Image recognition-based SMT patch quality detection system and method

Abstract

The present application relates to the technical field of electronic manufacturing quality detection, and particularly relates to an SMT patch quality detection system and method based on image recognition. The SMT patch quality detection method based on image recognition comprises the following steps: obtaining a patch image of FPC in a first state, and determining at least one to-be-detected patch object and a corresponding FPC board area of each to-be-detected patch object based on the patch image; obtaining first height data of the FPC in the first state; providing a micro-deformation excitation, applying the micro-deformation excitation to the FPC, so that the FPC generates a recoverable local micro-deformation; obtaining second height data of the FPC under the micro-deformation excitation; determining a height change amount of the to-be-detected patch object and a local deformation amount of the FPC board area based on the first height data and the second height data, and obtaining a deformation response difference feature between the two; and judging the SMT patch quality of the to-be-detected patch object according to the deformation response difference feature.

Description

Technical Field

[0001] This invention relates to the field of electronic manufacturing quality inspection technology, and in particular to an SMT (Surface Mount Technology) chip quality inspection system and method based on image recognition. Background Technology

[0002] Flexible printed circuit boards (FPCs) are characterized by their thinness, flexibility, and high wiring density, and are widely used in consumer electronics, medical electronics, wearable devices, and other fields. During the surface mount technology (SMT) production process of FPCs, it is necessary to perform online inspections of the quality status of mounted components, including component placement, solder joint condition, pin contact, and component coplanarity.

[0003] Current FPC SMT placement quality inspection typically employs two-dimensional or three-dimensional automated optical inspection. Two-dimensional automated optical inspection struggles to acquire height information, limiting its ability to identify three-dimensional defects such as abnormal placement height, floating leads, and coplanarity anomalies. While three-dimensional automated optical inspection can acquire height information, it usually requires establishing a height measurement baseline before defect identification.

[0004] However, FPC substrates are flexible and prone to non-rigid deformations such as warping, denting, and edge curling during production, transportation, mounting, and testing. These deformations can overlap with the height characteristics of components, solder joints, or leads themselves, making it difficult for height detection results to accurately reflect the true bonding state. This can easily lead to misjudgments or missed detections of hidden three-dimensional defects such as floating, lead misfitting, cold solder joints, or coplanarity abnormalities. Summary of the Invention

[0005] On the one hand, the present invention provides an SMT (Surface Mount Technology) chip quality inspection method based on image recognition, comprising the following steps: Acquire patch images of the FPC in a first state, and determine at least one patch object to be detected and the FPC board surface area corresponding to each patch object based on the patch images; Acquire the first height data of the FPC in the first state. The first height data includes the first object height data of the patch object to be detected and the first board surface height data of the FPC board surface area. Provide micro-deformation excitation, apply the micro-deformation excitation to the FPC, so that the FPC produces recoverable local micro-deformation; Acquire the second height data of the FPC under the micro-deformation excitation. The second height data includes the second object height data of the patch object to be detected and the second board surface height data of the FPC board surface area. Based on the first height data and the second height data, the height change of the patch object to be detected and the local deformation of the FPC board area are determined, and the deformation response difference characteristics between the two are obtained. The SMT patch quality of the patch object to be tested is determined based on the deformation response difference characteristics.

[0006] In some embodiments, any of the FPC board areas is an exposed FPC board area located adjacent to the corresponding patch object to be detected and not covered by the patch structure.

[0007] In some embodiments, the application location of the micro-deformation excitation is determined based on the location of the FPC board surface region, and the application direction and / or application intensity of the micro-deformation excitation are configured to cause the FPC board surface region to generate a local deformation for comparison with the height change of the patch object to be detected.

[0008] In some embodiments, the micro-deformation excitation includes at least one of negative pressure adsorption, local lifting, edge tensioning, and support height adjustment.

[0009] In some embodiments, the deformation response difference characteristics include at least one of the following: The height change difference is determined based on the difference between the height change of the patch object to be detected and the local deformation of the corresponding FPC board area. The deformation direction difference is determined based on whether the height change direction of the patch object to be detected is consistent with the local deformation direction of the corresponding FPC board area. The deformation transfer coefficient is determined based on the ratio between the height change of the patch object to be tested and the local deformation of the corresponding FPC board area.

[0010] In some embodiments, when the absolute value of the local deformation of the FPC board area is less than a preset effective deformation threshold, the micro-deformation excitation is adjusted, and the second height data is reacquired.

[0011] In some embodiments, the local deformation of any FPC panel area is obtained through the following steps: Obtain the height difference between multiple height sampling points within the FPC board area and the first height data and the second height data; A local deformation field is constructed for the FPC panel area based on multiple height differences; Based on the corresponding position of the patch object to be tested in the local deformation field corresponding to the FPC board area, a local deformation variable is determined for comparison with the height change of the patch object to be tested.

[0012] In some embodiments, the SMT patch quality of the patch object to be tested is determined based on the deformation response difference characteristics, including the following rules: The surface mount device to be inspected is deemed to have a risk of SMT surface mount quality abnormality when at least one of the following conditions is met: The absolute value of the height change difference is greater than a preset difference threshold; The height change direction of the patch object to be tested is inconsistent with the local deformation direction of the corresponding FPC board area; The deformation transfer coefficient did not fall within the preset normal coefficient range.

[0013] In some embodiments, when a patch object to be tested includes at least two detection sites, the deformation transfer coefficient corresponding to each detection site is determined, and the deformation response difference of multiple sites is determined based on the difference between the deformation transfer coefficients corresponding to any two detection sites; when the deformation response difference of multiple sites is greater than a preset consistency threshold, it is determined that the patch object to be tested has a local adhesion abnormality.

[0014] On the other hand, an SMT placement quality inspection system based on image recognition is characterized by comprising a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the SMT placement quality inspection method based on image recognition proposed in any of the above embodiments.

[0015] The SMT chip quality inspection system and method based on image recognition provided by this invention includes at least the following gains: This invention applies recoverable micro-deformation excitation to an FPC and uses the exposed FPC surface area adjacent to the patch object but not covered by the patch structure as a local deformation reference. It obtains the deformation response difference between the patch object and the corresponding FPC surface area, transforming patch quality inspection from static height judgment to deformation-following relationship judgment. This reduces the interference of non-rigid FPC deformation on height detection results and improves the reliability of detecting latent three-dimensional defects such as floating height, pin misalignment, poor soldering risk, and coplanarity anomalies. Furthermore, adjusting the micro-deformation excitation and re-acquiring height data when local deformation is insufficient improves the effectiveness and stability of the deformation response difference characteristics. Attached Figure Description

[0016] From the following description of embodiments in conjunction with the accompanying drawings, aspects, features, and advantages of the present invention will become clearer and more readily understood, in which: Figure 1 The flowchart illustrates the steps of the image recognition-based SMT chip quality inspection method provided by this invention. Figure 2 This is a schematic diagram of the SMT placement quality inspection system based on image recognition provided in an embodiment of the present invention. Detailed Implementation

[0017] In the following description, specific details such as systems, structures, and techniques are set forth for illustrative purposes and not limiting, in order to provide a thorough understanding of the embodiments of this application. Those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details.

[0018] In the description of this application, detailed descriptions of well-known systems, devices, circuits and methods have been omitted so as not to obscure the description of this application with unnecessary details; in addition, the terms "first", "second", "third", "fourth", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0019] In current FPC SMT placement quality inspection, two-dimensional inspection is difficult to obtain height information, while three-dimensional inspection is easily affected by non-rigid deformations such as FPC warping, depressions, and edge curling. This makes it difficult for the height detection results of components, solder joints, or leads to accurately reflect the true bonding state, which can easily lead to misjudgment or missed detection of hidden defects such as floating height, non-bonded leads, and abnormal coplanarity.

[0020] To address the aforementioned issues, this invention provides an image recognition-based SMT component quality inspection method, aiming to reduce the impact of non-rigid deformation of FPCs on three-dimensional height detection and improve the accuracy of identifying latent defects such as floating height, pin misalignment, cold solder joints, and coplanarity anomalies.

[0021] In one embodiment, see Figure 1 , Figure 1 This is a flowchart illustrating the steps of the image recognition-based SMT chip quality inspection method provided by the present invention. Figure 1 As shown, the SMT chip quality inspection method based on image recognition provided in this embodiment includes the following steps: S01. Obtain the patch image of the FPC in the first state, and determine at least one patch object to be detected and the FPC board surface area corresponding to each patch object based on the patch image.

[0022] In this embodiment, the first state refers to the detection state when the FPC is not subjected to the micro-deformation excitation; further, in different implementations, the first state can specifically be the state in which the FPC is naturally placed on the detection platform, the state in which it is carried by the conveying platform, or the reference clamping state held by the clamping mechanism.

[0023] In this embodiment, the patch image refers to image data used to characterize the surface state of the FPC patch area; further, in different implementations, the patch image can be a two-dimensional image acquired by an area scan camera, a line scan camera, or a microscopic imaging device, or it can be image data synchronously generated by a three-dimensional height detection device and established with a coordinate correspondence with the height data.

[0024] It should be noted that the patch image is used to identify the patch object to be detected on the FPC and to determine the position, outline and distribution area of ​​the patch object to be detected on the FPC.

[0025] Furthermore, based on image features such as grayscale, color, edge, contour, texture, or pad markings in the patch image, the corresponding areas of the component body, pins, solder joints, pads, or other objects that need to be inspected for patch quality can be identified.

[0026] In this embodiment, for any patch object to be detected, its corresponding FPC board area is determined based on the position and outline of the patch object. It can be understood that the FPC board area is used as a reference for the surface deformation of the local area where the patch object is located.

[0027] In one specific implementation, the FPC board surface area is an exposed FPC board surface area located adjacent to the corresponding patch object to be tested and not covered by the patch structure; wherein, the patch structure may include component body, pin, solder joint, solder, adhesive, silkscreen or other structures covering the FPC board surface.

[0028] Furthermore, the adjacent position can be a position within a preset distance range corresponding to the outer contour of the patch object to be detected; the preset distance range can be determined according to the size of the patch object to be detected, the spacing between adjacent components, the image resolution, the height data sampling accuracy, or the detection accuracy requirements.

[0029] In this specific implementation, by selecting the exposed FPC board area located near the patch object to be tested and not covered by the patch structure as a local deformation reference, the interference of long-distance board surface deformation differences and patch structure coverage on height data can be reduced, so that the local deformation can more accurately reflect the FPC board surface deformation state at the location of the patch object to be tested, thereby improving the reliability of deformation response difference characteristics.

[0030] S02. Obtain the first height data of the FPC in the first state. The first height data includes the first object height data of the patch object to be detected and the first board surface height data of the FPC board surface area.

[0031] In this embodiment, the first height data refers to the height detection data of the FPC in the first state, which is used to characterize the height state of the patch object to be detected and its corresponding FPC board area when it is not subjected to micro-deformation excitation.

[0032] Furthermore, the first height data can be a height map, point cloud data, or detection data containing height coordinates; the first height data can be acquired by a 3D camera, a structured light detection device, a laser displacement sensor, a confocal displacement sensor, or other height acquisition devices.

[0033] In this embodiment, based on the position and outline of the patch object to be detected determined in step S01, the height data of the area corresponding to the patch object to be detected is extracted from the first height data as the first object height data; and based on the FPC board area determined in step S01, the height data of the area corresponding to the FPC board area is extracted from the first height data as the first board height data.

[0034] Furthermore, the patch image and the first height data are in the same detection coordinate system; or, the patch image and the first height data are coordinate registered through FPC edges, positioning holes, pad features or image markers, so that the patch object to be detected and its corresponding FPC board area determined based on the patch image can be mapped to the first height data.

[0035] In one specific implementation, the height data of the first object is determined based on the height values ​​of multiple height sampling points within the area corresponding to the patch object to be detected; and the height data of the first board surface is determined based on the height values ​​of multiple height sampling points within the area of ​​the FPC board surface.

[0036] It is understood that the first object height data and the first board surface height data are used as reference data for subsequent calculation of the height change of the patch object to be detected and the local deformation of the FPC board surface area.

[0037] S03. Provide micro-deformation excitation, apply the micro-deformation excitation to the FPC, so that the FPC produces recoverable local micro-deformation.

[0038] In this embodiment, the micro-deformation excitation refers to a controllable external force applied to the FPC to cause a local height change in the FPC within its elastic range.

[0039] It should be noted that the micro-deformation excitation is not used to change the assembly state of the patch object to be detected, but to generate detectable local deformation in the FPC board area so as to compare it with the height change of the patch object to be detected later.

[0040] Furthermore, the application location of the micro-deformation excitation is determined based on the position of the FPC board surface area. Specifically, the micro-deformation excitation can be applied to the local detection area where the FPC board surface area is located, or to a location that can cause local deformation in the FPC board surface area. Thus, the FPC board surface area can generate local deformation corresponding to the patch object to be detected under the micro-deformation excitation.

[0041] Furthermore, the application direction and / or intensity of the micro-deformation excitation can be configured based on at least one of the following: the position of the FPC board area, the material properties of the FPC, the FPC thickness, the support state of the detection platform, and the height detection accuracy, so that the FPC board area generates local deformation that can be recognized by the height detection device. This local deformation is used in subsequent steps to compare with the height change of the patch object to be detected, thereby obtaining deformation response difference characteristics.

[0042] In different implementations, the micro-deformation excitation may include at least one of negative pressure adsorption, local lifting, edge tensioning, and support height adjustment.

[0043] Specifically, the negative pressure adsorption can be achieved through negative pressure holes, negative pressure chambers, or vacuum adsorption platforms located on the back side of the FPC, causing a local height change in the FPC surface area towards the adsorption side; the local lifting can be achieved through lifting components, micro-displacement rods, or liftable support components located on the back side of the FPC, causing a local height change in the FPC surface area away from the support surface; the edge tensioning can be achieved by clamping the edge of the FPC and applying tension force, causing the FPC to locally flatten or change its height under tension; the support height adjustment can be achieved by adjusting the height of the local support components or partitioned support platforms below the FPC, causing a controlled local height change in the corresponding FPC surface area.

[0044] In the above embodiments, by determining the application location of the micro-deformation excitation based on the location of the FPC board area, and configuring the application direction and / or application intensity, the FPC board area generates a stable, recoverable and detectable local deformation, thereby providing a data basis for subsequent judgment of the deformation response difference between the patch object to be detected and the FPC board area.

[0045] S04. Obtain the second height data of the FPC under the micro-deformation excitation. The second height data includes the second object height data of the patch to be detected and the second board surface height data of the FPC board surface area.

[0046] In this embodiment, the second height data refers to the height detection data of the FPC when it is subjected to the micro-deformation excitation, which is used to characterize the height state of the patch object to be detected and its corresponding FPC board area under micro-deformation.

[0047] Furthermore, while maintaining the micro-deformation excitation effect, a second height data of the FPC is acquired; the second height data can be a height map, point cloud data, or detection data containing height coordinates. The acquisition method of the second height data can be the same as that of the first height data, or it can use other height acquisition methods that can be coordinate-corresponded to the first height data.

[0048] In this embodiment, based on the patch object to be detected and its corresponding FPC board area determined in step S01, the second object height data of the patch object to be detected and the second board height data of the FPC board area are extracted from the second height data.

[0049] Furthermore, the second height data and the first height data correspond to the same patch object to be detected and the same FPC board area; the second height data and the first height data are in the same detection coordinate system, or are registered by coordinates through FPC edges, positioning holes, pad features or image markers.

[0050] Understandably, the second object height data and the second board surface height data are used to compare with the first object height data and the first board surface height data to determine the height change of the patch object to be detected and the local deformation of the FPC board surface area, respectively.

[0051] S05. Based on the first height data and the second height data, determine the height change of the patch object to be detected and the local deformation of the FPC board area, and obtain the deformation response difference characteristics between the two.

[0052] In this embodiment, the first height data includes the first object height data of the patch object to be detected and the first board surface height data of the corresponding FPC board area; the second height data includes the second object height data of the patch object to be detected and the second board surface height data of the corresponding FPC board area.

[0053] In this embodiment, the deformation response relationship between the same patch object to be detected and its corresponding FPC board area is determined based on the height change before and after micro-deformation excitation.

[0054] Specifically, for any patch object to be tested, the representative value of the object height in the first state and the representative value of the object height under micro-deformation excitation are determined based on the first object height data and the second object height data, respectively. The representative value of the object height is used to characterize the height state of the patch object under the corresponding detection state.

[0055] In different implementations, the representative value of the object height can be determined based on multiple height sampling points within the area where the patch object to be detected is located. For example, it can be determined using the average height value, median height value, highest point height value, lowest point height value, fitted plane height value, or preset feature point height value of multiple height sampling points. It should be noted that, to ensure comparability of data before and after, the same method for determining the representative value of the object height is preferably used in the first state and under micro-deformation excitation.

[0056] In one specific implementation, let the object height representative value of the patch object to be detected in the first state be H. obj1 The object height represented by H under micro-deformation excitation is obj2 The change in height ΔH of the patch object to be detected obj Determined by the following formula: ΔH obj =H obj2 -H obj1 , where ΔH obj Used to characterize the height change of the patch object under test between the first state and the micro-deformation excitation state.

[0057] Specifically, the local deformation of the FPC board area corresponding to the patch object to be tested is also determined by the first board height data and the second board height data.

[0058] In one specific implementation, multiple height sampling points can be obtained within the FPC board area, and let the height value of the k-th height sampling point in the first height data be H. b1,k The height value in the second altitude data is H. b2,k Then the sampling point deformation ΔH at that height sampling point b,k Determined by the following formula: ΔH b,k =H b2,k -H b1,k The sampling point deformation ΔH based on multiple height sampling points b,k This allows us to determine the local deformation state of the FPC board area under micro-deformation excitation.

[0059] Furthermore, in one specific implementation, based on the deformation ΔH of multiple sampling points b,k The average, median, or weighted average of the values ​​are used to determine the local deformation ΔH of the FPC panel area. b .

[0060] Furthermore, in another specific embodiment, the deformation ΔH of multiple sampling points can be used as a basis. b,k A local deformation field D(x,y) is constructed for the FPC panel area, where x and y represent the planar position coordinates within the FPC panel area, and D(x,y) represents the height change at that position before and after micro-deformation excitation. The local deformation field D(x,y) can be obtained by sampling the deformation ΔH at multiple points. b,k It is obtained through interpolation, fitting, or weighted calculation. For example, it can be constructed using at least one of the following methods: plane fitting, surface fitting, inverse distance weighting, spline interpolation, or local weighted fitting, to characterize the non-uniform deformation state of the FPC plate surface region under micro-deformation excitation.

[0061] After constructing the local deformation field, the local deformation value used for comparison can be determined based on the corresponding position of the patch object to be detected within this local deformation field. Specifically, the corresponding position can be the center position, pin end position, solder joint center position, pad center position, component end position of the patch object to be detected, or a preset feature position determined based on the contour of the patch object to be detected. Let the corresponding position be P. obj The local deformation ΔH is used for comparison with the height change of the patch object to be detected. b This can be expressed as: ΔH b =D(P obj ) Therefore, the local deformation ΔH b It can reflect the actual deformation state of the FPC board surface at the local location of the patch to be tested under micro-deformation excitation, avoiding the inaccurate characterization of local deformation caused by using the board surface deformation far away from the patch to be tested or the average deformation of the whole board.

[0062] Furthermore, based on the height change ΔH of the patch object to be detected obj And the corresponding local deformation ΔH of the FPC panel area b This allows us to obtain the deformation response difference characteristics between the two.

[0063] In one embodiment, the deformation response difference feature includes a height change difference E, which is determined according to the following formula: E=ΔH obj -ΔH bHere, E characterizes the degree of deviation of the height change of the patch object under test relative to the local deformation of the corresponding FPC board area. Furthermore, the absolute value of E is used to determine the consistency of the height changes between the two; when the absolute value of E is small, it indicates that the height changes of the patch object under test and the corresponding FPC board area are relatively consistent under micro-deformation excitation; when the absolute value of E is large, it indicates that the patch object under test deviates from the response of the corresponding FPC board area.

[0064] In another embodiment, the deformation response difference feature includes a difference in deformation direction. Specifically, comparing ΔH obj With ΔH b Whether the positive and negative directions are consistent: When the two signs are consistent, it means that the deformation direction of the patch object to be detected is consistent with that of the FPC board area; when the two signs are inconsistent, it means that the deformation direction of the patch object to be detected is different from that of the FPC board area.

[0065] In another embodiment, the deformation response difference characteristic includes a deformation transfer coefficient K, which is determined according to the following formula: K = ΔH obj / ΔH b , where K is used to characterize the degree to which the patch object being tested follows the deformation of the corresponding FPC board area.

[0066] When K is close to 1, it indicates that the height change of the patch object under test and the corresponding FPC board area are similar under micro-deformation excitation, indicating that the two have a good deformation following relationship. When K is significantly less than 1, it indicates that the patch object under test does not deform synchronously with the FPC board area, which may pose risks of floating, pin not being attached, or poor soldering. When K is significantly greater than 1, it indicates that the height change of the patch object under test is greater than the local deformation of the corresponding FPC board area, which may pose risks of local warping, end lifting, stress amplification, or coplanarity abnormalities. When K is negative, it indicates that the height change direction of the patch object under test and the corresponding FPC board area is opposite, indicating that there is a significant deformation response abnormality between the two.

[0067] To avoid distortion of the deformation transfer coefficient due to excessively small local deformation in the FPC panel area, when ΔH b When the absolute value of the deformation is less than the preset effective deformation threshold, the deformation transfer coefficient K is not calculated, or the current local deformation is determined to be an invalid local deformation.

[0068] Furthermore, when the local deformation ΔH of the FPC board surface area... b When the value is less than the preset effective deformation threshold, it indicates that the FPC board area has not generated an effective local deformation sufficient for deformation response comparison under the current micro-deformation excitation.

[0069] At this point, the micro-deformation excitation can be adjusted, and the second height data can be reacquired; further, the adjustment may include adjusting at least one of the following: the application position, application direction, application intensity, negative pressure magnitude, lifting height, tension force, or support height of the micro-deformation excitation, so that the FPC board area produces a local deformation that meets the comparison requirements.

[0070] The above processing can avoid the instability of deformation transfer coefficient calculation caused by the small local deformation of the FPC board surface area, thereby improving the reliability of deformation response difference characteristics.

[0071] In this embodiment, the deformation response difference feature is used to characterize the degree of deformation following of the patch object to be tested relative to the corresponding FPC board area: when the patch object to be tested and the FPC board area are in normal bonding state, the patch object to be tested can usually produce similar height changes with the corresponding FPC board area under micro-deformation excitation; when the patch object to be tested has floating height, non-bonding, poor soldering or coplanarity abnormality, the deformation following relationship between the patch object to be tested and the corresponding FPC board area is weakened or abnormal, and its height change difference, deformation direction difference or deformation transfer coefficient will deviate.

[0072] S06. Determine the SMT patch quality of the patch object to be tested based on the deformation response difference characteristics.

[0073] In this embodiment, based on the deformation response difference characteristics obtained in step S05, the deformation following state of the patch object to be detected relative to the corresponding FPC board area is determined, and the SMT patch quality of the patch object to be detected is determined according to the determination result.

[0074] Specifically, for any patch object to be tested, its corresponding height change difference, deformation direction difference and / or deformation transfer coefficient are obtained, and the above features are compared with preset quality judgment conditions. The preset quality judgment conditions include at least one of preset difference threshold, preset normal coefficient range and deformation direction consistency condition.

[0075] In one specific implementation, when the deformation response difference feature is a height change difference, the absolute value of the height change difference is compared with a preset difference threshold. If the absolute value of the height change difference is greater than the preset difference threshold, it indicates that there is a deviation beyond the allowable range between the height change of the patch object to be detected and the local deformation of the corresponding FPC board area, and it is determined that the patch object to be detected has an SMT patch quality abnormality risk.

[0076] In another specific implementation, when the deformation response difference feature is a deformation direction difference, it is determined whether the height change direction of the patch object to be tested is consistent with the local deformation direction of the corresponding FPC board area: if the two directions are inconsistent, it indicates that the patch object to be tested has not responded synchronously according to the deformation direction of the corresponding FPC board area, and it is determined that the patch object to be tested has an SMT patch quality abnormality risk.

[0077] In another specific implementation, when the deformation response difference characteristic is the deformation transfer coefficient, the deformation transfer coefficient is compared with a preset normal coefficient range: if the deformation transfer coefficient does not fall within the preset normal coefficient range, it indicates that the deformation following degree of the patch object to be tested relative to the corresponding FPC board area is abnormal, and it is determined that the patch object to be tested has an SMT patch quality abnormality risk.

[0078] The preset difference threshold and the preset normal coefficient range can be determined based on the FPC material characteristics, FPC thickness, type of patch to be tested, micro-deformation excitation method, high detection accuracy, or test data of good samples.

[0079] For example, steps S01 to S05 can be performed on multiple FPC samples with normal patch quality to obtain the height change difference and deformation transfer coefficient corresponding to the normal samples, and the preset difference threshold and the preset normal coefficient range can be determined based on the statistical distribution of the normal samples.

[0080] In one specific implementation, the surface mount device to be tested is determined to have an SMT surface mount quality abnormality risk when at least one of the following conditions is met: the absolute value of the height change difference is greater than a preset difference threshold; the height change direction of the surface mount device to be tested is inconsistent with the local deformation direction of the corresponding FPC board area; and the deformation transfer coefficient does not fall within the preset normal coefficient range.

[0081] Furthermore, the deformation transfer coefficient K is used to characterize the deformation response state of the patch object to be tested relative to the corresponding FPC board area. Specifically, when the corresponding FPC board area generates effective local deformation under micro-deformation excitation, the response relationship between the patch object to be tested and the corresponding FPC board area can be determined based on the value of K.

[0082] When K is within the preset normal coefficient range, it means that the height change of the patch object to be tested matches the local deformation of the corresponding FPC board area. That is, the patch object to be tested can produce synchronous or near-synchronous height changes with the corresponding FPC board area, which usually indicates that the patch object to be tested and the FPC board have a normal bonding or connection state.

[0083] When K is below the lower limit of the preset normal coefficient range, it indicates that the corresponding FPC board area has undergone local deformation, but the height change of the patch object under test is significantly smaller. Specifically, the FPC board area has moved, but the patch object under test has not fully followed the movement. This usually indicates insufficient adhesion or connection between the patch object under test and the FPC board, such as abnormalities such as pins not being attached, insufficient solder joint connection, local cold solder joints, or components floating.

[0084] When K exceeds the upper limit of the preset normal coefficient range, it indicates that the height change of the patch object under test is significantly greater than the local deformation of the corresponding FPC board area. Specifically, the patch object under test exhibits excessive height change relative to the FPC board area, which typically indicates that the patch object under test has undergone relative warping, end lifting, local rotation, or coplanarity abnormalities under micro-deformation excitation. For example, when one end of a component has a strong connection and the other end has a weak connection, local deformation of the FPC board may cause the component to produce a lever-like amplified end displacement.

[0085] When K is negative, or when the height change direction of the patch to be tested is inconsistent with the local deformation direction of the corresponding FPC board area, it indicates that the FPC board area undergoes local deformation in one direction, while the patch to be tested undergoes height change in the opposite direction. This usually indicates that there is obvious relative movement between the patch to be tested and the FPC board, such as local warping, reverse lifting of the end, bonding failure, or abnormal connection status.

[0086] Therefore, different values ​​of the deformation transfer coefficient K do not simply correspond to a single defect type, but rather characterize the deformation response state of the patch object under test relative to the FPC board surface area. When this deformation response state deviates from the normal range, it can be determined that the patch object under test has a risk of SMT patch quality abnormality.

[0087] In another specific embodiment, when a patch object to be tested includes at least two detection areas, step S05 can be performed on each detection area to obtain the deformation transfer coefficient corresponding to each detection area; the detection areas can be different ends, different pins, different solder joints, or other areas that need to be judged for local bonding status of the same component.

[0088] Furthermore, the difference between the deformation transfer coefficients corresponding to any two detection sites is calculated, and the multi-site deformation response difference is determined based on the difference. In a specific example, the maximum value among the differences in the deformation transfer coefficients corresponding to any two detection sites can be used as the multi-site deformation response difference.

[0089] When the difference in deformation response of the multiple parts is greater than the preset consistency threshold, it indicates that the degree of deformation following of different detection parts of the same patch object relative to the corresponding FPC board area is inconsistent. It is determined that the patch object under test has a local bonding abnormality. The local bonding abnormality may include at least one of the following: local non-bonding, local floating, risk of poor soldering, or coplanarity abnormality.

[0090] For example, when a chip component includes a first end and a second end, the deformation transfer coefficient corresponding to the first end and the deformation transfer coefficient corresponding to the second end are determined respectively. If the deformation transfer coefficient corresponding to the first end is within the preset normal coefficient range, while the deformation transfer coefficient corresponding to the second end deviates significantly from the preset normal coefficient range, or the difference between the two is greater than the preset consistency threshold, it indicates that the deformation following degree of the two ends of the chip component is inconsistent, and it can be determined that the chip component has the risk of one end floating, local non-fitting, or coplanarity abnormality.

[0091] For example, when a multi-pin component includes multiple pin detection areas, the deformation transfer coefficient corresponding to each pin detection area can be determined separately. If the difference between the deformation transfer coefficient corresponding to one or more pin detection areas and other pin detection areas is greater than a preset consistency threshold, it indicates that the deformation following degree between the pin detection area and the corresponding FPC board area is abnormal, and it can be determined that the pin detection area has the risk of non-fitting or poor soldering.

[0092] In this embodiment, the quality judgment is not based solely on the static height of the patch object in the first state, but rather on the deformation response relationship between the patch object and the corresponding FPC board area under micro-deformation excitation. Therefore, it can identify hidden three-dimensional defects in static height detection that are easily masked or misjudged by local warping, depressions, or edge curling of the FPC, thus improving the reliability of SMT patch quality inspection.

[0093] The image recognition-based SMT chip quality inspection method provided by this invention applies micro-deformation excitation to the FPC and obtains the height change relationship between the chip object to be inspected and the corresponding FPC board area before and after micro-deformation. It transforms the bonding state of components, pins or solder joints into quantifiable deformation response differences, thereby avoiding defect judgment based solely on static height and improving the accuracy of identifying hidden three-dimensional defects.

[0094] In some embodiments, the present invention also provides an SMT (Surface Mount Technology) chip quality inspection system based on image recognition; please refer to [link to relevant documentation]. Figure 2 , Figure 2 This is a schematic diagram of the SMT placement quality inspection system based on image recognition provided in an embodiment of the present invention.

[0095] In these embodiments, the image recognition-based SMT placement quality inspection system includes an input device, a processor, a memory, and an output device, wherein the input device, processor, memory, and output device are interconnected.

[0096] Further, the input device is used to acquire a patch image of the FPC in a first state, first height data of the FPC in the first state, and second height data of the FPC under micro-deformation excitation. The patch image is used to determine at least one patch object to be detected and the FPC board surface area corresponding to each patch object to be detected; the first height data includes first object height data of the patch object to be detected and first board surface height data of the FPC board surface area; the second height data includes second object height data of the patch object to be detected and second board surface height data of the FPC board surface area.

[0097] Furthermore, the input device may include an image acquisition device, a height acquisition device, and / or a data interface. The image acquisition device is used to acquire patch images of the FPC, the height acquisition device is used to acquire first height data and second height data, and the data interface is used to receive patch images, height data, or micro-deformation excitation state data output from an external acquisition device.

[0098] The memory is used to store computer programs, patch images, first height data, second height data, position and contour information of the patch object to be detected, FPC board area information, deformation response difference characteristics, preset difference threshold, preset normal coefficient range, preset effective deformation threshold and / or preset consistency threshold.

[0099] The processor is used to call and execute the computer program to implement the image recognition-based SMT component quality inspection method described in any of the above embodiments. Specifically, the processor can determine the component object to be inspected and its corresponding FPC board area based on the component image; determine the height change of the component object to be inspected and the local deformation of the FPC board area based on the first height data and the second height data; obtain the deformation response difference characteristics between the two; and determine the SMT component quality of the component object to be inspected based on the deformation response difference characteristics.

[0100] In one specific embodiment, the processor is further configured to output control information for adjusting the micro-deformation excitation when the absolute value of the local deformation in the FPC board area is less than a preset effective deformation threshold, so that the micro-deformation excitation device adjusts the micro-deformation excitation and reacquires the second height data.

[0101] In another specific embodiment, when the patch object to be tested includes at least two detection sites, the processor is further configured to determine the deformation transfer coefficient corresponding to each detection site, and determine the deformation response difference of multiple sites based on the difference between the deformation transfer coefficients corresponding to any two detection sites; when the deformation response difference of multiple sites is greater than a preset consistency threshold, it is determined that the patch object to be tested has a local bonding abnormality.

[0102] The output device is used to output SMT placement quality inspection results. Furthermore, the output device can output the location, defect type, anomaly risk level, deformation response difference characteristics, local deformation, deformation transfer coefficient, and / or re-inspection prompt information of the placement object to be inspected.

[0103] Furthermore, the output device can be at least one of a display, alarm device, printing device, communication interface or data storage interface, for outputting corresponding SMT placement quality inspection results to operators, host computer systems or production line control systems.

[0104] Furthermore, the memory may be a read-only memory, random access memory, flash memory, hard disk, solid-state drive, optical disk, or other non-transitory computer-readable storage medium.

[0105] With the above system structure, the input device can acquire FPC patch images and height data, the processor can obtain deformation response difference characteristics based on the height change relationship before and after micro-deformation excitation, and the output device can output SMT patch quality inspection results, thereby realizing the automatic detection of hidden three-dimensional defects such as floating height, local non-bonding, cold solder joint risk or coplanarity abnormality in FPC deformation scenarios.

[0106] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0107] It should be noted that the above embodiments can be freely combined as needed. The above are merely preferred embodiments of the present invention; it should be observed that those skilled in the art can make several improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for SMT (Surface Mount Technology) component placement quality inspection based on image recognition, characterized in that, Includes the following steps: Acquire patch images of the FPC in a first state, and determine at least one patch object to be detected and the FPC board surface area corresponding to each patch object based on the patch images; Acquire the first height data of the FPC in the first state. The first height data includes the first object height data of the patch object to be detected and the first board surface height data of the FPC board surface area. Provide micro-deformation excitation, apply the micro-deformation excitation to the FPC, so that the FPC produces recoverable local micro-deformation; Acquire the second height data of the FPC under the micro-deformation excitation. The second height data includes the second object height data of the patch object to be detected and the second board surface height data of the FPC board surface area. Based on the first height data and the second height data, the height change of the patch object to be detected and the local deformation of the FPC board area are determined, and the deformation response difference characteristics between the two are obtained. The SMT patch quality of the patch object to be tested is determined based on the deformation response difference characteristics.

2. The SMT chip quality inspection method based on image recognition according to claim 1, characterized in that, Any of the FPC board areas refers to the exposed FPC board area located adjacent to the corresponding patch object to be tested and not covered by the patch structure.

3. The SMT chip quality inspection method based on image recognition according to any one of claims 1 or 2, characterized in that, The application location of the micro-deformation excitation is determined based on the position of the FPC board surface area, and the application direction and / or application intensity of the micro-deformation excitation are configured to cause the FPC board surface area to generate a local deformation for comparison with the height change of the patch object to be detected.

4. The SMT chip quality inspection method based on image recognition according to claim 3, characterized in that, The micro-deformation excitation includes at least one of negative pressure adsorption, local lifting, edge tensioning, and support height adjustment.

5. The SMT chip quality inspection method based on image recognition according to claim 1, characterized in that, The deformation response difference characteristics include at least one of the following: The height change difference is determined based on the difference between the height change of the patch object to be detected and the local deformation of the corresponding FPC board area. The deformation direction difference is determined based on whether the height change direction of the patch object to be detected is consistent with the local deformation direction of the corresponding FPC board area. The deformation transfer coefficient is determined based on the ratio between the height change of the patch object to be tested and the local deformation of the corresponding FPC board area.

6. The SMT chip quality inspection method based on image recognition according to claim 5, characterized in that, When the absolute value of the local deformation of the FPC board area is less than the preset effective deformation threshold, the micro-deformation excitation is adjusted, and the second height data is reacquired.

7. The SMT chip quality inspection method based on image recognition according to claim 5, characterized in that, The local deformation of any of the FPC panel areas is obtained through the following steps: Obtain the height difference between multiple height sampling points within the FPC board area and the first height data and the second height data; A local deformation field is constructed for the FPC panel area based on multiple height differences; Based on the corresponding position of the patch object to be tested in the local deformation field corresponding to the FPC board area, a local deformation variable is determined for comparison with the height change of the patch object to be tested.

8. The SMT chip quality inspection method based on image recognition according to claim 5, characterized in that, Based on the deformation response difference characteristics, the SMT patch quality of the patch object to be tested is determined, including the following rules: The surface mount device to be inspected is deemed to have a risk of SMT surface mount quality abnormality when at least one of the following conditions is met: The absolute value of the height change difference is greater than a preset difference threshold; The height change direction of the patch object to be tested is inconsistent with the local deformation direction of the corresponding FPC board area; The deformation transfer coefficient did not fall within the preset normal coefficient range.

9. The SMT chip quality inspection method based on image recognition according to claim 5, characterized in that, When a patch object to be tested includes at least two detection sites, the deformation transfer coefficient corresponding to each detection site is determined, and the deformation response difference of multiple sites is determined based on the difference between the deformation transfer coefficients corresponding to any two detection sites; when the deformation response difference of multiple sites is greater than a preset consistency threshold, it is determined that the patch object to be tested has a local adhesion abnormality.

10. An SMT (Surface Mount Technology) chip quality inspection system based on image recognition, characterized in that, The device includes a processor and a memory, wherein the memory stores a computer program, which, when executed by the processor, implements the image recognition-based SMT placement quality inspection method as described in any one of claims 1 to 9.

Images