained based on the reference spacing. This corresponding diameter is marked as the second reference diameter. If the diameter of the solder joint obtained in the future time period is consistent with the reference diameter and / or the second reference diameter, the system determines that the PCB board corresponding to the solder joint is a normal PCB board; otherwise, no determination is made;

[0098] In this embodiment, a dynamic diameter reference value is introduced. By introducing a second reference diameter, the system can more accurately determine whether the diameter of the weld meets the requirements;

[0099] Multi-reference value comparison is also introduced. By comparing the reference diameter and the second reference diameter, the system can more accurately determine whether the shape characteristics of the weld are normal.

[0100] In summary, this application can comprehensively evaluate the quality of solder joints through multi-feature analysis, dynamic reference value adjustment and misjudgment correction mechanism, avoid misjudgment caused by single feature detection, and ensure the accuracy and reliability of the detection results.

[0101] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. The PCB board automatic detection system based on machine vision is characterized by: include: An image information acquisition module is used to acquire image information of a normal solder joint of a PCB board and mark the image information as reference image information; the reference image information includes shape features, position features, texture features, color features, and connectivity features of the solder joint; An image processing module, responsive to the reference image information, configured to preprocess the reference image information, the preprocessing comprising grayscale conversion, binarization, and filtering and denoising the reference image information; An image acquisition module is used to acquire an image of a solder joint on a PCB to be inspected and compare the image of the solder joint with the reference image information; An image fusion processing module, which responds to the result of comparing the solder joint image with the reference image information and is used to process the solder joint image according to the comparison result; the image fusion processing module includes a defect recognition unit, a distinction unit, and a judgment unit; The defect recognition unit is used to identify defects in the solder joint image based on the comparison result, and the defects include but are not limited to shape defects, position defects, texture defects, color defects and connectivity defects; The distinguishing unit is used to distinguish the shape feature, position feature, texture feature, color feature and connectivity feature respectively; The judgment unit responds to the distinction of the shape feature, position feature, texture feature, color feature and connectivity feature by the distinction unit, and is used to judge the solder joint image collected from the PCB board in a future time period. If the collected solder joint image is different from any one of the shape feature, position feature, texture feature, color feature and connectivity feature, the system determines that the PCB board corresponding to the solder joint image is an abnormal PCB board and issues an alarm; otherwise, no judgment is made.

2. The PCB automatic detection system based on machine vision according to claim 1, characterized in that: In the image information acquisition module, the shape features include solder joint contour, solder joint size, and solder joint shape symmetry; The position features include the coordinates of the soldering point and the relative position of the soldering point and surrounding components; The texture characteristics include the surface texture of the solder joint and the uniformity of the texture; The color characteristics include solder joint color and color variation; The connectivity features include the connection between the solder joints and the circuits; The solder joint size includes the diameter, height and width of the solder joint.

3. The PCB automatic detection system based on machine vision according to claim 2, characterized in that: In the distinguishing unit, the shape features are distinguished, including distinguishing the shape features into circular, elliptical, and irregular shapes according to the solder joint contour; if the solder joint corresponding to the collected solder joint image is an irregular shape, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board; Distinguishing the position features, including distinguishing the position features based on the relative positions of the solder joints and surrounding components, wherein the distinguishing method includes distinguishing based on the relative position relationship between the solder joints and adjacent components; If the relative position of the solder joint corresponding to the captured solder joint image and the adjacent component deviates, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board; Differentiating the texture features, including differentiating the texture features into textured and non-textured based on the surface texture of the solder joints, wherein if both textured and non-textured solder joints appear on the same PCB board, the system determines that the PCB board is an abnormal PCB board; Differentiating the color features, including differentiating the color features based on the color of the solder joints, wherein the color of the solder joints is the same color as the soldering material; if solder joints of different colors appear on the same PCB board, the system determines that the PCB board is an abnormal PCB board; The connectivity features are distinguished, including distinguishing the connectivity features based on the connection status of the solder joints and the circuits.

4. The PCB automatic detection system based on machine vision according to claim 3, characterized in that: Distinguishing the solder joint from adjacent components based on their relative positional relationship, wherein the distinguishing method includes distinguishing the components adjacent to the solder joint into a first component, a second component, and a third component, calculating the distances between the solder joint and the first component, the second component, and the third component, and marking the distances as reference distances; If, among the distinguished components, the distance between the solder joint and any one of the components is greater than and / or less than the reference distance, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board; Otherwise, no judgment is made.

5. The PCB automatic detection system based on machine vision according to claim 3, characterized in that: The connectivity features are distinguished based on the connection between the solder joint and the circuit. The distinguishing method includes counting the number of circuits connected to the solder joint and dividing the circuits into θ1, θ2, ..., θ n , where θ n Indicates the nth line that has been distinguished. The connection status between the line and the solder joint is collected in each distinguished line. If any line is disconnected from the solder joint, the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board; otherwise, no determination is made.

6. The PCB automatic detection system based on machine vision according to claim 5, characterized in that: If the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board, it obtains the circuit corresponding to the disconnection of the solder joint from the distinguished circuits, and obtains a length of at least 1 cm of the circuit based on the edge of the solder joint. The width of the circuit is obtained from the length and the width is evenly divided into upper layer width, middle layer width and lower layer width. If any one of the width layers is connected to the solder joint, the system cancels the determination that the PCB board corresponding to the solder joint is an abnormal PCB board; otherwise, the original determination is maintained.

7. The PCB automatic detection system based on machine vision according to claim 4, characterized in that: If the system determines that the PCB board corresponding to the solder joint is an abnormal PCB board, the diameter of the solder joint is obtained, a center point about the solder joint is selected from the diameter, and a distance from the center point to the edge of the solder joint is calculated. The calculation method includes: Dividing the edge of the weld into at least 6 edge points according to the diameter; Calculating the distance between the center point and each edge point; If the distances to each edge point are the same, the system cancels the determination that the PCB board corresponding to the solder joint is an abnormal PCB board; otherwise, the original determination is maintained.

8. The machine vision-based PCB automatic detection system according to claim 7, wherein: reacquire the spacing between the solder joint and the first component, the second component, and the third component based on the diameter, reassign the reference spacing based on the reacquired spacing, and mark the reassigned reference spacing as a second reference spacing; if the spacing between the solder joint and the first component, the second component, and the third component acquired in a future time period is equal to the reference spacing and / or the second reference spacing, the system determines that the PCB board corresponding to the solder joint is a normal PCB board; Otherwise, no judgment is made.

9. The PCB automatic detection system based on machine vision according to claim 8, characterized in that: The diameter is marked as a reference diameter, and the diameter corresponding to the solder joint is obtained based on the reference spacing, and this corresponding diameter is marked as a second reference diameter. If the diameter of the solder joint obtained in a future time period is consistent with the reference diameter and / or the second reference diameter, the system determines that the PCB board corresponding to the solder joint is a normal PCB board; otherwise, no determination is made.

Images