Soldering defect intelligent detection method of AOI device and AOI device

Abstract

This invention relates to the field of welding defect detection technology, and discloses an intelligent welding defect detection method and AOI equipment. The method includes: acquiring a field of view (FOV) image of the pad to be inspected; extracting a first red ratio at the center of the pad, a second red ratio at the outer end of the pad, a target row number on the side of the pad, and a third red ratio of the pins based on the FOV image; generating a first comparison result based on the first red ratio, a second comparison result based on the second red ratio, a third comparison result based on the target row number, and a fourth comparison result based on the third red ratio; marking the type of open solder joint defect and overlaying a colored marker box on the FOV image; and outputting an open solder joint defect detection report. This invention establishes detection areas at the center of the pad, the outer end of the pad, the side of the pad, and the pin portion, which can accurately mark the specific type and spatial location of open solder joint defects, thus improving the intelligence level of the AOI inspection system.

Description

Technical Field

[0001] This invention relates to the field of welding defect detection technology, and in particular to an intelligent welding defect detection method and AOI equipment. Background Technology

[0002] In the surface mount technology (SMT) manufacturing process, soldering quality directly affects the reliability and lifespan of electronic products. Traditional AOI (Automated Optical Inspection) methods use threshold judgment for a single inspection area to uniformly identify defects across the entire pad. However, due to significant differences in soldering conditions across different parts of the pad—with varying solder distribution characteristics at the center, outer edges, sides, and leads—a single inspection standard struggles to accurately distinguish between normal soldering and open solder joint defects. This leads to missed detections or misjudgments and an inability to flexibly adapt to the soldering characteristics of different component types, resulting in low inspection accuracy. Summary of the Invention

[0003] This invention provides an intelligent detection method and AOI equipment for welding defects. The invention establishes detection areas for the center of the solder pad, the outer end of the solder pad, the side of the solder pad, and the pin portion, which can accurately mark the specific type and spatial location of the solder joint defect, thereby improving the intelligence level of the AOI detection system.

[0004] In a first aspect, the present invention provides an intelligent detection method for welding defects in AOI equipment, the intelligent detection method for welding defects in AOI equipment comprising: Obtain the FOV image of the pad to be inspected, and extract the first red ratio in the center of the pad, the second red ratio at the outer end of the pad, the target row number on the side of the pad, and the third red ratio of the pins based on the FOV image. A first comparison result is generated based on the first red ratio, a second comparison result is generated based on the second red ratio, a third comparison result is generated based on the target number of rows, and a fourth comparison result is generated based on the third red ratio. Based on the first comparison result, the second comparison result, the third comparison result, and the fourth comparison result, the type of open weld defect is marked and a colored marker box is superimposed on the FOV image to output an open weld defect detection report.

[0005] In conjunction with the first aspect, in a first implementation of the first aspect of the present invention, the step of acquiring the FOV image of the pad to be detected, and extracting the first red ratio at the center of the pad, the second red ratio at the outer end of the pad, the target row number on the side of the pad, and the third red ratio of the pins based on the FOV image, includes: Acquire the field of view (FOV) image of the pad to be inspected, and identify the pin window and pad window in the FOV image; Measure the first dimension of the pin window and the second dimension of the pad window; Based on the second dimension and the pad window, a first detection area in the center of the pad and a second detection area at the outer end of the pad are determined. The third detection area on the side of the pad and the fourth detection area on the pin are determined based on the first size and the pin window. Calculate the first red ratio of the first detection area, the second red ratio of the second detection area, the target row number of the third detection area, and the third red ratio of the fourth detection area.

[0006] In conjunction with the first aspect, in a second implementation of the first aspect of the present invention, the calculation of the first red ratio of the first detection region, the second red ratio of the second detection region, the target row number of the third detection region, and the third red ratio of the fourth detection region includes: Extract the number of first pixels in the red channel in the first detection area, and calculate the first proportion of the number of first pixels to the first total number of red pixels in the first detection area; Extract the number of second pixels in the red channel of the second detection region, and calculate the second red ratio of the number of second pixels to the second total number of pixels in the second detection region; The left and right regions of the third detection area are scanned line by line to extract the number of target rows that meet the preset first condition in the red-green mixed channel; Extract the number of third pixels in the red channel of the fourth detection area that meet the preset second condition, and calculate the third proportion of the number of third pixels to the total number of third red pixels in the fourth detection area.

[0007] In conjunction with the first aspect, in a third implementation of the first aspect of the present invention, the step of generating a first comparison result based on the first red ratio, generating a second comparison result based on the second red ratio, generating a third comparison result based on the target row number, and generating a fourth comparison result based on the third red ratio includes: The first red ratio is compared with the first threshold. When the first red ratio is less than or equal to the first threshold, the first comparison result is assigned to 0. When the first red ratio is greater than the first threshold, the first comparison result is assigned to 1. The second red ratio is compared with the second threshold. When the second red ratio is greater than or equal to the second threshold, the second comparison result is assigned a value of 0. When the second red ratio is less than the second threshold, the second comparison result is assigned a value of 1. The target number of rows is compared with the row number threshold. When the target number of rows in both the left and right regions of the third detection area is less than or equal to the row number threshold, the third comparison result is assigned a value of 1; otherwise, the third comparison result is assigned a value of 0. The third red ratio is compared with the third threshold. When the third red ratio is greater than or equal to the third threshold, the fourth comparison result is assigned a value of 0. When the third red ratio is less than the third threshold, the fourth comparison result is assigned a value of 1.

[0008] In conjunction with the first aspect, in the fourth implementation of the first aspect of the present invention, the step of comparing the target row count with a row count threshold, assigning a value of 1 when the target row counts in both the left and right regions of the third detection region are less than or equal to the row count threshold, and assigning a value of 0 otherwise, to obtain a third comparison result, includes: Compare the target row count in the left region of the third detection region with the row count threshold. When the target number of rows in the left region is less than or equal to the number of rows threshold, the left judgment value is assigned to 1; otherwise, the left judgment value is assigned to 0. The number of target rows in the right region of the third detection area is compared with the row count threshold. When the number of target rows in the right region is less than or equal to the row count threshold, the right judgment value is assigned to 1; otherwise, the right judgment value is assigned to 0. Perform an AND operation on the left-hand judgment value and the right-hand judgment value. If both the left-hand judgment value and the right-hand judgment value are 1, assign the third comparison result as 1; otherwise, assign the third comparison result as 0.

[0009] In conjunction with the first aspect, in a fifth implementation of the first aspect of the present invention, the step of marking the type of open weld defect based on the first comparison result, the second comparison result, the third comparison result, and the fourth comparison result, superimposing a colored marker box on the FOV image, and outputting an open weld defect detection report includes: The target combination strategy expression is selected based on the type of the component to be detected. The target combination strategy expression includes addition logical operators and multiplication logical operators as well as operation priority indicators. The first comparison result, the second comparison result, the third comparison result, and the fourth comparison result are substituted into the target combination strategy expression according to the operation priority identifier to obtain the defect comprehensive analysis result; Based on the comprehensive defect analysis results, the type of open weld defect is marked and a colored marker box is superimposed on the FOV image to output an open weld defect detection report.

[0010] In conjunction with the first aspect, in a sixth implementation of the first aspect of the present invention, the step of substituting the first comparison result, the second comparison result, the third comparison result, and the fourth comparison result into the target combination strategy expression according to the operation priority identifier to obtain the defect comprehensive analysis result includes: Substitute the first comparison result and the second comparison result into the parentheses of the target combination strategy expression, and perform an addition logical operation. When the first comparison result is 0 or the second comparison result is 0, the first intermediate operation result is 0. When both the first comparison result and the second comparison result are 1, the first intermediate operation result is 1. Perform a multiplication logic operation between the first intermediate operation result and the third comparison result. When the first intermediate operation result is 1 or the third comparison result is 1, the operation outputs a second intermediate operation result of 1. When both the first intermediate operation result and the third comparison result are 0, the operation outputs a second intermediate operation result of 0. When the result of the second intermediate calculation is 1, the result of the comprehensive defect analysis is determined to be that the solder joint has a weld breakage defect. When the result of the second intermediate calculation is 0, the result of the comprehensive defect analysis is determined to be that the solder joint does not have a weld breakage defect.

[0011] In conjunction with the first aspect, in the seventh implementation of the first aspect of the present invention, the step of marking the type of open weld defect based on the comprehensive defect analysis results and overlaying a colored marker box on the FOV image, and outputting an open weld defect detection report, includes: When the comprehensive defect analysis result indicates that there is no open weld defect at the weld point, it is marked as qualified. When the comprehensive defect analysis result indicates that there is a weld breakage defect at the weld point, the defect type subdivision process is initiated and the weld breakage defect type is marked. A colored marker box is then overlaid on the FOV image to obtain a weld breakage defect detection report.

[0012] In conjunction with the first aspect, in the eighth implementation of the first aspect of the present invention, when the comprehensive defect analysis result indicates that the solder joint has a weld breakage defect, the defect type subdivision process is entered and the weld breakage defect type is marked. A color-coded marker box is superimposed on the FOV image to obtain a weld breakage defect detection report, including: When the comprehensive defect analysis result indicates that the weld joint has a weld breakage defect, the defect type subdivision process is entered and the weld breakage defect type is marked. A colored marker box is superimposed on the FOV image. The first red ratio, the second red ratio, the target row number, the third red ratio, and the colored marker box are integrated to obtain a weld defect detection report.

[0013] In a second aspect, the present invention provides an AOI device, the AOI device comprising: The acquisition module is used to acquire the FOV image of the pad to be inspected, and extract the first red ratio in the center of the pad, the second red ratio at the outer end of the pad, the target row number on the side of the pad, and the third red ratio of the pin based on the FOV image. The comparison module is used to generate a first comparison result based on the first red ratio, a second comparison result based on the second red ratio, a third comparison result based on the target number of rows, and a fourth comparison result based on the third red ratio. The output module is used to mark the type of open weld defect based on the first comparison result, the second comparison result, the third comparison result and the fourth comparison result, and to overlay a colored marker box on the FOV image to output an open weld defect detection report.

[0014] The technical solution provided by this invention establishes detection areas for the center, outer end, side, and pin portions of the solder pad. RGB three-channel color features are extracted for the soldering characteristics of different parts of the pad, generating binary judgment results for each. A combined logic strategy expression, including addition and multiplication operators, is used for logical operations, achieving intelligent comprehensive analysis of multiple detection results. The addition logic operator allows any area, whether in the center or outer end of the pad, to pass the initial screening if it is normal, while the multiplication logic operator mandates that critical areas must simultaneously meet the detection criteria. This flexible logical combination mechanism can select the optimal judgment strategy based on different component types and soldering process characteristics, effectively reducing the false positive and false negative rates. This invention can accurately label the specific type and spatial location of solder joint defects, improving the intelligence level of the AOI inspection system. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A schematic flowchart illustrating the intelligent detection method for welding defects in AOI equipment provided in this application embodiment; Figure 2 This is a schematic diagram of the PCB board Mark point location provided in an embodiment of this application; Figure 3 This is a schematic diagram illustrating the relationship between the component window and the automatic retrieval window provided in an embodiment of this application; Figure 4 A schematic diagram of the detection window layout for a CHIP-type element provided in an embodiment of this application; Figure 5 This is a schematic diagram illustrating the vertical and horizontal direction definitions of CHIP-type elements provided in an embodiment of this application; Figure 6 This is a schematic diagram illustrating the number of consecutive rows of pads on the side, provided in an embodiment of this application. Figure 7A comparison table of actual image samples of various welding defects provided for embodiments of this application; Figure 8 A schematic block diagram of the structure of the AOI device provided in the embodiments of this application. Detailed Implementation

[0017] 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, not all, of the embodiments of the present invention. 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.

[0018] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the described order. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change based on the actual situation.

[0019] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0020] It should also be further understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0021] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features described herein can be combined with each other.

[0022] Please see Figure 1 , Figure 1 This is a flowchart illustrating the intelligent detection method for welding defects in AOI equipment provided in the embodiments of this application, as shown below. Figure 1 As shown in the embodiments of this application, the intelligent detection method for welding defects in AOI equipment includes the following steps: Step S100: Obtain the FOV image of the pad to be inspected, and extract the first red ratio in the center of the pad, the second red ratio at the outer end of the pad, the target row number on the side of the pad, and the third red ratio of the pin based on the FOV image. Step S200: Generate a first comparison result based on the first red ratio, generate a second comparison result based on the second red ratio, generate a third comparison result based on the target number of rows, and generate a fourth comparison result based on the third red ratio; Step S300: Based on the first comparison result, the second comparison result, the third comparison result, and the fourth comparison result, mark the type of open weld defect and overlay a colored marker box on the FOV image, and output the open weld defect detection report.

[0023] In one specific embodiment, the process of performing step S100 may specifically include the following steps: Acquire the FOV image of the pad to be inspected, and identify the pin window and pad window in the FOV image; Measure the first dimension of the pin window and the second dimension of the pad window; The first detection area in the center of the pad and the second detection area at the outer end of the pad are determined based on the second size and the pad window. The third detection area on the side of the pad and the fourth detection area on the pin are determined based on the first size and the pin window. Calculate the first red ratio of the first detection area, the second red ratio of the second detection area, the target row number of the third detection area, and the third red ratio of the fourth detection area.

[0024] Specifically, the FOV image containing the target pad is acquired through the high-resolution imaging system built into the AOI device. The FOV image covers the entire pin and pad structure of the component and has high color fidelity. Figure 2 This is a schematic diagram of the Mark point locations on the PCB board, including the distribution of two Mark points on the PCB board. Mark point 1 is located in the lower left corner of the PCB board, and Mark point 2 is located in the upper right corner of the PCB board. The two Mark points are used to perform global correction of the XY coordinates and θ angle of the PCB board, ensuring that the fourth-level detection area can accurately locate the actual position of the pads and compensate for the positional offset of the PCB board caused by transportation or deformation. Figure 3This diagram illustrates the relationship between the component window and the automatic retrieval window, including their spatial position. The automatic retrieval window, acting as the outer window, defines the retrieval range for components, while the component window, located inside, defines the actual dimensions of the component body. Together, they enable precise retrieval and positioning of components. A structure recognition module based on edge detection and template matching algorithms identifies pin windows and pad windows in the FOV image and extracts their boundary coordinate information to establish a unified spatial coordinate reference system. Based on this, the vertical length and horizontal width of the pin window are measured as the first dimension, and the vertical length and horizontal width of the pad window are measured as the second dimension. During the detection area division process, based on the second size of the pad window, by setting the boundary removal ratio parameter, the rectangular area remaining after removing the edges in the pad window is defined as the first detection area in the center of the pad. A fixed ratio is set to extend inward in the vertical direction to form a second detection area near the pad boundary area for analyzing the solder coverage at the outer end. At the same time, referring to the first size of the pin window, the vertical extension range and the horizontal width range are set on the left and right sides of the pin window respectively to form a third detection area on the side of the pad, which is used to identify whether the solder can well cover both sides of the pad. Meanwhile, a fixed ratio range is defined inside the pin window in the vertical direction as a fourth detection area to capture the soldering feature information on the pin body. After the regions are divided, pixel-level color feature extraction is performed on the four detection regions. In the first detection region, the red channel brightness value is extracted, and the number of red pixels that meet the preset brightness threshold range is divided by the total number of pixels in the region to obtain the first red ratio, which reflects the solder coverage in the central area of ​​the pad. In the second detection region, similar processing is performed using different red threshold parameters to obtain the second red ratio, which indicates whether the solder creep at the outer end of the pad meets the process requirements. In the third detection region, each pixel row is scanned line by line, and the ratio of pixels in the red and green channels that simultaneously meet the brightness conditions is counted. If the pixel ratio exceeds the set threshold in the row, the row is counted as the effective target row. The target rows on the left and right sides are counted separately as the target row number in the third region. In the fourth detection region, the red channel information is extracted after eliminating blue interference, and the ratio of red pixels that meet the threshold requirements is counted as the third red ratio, which is used to determine the soldering validity of the pin area.

[0025] Figure 4 This diagram illustrates the detection window layout for chip-type components, including a complete detection window system for standard chip-type components. It consists of an outermost automatic retrieval window, two pad windows (pad window 1 and pad window 2), a central component window, a polarity window, four short-circuit windows (short-circuit window 1 to short-circuit window 4), and three missing component windows (missing component window 1 to missing component window 3). All detection windows work together to perform comprehensive detection of the center of the pad, the outer edge of the pad, the side of the pad, and the pin portion.

[0026] Figure 5 This diagram illustrates the definition of the vertical and horizontal directions for CHIP-type components, including the definition methods for the vertical and horizontal directions in CHIP-type components. The vertical direction is defined as the direction of the connection line between two pads, and the horizontal direction is defined as the direction perpendicular to the connection line between two pads. This direction definition is used for the directional determination of the spatial coordinate system establishment and color feature extraction of the subsequent four-level detection area.

[0027] Figure 6 This is a schematic diagram illustrating the number of consecutive rows matching the color characteristics on the side of the pad. It includes the row-by-row scanning process of the detection area on the side of the pad. Dark squares indicate that there are pixels matching the color characteristics (with solder) in that row, and light squares indicate that there are no pixels matching the color characteristics in that row (without solder). The number of consecutive rows of dark squares is counted from the outer edge of the pin window outwards. Figure 6 The corresponding bar chart shows the number of consecutive matching rows. When the number of consecutive matching rows is less than the threshold, it is determined to be a solder pad side open solder defect.

[0028] Before determining the first detection area in the center of the pad and the second detection area at the outer end of the pad based on the second size and the pad window, the process includes: calculating the aspect ratio for the horizontal and vertical lengths of the pad window; determining the pad type as wide when the aspect ratio is greater than a preset threshold, and as a regular pad when the aspect ratio is less than or equal to the preset threshold; selecting a removal ratio parameter based on the pad type determination result; selecting a first removal ratio parameter when the pad type determination result is wide, and selecting a second removal ratio parameter when the pad type determination result is regular, wherein the first removal ratio parameter is less than the second removal ratio parameter; obtaining the horizontal removal length by multiplying the horizontal length of the pad window by the selected removal ratio parameter, and obtaining the vertical removal length by multiplying the vertical length of the pad window by the selected removal ratio parameter; removing the horizontal removal length inward from both horizontal ends of the pad window, and removing the vertical removal length inward from both vertical ends of the pad window, and determining the central rectangular area after removal as the first detection area in the center of the pad.

[0029] The method for determining the third detection area on the side of the pad based on the first size and the pin window includes: measuring the lateral distance between the side end of the pin window and the corresponding side end of the pad window to obtain the available width of the side; obtaining the longitudinal length of the side detection based on the longitudinal length of the pin window multiplied by a preset side length ratio parameter; determining whether the available width of the side is greater than a preset minimum width threshold; when the available width of the side is greater than the minimum width threshold, removing a preset exclusion width pixel from the side end of the pad window inward, and using the lateral width of the removed area as the lateral width of the side detection; when the available width of the side is less than or equal to the minimum width threshold, directly using the available width of the side as the lateral width of the side detection; starting from the outer end of the pin window, extending the longitudinal length of the side detection along the longitudinal direction and extending the lateral width of the side detection along the lateral direction, generating a left detection area and a right detection area on the left and right sides of the pin window respectively, and combining the left detection area and the right detection area to obtain the third detection area on the side of the pad.

[0030] In one specific embodiment, the process of calculating the first red ratio of the first detection area, the second red ratio of the second detection area, the target row number of the third detection area, and the third red ratio of the fourth detection area can specifically include the following steps: Extract the number of first pixels in the red channel in the first detection area, and calculate the first proportion of the number of first pixels to the first total number of red pixels in the first detection area; Extract the number of second pixels in the red channel of the second detection region, and calculate the proportion of the number of second pixels to the total number of second red pixels in the second detection region; The left and right regions of the third detection area are scanned line by line to extract the number of target rows that meet the preset first condition in the red-green mixed channel. Extract the number of third pixels in the red channel of the fourth detection area that meet the preset second condition, and calculate the proportion of the number of third pixels to the total number of third red pixels in the fourth detection area.

[0031] Specifically, in the first detection area, each pixel is traversed, its red channel brightness value is extracted, and a threshold range of 120 to 180 is set. Simultaneously, green channel brightness less than 100 and blue channel brightness less than 100 are used as joint filtering conditions. The number of pixels satisfying these combined conditions is defined as the first pixel count. The total number of pixels in the first detection area is also counted. Assuming the first detection area size is 80×60 pixels, the first total pixel count is 4800. The first red ratio is obtained by dividing the first pixel count by 4800 and multiplying by 100%. If the first pixel count is 720, the first red ratio is 15%. In the second detection area, the red channel brightness value is judged for each pixel, with a threshold range of 100 to 160. Green and blue channel brightness values ​​less than 120 are used as filtering conditions. The number of second pixels satisfying these conditions is counted. If the second detection area size is 100×40 pixels, the total pixel count is 4000. If the second pixel count is 3200, the second red ratio is 80%. In the analysis of the third detection area, the left and right areas are scanned line by line. The red and green channel brightness values ​​of all pixels in each line are extracted. The threshold for red is set to 80 to 140, and the threshold for green is set to 60 to 100. It is determined whether the number of pixels in the row that meet the red and green channel brightness conditions accounts for more than 50% of the total number of pixels in the row (as the lower limit of the target row determination ratio). If it meets the condition, the row is recorded as the target row. Assuming that there are 60 rows in the left area, and 12 of them meet the above conditions, the number of consecutive valid rows on the left is 12. If there are 10 rows in the right area that meet the conditions, the number of consecutive valid rows on the right is 10. In the fourth detection area, the red channel value is extracted, and the brightness threshold is set to 100 to 150. The blue channel brightness is required to be less than 80. The number of pixels that meet this condition is counted as the third pixel count. If the size of the fourth detection area is 50×40 pixels, the total number of pixels is 2000, and the third pixel count is 1500, then the third red ratio is 75%.

[0032] The process of extracting the number of first pixels in the red channel within the first detection region and calculating the first red proportion of the first pixel count relative to the total number of pixels in the first detection region includes: traversing each pixel in the first detection region and obtaining the red channel luminance value R, green channel luminance value G, and blue channel luminance value B for each pixel; comparing the red channel luminance value R of each pixel with the lower limit threshold and the upper limit threshold of the red channel luminance; marking the pixel as a candidate pixel for the red channel when the red channel luminance value R is greater than or equal to the lower limit threshold and less than or equal to the upper limit threshold; further determining whether the green channel luminance value G is less than the green channel limit threshold and whether the blue channel luminance value B is less than the blue channel limit threshold for each pixel; confirming the pixel as a red feature pixel and including it in the first pixel count when both the green channel luminance value G and the blue channel luminance value B are less than the blue channel limit threshold; and calculating the first red proportion by dividing the first pixel count by the total number of pixels in the first detection region.

[0033] The process involves scanning the left and right regions of the third detection area line by line to extract the target number of rows in the red-green mixed channel that meet a preset first condition. This includes: scanning the left region of the third detection area line by line from the outer edge of the pin window, extracting the red channel brightness value and green channel brightness value of all pixels in each row; determining the number of pixels in each row that simultaneously meet the condition that the red channel brightness value is within the first red brightness range and the green channel brightness value is within the first green brightness range, and calculating the proportion of this number of pixels to the total number of pixels in the row to obtain the row color ratio; comparing the row color ratio with a preset row color ratio threshold, and marking the row as a compliant row when the row color ratio is greater than the threshold, and counting the number of consecutive rows marked as compliant rows starting from the outer edge of the pin window to obtain the number of consecutive compliant rows in the left region; applying the same line-by-line scanning and counting method to the right region of the third detection area to obtain the number of consecutive compliant rows in the right region; and taking the smaller value between the number of consecutive compliant rows in the left region and the number of consecutive compliant rows in the right region as the target number of the third detection area.

[0034] The process involves extracting the number of third pixels in the fourth detection region whose red channel meets the preset second condition, and calculating the proportion of the number of third pixels to the total number of third red pixels in the fourth detection region. This includes: traversing each pixel in the fourth detection region and extracting the red channel brightness value R, green channel brightness value G, and blue channel brightness value B for each pixel; calculating the red-blue brightness difference for each pixel, where the red-blue brightness difference equals the red channel brightness value R minus the blue channel brightness value B; comparing the red channel brightness value R with the lower threshold and upper threshold of the pin red channel, and determining the third red channel brightness value when R falls below the lower threshold and upper threshold of the pin red channel. When the value of a pixel falls within the upper threshold of the red channel of the pin, it is marked as a candidate pixel for red pin. For a pixel marked as a candidate pixel for red pin, it is determined whether its blue channel brightness value B is less than the blue channel exclusion threshold and whether the difference between red and blue brightness is greater than the lower threshold of the difference. When both the blue channel brightness value B and the difference between red and blue brightness are satisfied, the pixel is confirmed as a feature pixel of red pin and is included in the number of third pixels. The total number of pixels confirmed as feature pixels of red pin in the fourth detection area is counted to obtain the number of third pixels. The number of third pixels is calculated and divided by the total number of third pixels in the fourth detection area to obtain the proportion of third red.

[0035] In one specific embodiment, the process of performing step S200 may specifically include the following steps: The first red ratio is compared with the first threshold. When the first red ratio is less than or equal to the first threshold, the first comparison result is assigned to 0. When the first red ratio is greater than the first threshold, the first comparison result is assigned to 1. The second red ratio is compared with the second threshold. When the second red ratio is greater than or equal to the second threshold, the second comparison result is assigned to 0. When the second red ratio is less than the second threshold, the second comparison result is assigned to 1. The target number of rows is compared with the row number threshold. When the target number of rows in both the left and right regions of the third detection area is less than or equal to the row number threshold, the third comparison result is assigned a value of 1; otherwise, the third comparison result is assigned a value of 0. The third red ratio is compared with the third threshold. When the third red ratio is greater than or equal to the third threshold, the fourth comparison result is assigned a value of 0. When the third red ratio is less than the third threshold, the fourth comparison result is assigned a value of 1.

[0036] Specifically, the first red ratio calculated for the central area of ​​the pad is read, for example, if its value is 15%, and compared with a preset first threshold. If the first threshold is set to 20%, then since 15% ≤ 20%, the red exposure of the central area of ​​the pad is determined to be lower than the threshold, indicating that the solder coverage in the central area is normal, and the first comparison result is assigned to 0. If the first red ratio is 23%, then since 23% > 20%, it is inferred that the exposed copper area in the center of the pad is too large, and the first comparison result is assigned to 1. The second red ratio of the outer area of ​​the pad is processed, for example, if its value is 72%, if the preset second threshold is 70%, then 72% ≥ 70% indicates that the red pixel distribution has reached a sufficient solder climb level, and there is no obvious solder shortage or poor soldering at the outer end, so the second comparison result is assigned to 0. If the measured red ratio is 65%, which is lower than the 70% threshold, it is inferred that there is a risk of insufficient solder in the area, and the second comparison result is assigned to 1. The third detection area is processed by counting the target rows. The number of consecutive target rows satisfying the red-green condition within a row is recorded in both the left and right regions. For example, if the left side has 8 rows and the right side has 6 rows, and the preset row threshold is 10 rows, then since 8 ≤ 10 and 6 ≤ 10, the "neither exceeding" condition is met, and the third comparison result is assigned as 1, indicating anomalies such as discontinuous coverage or insufficient solder creep on the side of the pad. If the left side has 12 rows and the right side has 9 rows, since the target row count on the left side exceeds the threshold, the "simultaneously less than or equal to" condition is not met, and the soldering on the side of the pad is judged to be relatively normal, and the third comparison result is assigned as 0. The third red ratio extracted from the pin area is numerically compared. For example, if the third red ratio is 60%, it is compared with the set third threshold of 55%. Since 60% ≥ 55%, it indicates sufficient solder coverage on the pin, and there are no obvious thin solder or open solder joint issues at the soldering interface, so the fourth comparison result is assigned as 0. If the ratio is only 50%, it is below the lower threshold, and an soldering anomaly is judged at the pin, so the fourth comparison result is assigned as 1.

[0037] In one specific embodiment, the execution step compares the target row count with a row count threshold. When the target row counts in both the left and right regions of the third detection region are less than or equal to the row count threshold, a value of 1 is assigned; otherwise, a value of 0 is assigned. The process of obtaining the third comparison result can specifically include the following steps: Compare the number of target rows in the left region of the third detection area with the row count threshold; When the target number of rows in the left region is less than or equal to the row number threshold, the left judgment value is assigned to 1; otherwise, the left judgment value is assigned to 0. Compare the number of target rows in the right region of the third detection area with the row count threshold. If the number of target rows in the right region is less than or equal to the row count threshold, assign the right judgment value to 1; otherwise, assign the right judgment value to 0. Perform a logical AND operation on the left and right comparison values. If both the left and right comparison values ​​are 1, assign the third comparison result to 1; otherwise, assign the third comparison result to 0.

[0038] Specifically, for the pixel rows on the left side of the third detection area, a row-by-row scan is performed sequentially. Combining the conditions that the red channel brightness value is between 80 and 140 and the green channel brightness value is between 60 and 100, it is determined whether there are pixels in each row whose number satisfies the red-green mixed brightness characteristic, accounting for more than 50% of the total number of pixels in that row. When the condition is met, a valid target row is counted. After all scans are completed, the number of target rows in the left area is obtained, for example, 8 rows. A row count threshold of 10 rows is set. The left target row count of 8 is compared with 10. Since 8 is less than or equal to 10, the left judgment value is assigned 1, indicating a potential welding abnormality on that side. If the left target row count is 12 rows, exceeding the threshold, the left judgment value is assigned 0, indicating that the welding status on that side is normal. The same judgment process is repeated for the right area. Assuming the right target row count is 6 rows, it is compared with the row count threshold of 10. Since 6 is less than or equal to 10, the right judgment value is assigned 1; if it is 11 rows, the right judgment value is assigned 0. The left and right judgment values ​​are ANDed. Only when both the left and right judgment values ​​are 1, it is confirmed that there is insufficient continuous welding feature on both sides of the pad. The third comparison result is assigned as 1, indicating that the entire third detection area has welding abnormality. Otherwise, as long as the judgment value on either side is 0, it means that at least one side has effective welding. The third comparison result is assigned as 0, indicating that the area is not judged as an abnormal solder joint.

[0039] In one specific embodiment, the process of performing step S300 may specifically include the following steps: Select the target combination strategy expression based on the type of the component to be detected. The target combination strategy expression includes addition logic operators and multiplication logic operators as well as operation priority indicators. Substitute the first comparison result, the second comparison result, the third comparison result, and the fourth comparison result into the target combination strategy expression according to the operation priority identifier to obtain the defect comprehensive analysis result; Based on the comprehensive defect analysis results, the type of weld breakage defect is marked and a colored marker box is superimposed on the FOV image to output a weld breakage defect detection report.

[0040] Specifically, after performing color feature extraction and threshold comparison in the four-level detection areas, the first, second, third, and fourth comparison results are obtained respectively. These results represent the presence of open solder joints in the four areas in numerical form. A value of 0 indicates that the area detection is normal, and a value of 1 indicates that the area has suspected open solder joints. Based on the package structure type, pin distribution characteristics, and process requirements of the component under test, the most suitable combination strategy expression is selected from the preset strategy expression library as the current detection logic formula. The combination strategy expression consists of "addition logic operators" and "multiplication logic operators". The addition logic means that the result is 1 when all participating terms are 1, and 0 if any one is 0. The multiplication logic means that the result is 0 when all participating terms are 0, and 1 if any one is 1. Priority is indicated by parentheses. For example, for surface mount resistors, the combined expression (first comparison result + second comparison result) × third comparison result is selected, meaning that an overall open solder joint is considered only if there is at least one abnormality in the center and outer edge of the pad, and an abnormality also exists in the side area. For pin-type components, (first comparison result + second comparison result) × (third comparison result + fourth comparison result) is selected, requiring at least one abnormality in the pad area and at least one abnormality in the pin-related area for a failure. For high-reliability devices, the most stringent first comparison result + second comparison result + third comparison result + fourth comparison result is used, meaning that an abnormality in any detection area is considered an overall failure. Substituting the four comparison results into the selected strategy expression and performing logical operations according to the calculation order specified in parentheses yields the comprehensive defect analysis result. If the result is 1, it indicates that the solder joint has an open solder joint defect; if it is 0, it is considered defect-free. When the overall defect analysis result is 1, the individual comparison results are analyzed back to determine the defect type by identifying which regions have a comparison result of 1. For example, if only the first comparison result is 1 and the rest are 0, it is marked as "center solder joint open". If the third comparison result is 1 and the rest are 0, it is marked as "side solder joint open". If multiple regions have a result of 1, it is marked as "compound open". At the same time, color-coded boxes are superimposed on the original FOV image to visually display the defect area. Red rectangles are used to mark the center of the solder pad, yellow boxes to mark the outer edge of the solder pad, green boxes to mark the side of the solder pad, and blue boxes to mark the pin area. The image, along with the judgment result, defect type, and four characteristic parameters, are used to generate a combined text and image open solder joint defect detection report.

[0041] The selection of a target combination strategy expression based on the type of the component to be detected includes addition and multiplication logic operators, as well as operation priority indicators. This includes: establishing a preset combination strategy expression library containing a first strategy expression, a second strategy expression, and a third strategy expression; the first strategy expression is defined as performing an addition operation between the first and second comparison results, followed by a multiplication operation between the result and the third comparison result, applicable to conventional CHIP-type components; the second strategy expression is defined as performing an addition operation between the first and second comparison results, followed by an addition operation between the result and the third and fourth comparison results. The first strategy expression performs a multiplication operation, suitable for pin-mounted chip components; the second strategy expression is defined as performing an addition operation on the first comparison result, the second comparison result, the third comparison result, and the fourth comparison result in sequence, suitable for components with high reliability requirements; the component type identifier of the component to be tested is obtained, and the corresponding strategy expression is matched from the preset combination strategy expression library according to the component type identifier. When the component type identifier is a CHIP type component, the first strategy expression is selected as the target combination strategy expression; when the component type identifier is a pin-mounted chip component, the second strategy expression is selected as the target combination strategy expression; when the component type identifier is a component with high reliability requirements, the third strategy expression is selected as the target combination strategy expression.

[0042] In one specific embodiment, the process of substituting the first comparison result, the second comparison result, the third comparison result, and the fourth comparison result into the target combination strategy expression according to the operation priority identifier to obtain the defect comprehensive analysis result can specifically include the following steps: Substitute the first comparison result and the second comparison result into the parentheses of the target combination strategy expression, perform addition logic operation, and output the first intermediate operation result as 0 when the first comparison result is 0 or the second comparison result is 0, and output the first intermediate operation result as 1 when both the first comparison result and the second comparison result are 1. Perform a multiplication logic operation between the first intermediate operation result and the third comparison result. When the first intermediate operation result is 1 or the third comparison result is 1, the operation outputs the second intermediate operation result as 1. When both the first intermediate operation result and the third comparison result are 0, the operation outputs the second intermediate operation result as 0. When the result of the second intermediate calculation is 1, the result of the comprehensive defect analysis is determined to be that the weld joint has an open weld defect. When the result of the second intermediate calculation is 0, the result of the comprehensive defect analysis is determined to be that the weld joint does not have an open weld defect.

[0043] Specifically, the first comparison result and the second comparison result are substituted into the parentheses of the combination strategy expression as input values ​​and processed using the addition logic operation rule. That is, when either of the two input values ​​is 0, the first intermediate operation result is 0. The first intermediate operation result is 1 only when both the first comparison result and the second comparison result are 1. This reflects the constraint that "all anomalies must be present to constitute a comprehensive anomaly in the pad area". For example, when the first comparison result is 1 and the second comparison result is 0, the first intermediate operation result is 0; and when both are 1, the output value is 1. The first intermediate operation result and the third comparison result are used as the two input values ​​for the multiplication logic operation, and the logic judgment is performed again. The logic rule requires that if either of the two input values ​​is 1, the multiplication logic output is 1; otherwise, it only outputs 0 when both input values ​​are 0. This expresses the idea that "if an anomaly is confirmed in the solder pad area, or if there is a soldering defect on the side of the solder pad, the solder joint is considered to have a risk of open solder joint." For example, if the first intermediate operation result is 0 but the third comparison result is 1, the multiplication operation result, i.e., the second intermediate operation result, is 1; if both are 0, the final output of the multiplication logic is 0. Based on the value of the second intermediate operation result, the defect comprehensive analysis conclusion is determined. When the second intermediate operation result is 1, it clearly indicates that the current solder joint has an identifiable open solder joint defect, and the judgment result is a non-conforming product; when the second intermediate operation result is 0, it is determined that all critical areas have not formed a joint abnormal state, confirming that the solder joint has not shown an identifiable structural open solder joint, and the output result is qualified.

[0044] In one specific embodiment, the process of marking the type of open weld defect based on the comprehensive defect analysis results, overlaying a colored marker box on the FOV image, and outputting an open weld defect detection report can specifically include the following steps: When the comprehensive defect analysis result shows that there is no open weld defect at the weld point, it is marked as qualified. When the comprehensive defect analysis results indicate that there is a weld breakage defect at the weld point, the defect type subdivision process is initiated and the weld breakage defect type is marked. A colored marker box is then overlaid on the FOV image to obtain a weld breakage defect detection report.

[0045] Specifically, when the comprehensive analysis result is 0, it indicates that the logical combination of the first, second, third, and fourth comparison results did not trigger the corresponding defect judgment condition. This means that no obvious abnormalities were found in any of the key inspection areas for the current solder joint, and the current solder joint is marked as "qualified." No defect identification or marking diagram is included in the inspection report. When the comprehensive analysis result is 1, it indicates that at least one or more inspection areas meet the conditions for triggering a solder joint opening defect. This triggers the defect type subdivision process, which determines the specific area type of the defect by tracing back the values ​​of the four individual comparison results. For example, if the first comparison result is 1 and the rest are 0, it is judged as "central solder joint opening"; if the second comparison result is 1 and the rest are 0, it is "outer solder joint opening"; if the third comparison result is 1, it is "side solder joint opening"; if the fourth comparison result is 1, it is "pin area solder joint opening"; and if two or more comparison results are 1, it is marked as "compound solder joint opening" based on the specific combination, and all trigger area names are listed. After determining the defect type, corresponding colored rectangular marker boxes are superimposed on the FOV image based on the spatial coordinate range of each detection area. A red box is superimposed on the center of the pad, a yellow box on the outer edge of the pad, a green box on the side of the pad, and a blue box on the pin. Different colors correspond to different areas. The position of the marker boxes is automatically derived from the original coordinates of the detection area, ensuring accurate image visualization and positioning. The determination result, defect type, values ​​of each color feature parameter, various set thresholds, and the colored marked image are combined to generate a solder joint defect detection report. The report can be automatically output to the interface display module or uploaded to the quality traceability database.

[0046] In one specific embodiment, the process of executing the steps when the comprehensive defect analysis result indicates that there is a weld breakage defect at the solder joint, entering the defect type subdivision process and marking the weld breakage defect type, and overlaying a colored marker box on the FOV image to obtain a weld breakage defect detection report can specifically include the following steps: When the comprehensive defect analysis result indicates that there is a weld breakage defect at the weld point, the defect type subdivision process is initiated and the weld breakage defect type is marked, and a colored marker box is superimposed on the FOV image. By integrating the first red ratio, the second red ratio, the target row number, the third red ratio, and the colored marker boxes, a weld defect detection report is obtained.

[0047] Figure 7This is a reference table of actual images of various welding defects, including real images of various welding defect types commonly seen in AOI inspection, such as good products, missing parts, solder joint offset, defect offset, deviation, tombstoning, open solder joint, missing solder, insufficient solder, excessive solder, red glue overflow, short circuit, porosity, missing pins, and many other defect forms. Specifically, based on the comprehensive defect analysis results, differentiated processing is performed. When the result value is 1, it indicates the presence of a solder joint defect. The defect type subdivision module is then invoked, and the individual comparison results of each detection area are analyzed retrospectively. The value status of the first, second, third, and fourth comparison results is determined item by item to identify which detection area the defect is concentrated in. If only the first comparison result is 1 and the other results are 0, the defect type is labeled as "central solder joint defect". If only the second comparison result is 1, it is labeled as "outer solder joint defect". If the third comparison result is 1, it is labeled as "side solder joint defect". If the fourth comparison result is 1, it is labeled as "pin solder joint defect". If multiple comparison results are 1 simultaneously, it is labeled as "compound solder joint defect" and all the names of the areas with anomalies are listed in sequence. After the defect type is determined, colored marker boxes are superimposed on the corresponding detection areas based on the image coordinate mapping relationship. The first detection area corresponds to a red rectangle, the second to a yellow rectangle, the third to green rectangles on both the left and right sides, and the fourth to a blue rectangle. The coordinates of all marker boxes are automatically derived from the initial area division parameters. The quantized parameters extracted from each area in this inspection—namely, the first red proportion in the center of the pad, the second red proportion at the outer edge of the pad, the number of consecutive target rows on the side of the pad, and the third red proportion in the pin area—are encapsulated together with the colored marker image, the inspection timestamp, the inspection number, and the determination result into a structured data object, and uniformly output to the solder joint defect inspection report template.

[0048] Please see Figure 8 , Figure 8 This is a schematic block diagram of the structure of the AOI device 200 provided in the embodiments of this application, such as... Figure 8 As shown, the AOI device 800 includes: The acquisition module 810 is used to acquire the FOV image of the pad to be inspected, and extract the first red ratio in the center of the pad, the second red ratio at the outer end of the pad, the target row number on the side of the pad, and the third red ratio of the pin based on the FOV image. The comparison module 820 is used to generate a first comparison result based on a first red ratio, a second comparison result based on a second red ratio, a third comparison result based on the target number of rows, and a fourth comparison result based on the third red ratio. The output module 830 is used to mark the type of open weld defect based on the first comparison result, the second comparison result, the third comparison result and the fourth comparison result, and to overlay a colored marker box on the FOV image to output an open weld defect detection report.

[0049] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0050] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0051] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for intelligent detection of welding defects in AOI equipment, characterized in that, include: Acquire an FOV image of the pad to be inspected, and extract the first red ratio in the center of the pad and the second red ratio at the outer end of the pad based on the FOV image; identify the pin window in the FOV image and measure the first size of the pin window; determine the third detection area on the side of the pad and the fourth detection area of ​​the pin based on the first size and the pin window; scan the left and right areas of the third detection area line by line respectively, and extract the target number of rows in the red-green mixed channel that meet the preset first condition; Extract the number of third pixels in the red channel of the fourth detection area that satisfy the preset second condition, and calculate the third proportion of the number of third pixels to the total number of third red pixels in the fourth detection area; A first comparison result is generated based on the first red ratio, a second comparison result is generated based on the second red ratio, a third comparison result is generated based on the target number of rows, and a fourth comparison result is generated based on the third red ratio. Based on the first comparison result, the second comparison result, the third comparison result, and the fourth comparison result, the type of open solder joint defect is marked and a colored marker box is superimposed on the FOV image, and an open solder joint defect detection report is output. Specifically, this includes: establishing a preset combination strategy expression library, which contains a first strategy expression, a second strategy expression, and a third strategy expression; the first strategy expression is defined as performing an addition logical operation between the first comparison result and the second comparison result, followed by a multiplication logical operation with the third comparison result, applicable to conventional CHIP-type components; the second strategy expression is defined as performing a multiplication logical operation between the addition logical operation between the first comparison result and the second comparison result, followed by an addition logical operation with the addition logical operation results of the third and fourth comparison results. This method is applicable to pin-mounted surface mount components. The third strategy expression is defined as performing an addition logic operation sequentially on the first comparison result, the second comparison result, the third comparison result, and the fourth comparison result, and is suitable for components with high reliability requirements. A target combination strategy expression is selected from a preset combination strategy expression library based on the type of component to be tested. The target combination strategy expression includes addition logic operators, multiplication logic operators, and an operation priority identifier. The first comparison result, the second comparison result, the third comparison result, and the fourth comparison result are substituted into the target combination strategy expression according to the operation priority identifier to obtain a comprehensive defect analysis result. Based on the comprehensive defect analysis result, the type of open solder joint defect is marked and a colored marker box is superimposed on the FOV image to output an open solder joint defect detection report.

2. The intelligent detection method for welding defects in AOI equipment according to claim 1, characterized in that, The process of acquiring the field of view (FOV) image of the pad to be inspected, and extracting the first red proportion in the center of the pad and the second red proportion at the outer edge of the pad based on the FOV image, includes: Acquire a field of view (FOV) image of the pads to be inspected, and identify the pad windows in the FOV image; Measure the second dimension of the pad window; Based on the second dimension and the pad window, a first detection area in the center of the pad and a second detection area at the outer end of the pad are determined. Calculate the first red ratio of the first detection area and the second red ratio of the second detection area.

3. The intelligent detection method for welding defects in AOI equipment according to claim 2, characterized in that, The calculation of the first red ratio of the first detection area and the second red ratio of the second detection area includes: Extract the number of first pixels in the red channel in the first detection area, and calculate the first proportion of the number of first pixels to the first total number of red pixels in the first detection area; Extract the number of second pixels in the red channel of the second detection area, and calculate the second red ratio of the number of second pixels to the second total number of pixels in the second detection area.

4. The intelligent detection method for welding defects in AOI equipment according to claim 3, characterized in that, The steps of generating a first comparison result based on the first red ratio, generating a second comparison result based on the second red ratio, generating a third comparison result based on the target row number, and generating a fourth comparison result based on the third red ratio include: The first red ratio is compared with the first threshold. When the first red ratio is less than or equal to the first threshold, the first comparison result is assigned to 0. When the first red ratio is greater than the first threshold, the first comparison result is assigned to 1. The second red ratio is compared with the second threshold. When the second red ratio is greater than or equal to the second threshold, the second comparison result is assigned a value of 0. When the second red ratio is less than the second threshold, the second comparison result is assigned a value of 1. The target number of rows is compared with the row number threshold. When the target number of rows in both the left and right regions of the third detection area is less than or equal to the row number threshold, the third comparison result is assigned a value of 1; otherwise, the third comparison result is assigned a value of 0. The third red ratio is compared with the third threshold. When the third red ratio is greater than or equal to the third threshold, the fourth comparison result is assigned a value of 0. When the third red ratio is less than the third threshold, the fourth comparison result is assigned a value of 1.

5. The intelligent detection method for welding defects in AOI equipment according to claim 4, characterized in that, The step of comparing the target row count with a row count threshold, assigning a value of 1 when the target row counts in both the left and right regions of the third detection region are less than or equal to the row count threshold, and assigning a value of 0 otherwise, yields a third comparison result, including: Compare the target row count in the left region of the third detection region with the row count threshold. When the target number of rows in the left region is less than or equal to the number of rows threshold, the left judgment value is assigned to 1; otherwise, the left judgment value is assigned to 0. The number of target rows in the right region of the third detection area is compared with the row count threshold. When the number of target rows in the right region is less than or equal to the row count threshold, the right judgment value is assigned to 1; otherwise, the right judgment value is assigned to 0. Perform an AND operation on the left-hand judgment value and the right-hand judgment value. If both the left-hand judgment value and the right-hand judgment value are 1, assign the third comparison result as 1; otherwise, assign the third comparison result as 0.

6. The intelligent detection method for welding defects in AOI equipment according to claim 1, characterized in that, The step of substituting the first comparison result, the second comparison result, the third comparison result, and the fourth comparison result into the target combination strategy expression according to the operation priority identifier to obtain the defect comprehensive analysis result includes: Substitute the first comparison result and the second comparison result into the parentheses of the target combination strategy expression, and perform an addition logical operation. When the first comparison result is 0 or the second comparison result is 0, the first intermediate operation result is 0. When both the first comparison result and the second comparison result are 1, the first intermediate operation result is 1. Perform a multiplication logic operation between the first intermediate operation result and the third comparison result. When the first intermediate operation result is 1 or the third comparison result is 1, the operation outputs a second intermediate operation result of 1. When both the first intermediate operation result and the third comparison result are 0, the operation outputs a second intermediate operation result of 0. When the result of the second intermediate calculation is 1, the result of the comprehensive defect analysis is determined to be that the solder joint has a weld breakage defect. When the result of the second intermediate calculation is 0, the result of the comprehensive defect analysis is determined to be that the solder joint does not have a weld breakage defect.

7. The intelligent detection method for welding defects in AOI equipment according to claim 6, characterized in that, The process of labeling the type of open weld defect based on the comprehensive defect analysis results and overlaying a colored marker box on the FOV image, and outputting an open weld defect detection report, includes: When the comprehensive defect analysis result indicates that there is no open weld defect at the weld point, it is marked as qualified. When the comprehensive defect analysis result indicates that there is a weld breakage defect at the weld point, the defect type subdivision process is initiated and the weld breakage defect type is marked. A colored marker box is then overlaid on the FOV image to obtain a weld breakage defect detection report.

8. The intelligent detection method for welding defects in AOI equipment according to claim 7, characterized in that, When the comprehensive defect analysis result indicates that the solder joint has a weld breakage defect, the defect type subdivision process is initiated and the weld breakage defect type is marked. A color-coded marker box is overlaid on the FOV image to obtain a weld breakage defect detection report, including: When the comprehensive defect analysis result indicates that the weld joint has a weld breakage defect, the defect type subdivision process is entered and the weld breakage defect type is marked. A colored marker box is superimposed on the FOV image. The first red ratio, the second red ratio, the target row number, the third red ratio, and the colored marker box are integrated to obtain a weld defect detection report.

9. An AOI device, characterized in that, A method for intelligent detection of welding defects in AOI equipment as described in any one of claims 1-8, comprising: The acquisition module is used to acquire the FOV image of the pad to be inspected, and extract the first red ratio in the center of the pad, the second red ratio at the outer end of the pad, the target row number on the side of the pad, and the third red ratio of the pin based on the FOV image. The comparison module is used to generate a first comparison result based on the first red ratio, a second comparison result based on the second red ratio, a third comparison result based on the target number of rows, and a fourth comparison result based on the third red ratio. The output module is used to mark the type of open weld defect based on the first comparison result, the second comparison result, the third comparison result and the fourth comparison result, and to overlay a colored marker box on the FOV image to output an open weld defect detection report.

Images