Intelligent detection system and method for copper plating quality

Abstract

The present invention relates to the technical field of intelligent detection of coating quality, and specifically to an intelligent detection system and method for copper coating quality. First, copper-plated samples produced in the same batch are collected by a collection unit to construct a first copper-plated sample set; secondly, the copper-plated samples in the first copper-plated sample set are subjected to appearance quality detection, coating thickness uniformity detection, coating adhesion detection, and coating conductivity detection by a coating quality detection unit to obtain the coating quality detection results; finally, the quality detection pass rate is obtained by an output unit. Through this method, the present invention can not only detect appearance defects, but also detect intrinsic defects in the coating area of ​​the copper-plated sample without destroying the coating of the copper-plated sample, thereby improving the accuracy of coating quality detection.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent detection of coating quality, and in particular to an intelligent detection system and method for copper plating coating quality. Background Art

[0002] In modern manufacturing, copper plating technology is widely used in industries such as electronics, automobiles and aviation to improve the conductivity, corrosion resistance and mechanical strength of metal surfaces. The copper plating process usually involves depositing a copper layer evenly on a substrate. This process is critical to the quality and performance of the product, so the quality of the plating layer needs to be tested.

[0003] The quality of the coating is affected by many factors, including the thickness, uniformity and adhesion of the coating. In the prior art, the quality inspection of the coating is mostly based on machine vision, but machine vision can only detect the surface defects of the coating and cannot evaluate the quality of the coating in combination with internal defects such as the thickness and adhesion of the coating, which will reduce the accuracy of the coating quality inspection results.

[0004] Therefore, an intelligent detection system and method for copper plating quality are proposed. Summary of the invention

[0005] The object of the present invention is to provide an intelligent detection system and method for copper plating quality. Firstly, copper-plated samples produced in the same batch are collected by a collection unit to construct a first copper-plated sample set; secondly, the copper-plated samples in the first copper-plated sample set are subjected to appearance quality detection, coating thickness uniformity detection, coating adhesion detection and coating conductivity detection by a coating quality detection unit to obtain the coating quality detection results; finally, the quality detection pass rate is obtained by an output unit.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A copper plating coating quality intelligent detection system comprises: a collection unit, a plating coating quality detection unit and an output unit.

[0008] Furthermore, the collection unit is used to collect copper-plated samples produced in the same batch to construct a first copper-plated sample set; the detection unit is used to perform quality detection on the copper-plated samples in the first copper-plated sample set to obtain quality detection results; and the output unit is used to obtain the quality detection pass rate based on the quality detection results.

[0009] Furthermore, the detection unit includes: an appearance detection module, a thickness uniformity detection module, an adhesion detection module and a conductivity detection module.

[0010] Furthermore, the appearance inspection module is used to obtain appearance quality inspection results of the coating of the copper-plated samples in the first copper-plated sample set; the appearance quality inspection results include qualified appearance quality inspection and unqualified appearance quality inspection.

[0011] Furthermore, the thickness uniformity detection module is used to detect the uniformity of the thickness of the copper-plated samples with qualified appearance quality detection results. The specific process includes:

[0012] Acquire copper-plated samples with qualified appearance quality test results, and construct a second copper-plated sample set;

[0013] Obtaining the profile of each copper-plated sample in the second copper-plated sample set before copper plating, recorded as a first profile; obtaining the profile of the copper-plated sample, recorded as a second profile; obtaining a plating profile through the first profile and the second profile;

[0014] Selecting N thickness detection points on the coating profile, obtaining the coating thickness of the N thickness detection points, and obtaining a thickness uniformity value according to the coating thickness;

[0015] The thickness uniformity value of each copper-plated sample is compared with the standard thickness uniformity value; when the thickness uniformity value is less than the standard thickness uniformity value, the coating thickness quality test result of the copper-plated sample is that the coating thickness quality test fails; otherwise, the coating thickness quality test result of the copper-plated sample is that the coating thickness quality test passes.

[0016] Furthermore, the calculation formula of the thickness uniformity value is:

[0017] ;

[0018] in, Expressed as the uniform thickness value of the i-th copper-plated sample; Expressed as the coating thickness at the jth thickness detection point; Indicated as the set standard value of coating thickness; It is expressed as the variance of the coating thickness at N thickness detection points in the i-th copper plating sample; Expressed as the uniformity coefficient of the first coating thickness; Expressed as the second coating thickness uniformity coefficient.

[0019] Furthermore, the adhesion detection module is used to perform adhesion detection on the copper-plated samples that have passed the coating thickness quality detection in the second copper-plated sample set to obtain the coating adhesion quality detection results; the specific steps include:

[0020] Acquire copper-plated samples with qualified coating thickness quality test results in the second copper-plated sample set to construct a third copper-plated sample set;

[0021] Selecting M adhesion testing areas of equal area on each of the copper-plated samples;

[0022] Adhere the adhesion test sheet to the M adhesion detection areas, apply tension in opposite directions to the adhesion test sheet and the copper-plated sample, and the tension value is a set fixed value;

[0023] After the force is applied at time T, the pulling force is cancelled and the adhesion test sheet and the adhesive on the copper-plated sample are cleaned;

[0024] Collecting a coating image of the adhesion detection area and obtaining an appearance quality detection result of the adhesion detection area;

[0025] When the appearance quality inspection results of the M adhesion inspection areas are all qualified in the appearance quality inspection, obtaining the third profile of the copper-plated sample; otherwise, the coating adhesion quality inspection result of the copper-plated sample is unqualified in the coating adhesion quality inspection;

[0026] Obtaining the degree of overlap between the third profile and the second profile of each of the copper-plated samples; when the degree of overlap is greater than a set degree of overlap threshold, the coating adhesion quality test result of the corresponding copper-plated sample is that the coating adhesion quality test is qualified; otherwise, the coating adhesion quality test result is that the coating adhesion quality test is unqualified.

[0027] Furthermore, the conductivity detection module performs conductivity detection on the copper-plated samples that have passed the coating adhesion quality detection in the third copper-plated sample set, and the specific steps include:

[0028] Obtaining copper-plated samples with qualified coating adhesion quality test results in the third copper-plated sample set to construct a fourth copper-plated sample set; cleaning impurities on the surface of each copper-plated sample in the fourth copper-plated sample set to obtain multiple conductivity test positions of the copper-plated samples;

[0029] Obtain a first resistivity of the copper-plated sample in each of the conductivity detection positions; compare the first resistivity with the second resistivity of the copper-plated sample before copper plating to obtain a relative resistivity difference, and the calculation formula is: ;in, It is represented by the relative difference in resistivity of the kth copper-plated sample in the fourth copper-plated sample set at the fth conductivity detection position; Expressed as The corresponding first resistivity; Expressed as The corresponding second resistivity;

[0030] When the relative resistivity differences of the kth copper-plated sample at all conductivity detection positions belong to the set relative resistivity difference standard interval, the variance of the relative resistivity difference of the kth copper-plated sample at all conductivity detection positions is obtained; otherwise, the conductivity quality test result of the conductive plating layer is unqualified;

[0031] When the variance is less than a set variance threshold, the test result of the conductivity quality of the copper-plated layer is qualified; otherwise, the test result of the conductivity quality of the copper-plated layer is unqualified.

[0032] Furthermore, the second resistivity includes: detecting the resistivity of each of the copper-plated samples at each of the conductivity detection positions before copper plating, which is recorded as the second resistivity.

[0033] Furthermore, the output unit is used to obtain the quality inspection pass rate based on the quality inspection result, and the specific steps include: comparing the number of copper-plated samples that pass the coating thickness quality inspection with the number of copper-plated samples in the first copper-plated sample set to obtain the quality inspection pass rate.

[0034] An intelligent detection method for copper plating coating quality, comprising:

[0035] Collect copper-plated samples produced in the same batch to construct a first copper-plated sample set;

[0036] Performing quality inspection on the copper-plated samples in the first copper-plated sample set to obtain quality inspection results;

[0037] The quality inspection includes: appearance inspection, thickness uniformity inspection, adhesion inspection and conductivity inspection; the thickness uniformity inspection is to inspect the thickness uniformity of the copper-plated samples with qualified appearance quality inspection results, and the specific process includes:

[0038] Acquire copper-plated samples with qualified appearance quality test results, and construct a second copper-plated sample set;

[0039] Obtaining the profile of each copper-plated sample in the second copper-plated sample set before copper plating, recorded as a first profile; obtaining the profile of the copper-plated sample, recorded as a second profile; obtaining a plating profile through the first profile and the second profile;

[0040] Selecting N thickness detection points on the coating profile, obtaining the coating thickness of the N thickness detection points, and obtaining a thickness uniformity value according to the coating thickness;

[0041] Comparing the thickness uniformity value of each of the copper-plated samples with the standard thickness uniformity value;

[0042] When the thickness uniformity value is less than the standard thickness uniformity value, the coating thickness quality test result of the copper-plated sample is that the coating thickness quality test fails; otherwise, the coating thickness quality test result of the copper-plated sample is that the coating thickness quality test passes;

[0043] The quality inspection pass rate is obtained according to the quality inspection result.

[0044] Furthermore, the appearance inspection is used to obtain appearance quality inspection results of the coating of the copper-plated samples in the first copper-plated sample set; the appearance quality inspection results include qualified appearance quality inspection and unqualified appearance quality inspection.

[0045] Furthermore, the calculation formula of the thickness uniformity value is:

[0046] ;

[0047] in, Expressed as the uniform thickness value of the i-th copper-plated sample; Expressed as the coating thickness at the jth thickness detection point; Indicated as the set standard value of coating thickness; It is expressed as the variance of the coating thickness at N thickness detection points in the i-th copper plating sample; Expressed as the uniformity coefficient of the first coating thickness; Expressed as the second coating thickness uniformity coefficient.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] 1. The present invention performs appearance inspection, thickness uniformity inspection, adhesion inspection and conductivity inspection on the copper-plated sample, and all four inspections are non-destructive inspections. This method can detect not only appearance defects but also intrinsic defects in the copper-plated sample coating area without destroying the copper-plated sample coating. The quality of the copper-plated layer can be fully understood to ensure that its performance in practical applications meets the expected requirements and improve the reliability and durability of the product.

[0050] 2. The present invention performs thickness uniformity testing on copper-plated samples that have passed the appearance quality test, which can avoid the influence of appearance defects of copper-plated samples on the thickness uniformity test results, and uses the profile changes of samples before and after copper plating to obtain the coating profile, and then obtain the thickness of the coating at multiple thickness detection points, and calculate the thickness uniformity value. The thickness uniformity value can not only measure the difference between the coating thickness at each point and the standard thickness, but also measure the variance of the coating thickness at each point. This method can more accurately obtain the coating thickness quality test results.

[0051] 3. The present invention performs conductivity testing on copper-plated samples that have passed the appearance test, thickness uniformity test, and adhesion test. The resistivity of different positions of the copper-plated samples before and after copper plating is tested to obtain the relative difference in resistivity at different positions, and then obtain the conductivity quality test results of the coating. This method can avoid the situation where the appearance test, thickness uniformity test, and adhesion test are unqualified, which affects the conductivity quality test results of the coating. At the same time, the use of multi-point detection can more comprehensively and accurately evaluate the resistivity of the copper-plated samples, avoid potential errors caused by single-position measurement, and ensure higher quality test results. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A structural diagram of an intelligent detection system for copper plating quality provided by an embodiment of the present invention;

[0053] Figure 2 A flow chart of an intelligent detection method for copper plating quality provided by an embodiment of the present invention;

[0054] Figure 3 A flow chart of coating quality detection provided by an embodiment of the present invention;

[0055] Figure 4 This is a structural diagram of a coating quality detection unit provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] In modern manufacturing, copper plating technology is widely used in industries such as electronics, automobiles and aviation to improve the conductivity, corrosion resistance and mechanical strength of metal surfaces. The copper plating process usually involves depositing a copper layer evenly on a substrate. This process is critical to the quality and performance of the product, so the quality of the plating layer needs to be tested.

[0058] The quality of the coating is affected by many factors, including the thickness, uniformity and adhesion of the coating. In the prior art, the quality inspection of the coating is mostly based on machine vision, but machine vision can only detect the surface defects of the coating and cannot evaluate the quality of the coating in combination with internal defects such as the thickness and adhesion of the coating, which will reduce the accuracy of the coating quality inspection results.

[0059] To this end, the present invention provides a copper plating coating quality intelligent detection system, the system is applied to a copper plating coating quality intelligent detection method, the specific system structure diagram and method flow chart refer to Figure 1 and Figure 2 .

[0060] Embodiment 1

[0061] Reference Figure 1 The collection unit in the embodiment is applied to S10 of an intelligent detection method for copper plating quality.

[0062] Furthermore, in this embodiment, 134 copper-plated circuit boards produced by Company A are collected as copper-plated samples to construct a first copper-plated sample set.

[0063] Reference Figure 1 The coating quality detection unit in the embodiment of the present invention is applied to S20 of an intelligent detection method for coating quality of copper plating, and the step of obtaining the quality detection result of S20 is referred to Figure 3 .

[0064] Further, refer to Figure 4 The coating quality detection unit includes: an appearance detection module, a thickness uniformity detection module, an adhesion detection module and a conductivity detection module.

[0065] Furthermore, the appearance inspection module is used to obtain the appearance quality inspection results of the coating of the copper-plated samples in the first copper-plated sample set; the appearance quality inspection results include qualified appearance quality inspection and unqualified appearance quality inspection; the specific steps include: collecting the coating image of each copper-plated sample in the copper-plated sample set; inputting the coating image into the coating defect detection model to obtain the appearance quality inspection result;

[0066] The coating defect detection model includes an input layer, a coating feature map acquisition layer, a coating feature fusion layer, a defect recognition layer and an appearance quality detection result layer;

[0067] The input layer is used to input the coating image into the coating defect detection model;

[0068] The coating feature map acquisition layer is used to perform convolution on the coating image to obtain a plurality of coating feature maps;

[0069] The coating feature fusion layer is used to fuse the multiple coating feature maps to obtain coating appearance features; the coating appearance features include color defect features, bubble defect features, crack defect features, scratch defect features and foreign body defect features; the defect recognition layer is used to recognize defects of the coating based on the appearance features;

[0070] The appearance quality inspection result layer is used to obtain the appearance quality inspection result based on the defect recognition result output by the defect recognition layer; when the defect recognition result is that color defects, bubble defects, crack defects, scratch defects and foreign matter defects do not exist, the appearance quality inspection result is that the appearance quality inspection is qualified; otherwise, the appearance quality inspection result is that the appearance quality inspection is unqualified.

[0071] Furthermore, the step of acquiring a plurality of the coating characteristic maps in the coating characteristic map acquisition layer comprises:

[0072] Input the coating image into a 3×3 convolutional layer with 64 convolution kernels, and add a CBAM attention mechanism after the convolutional layer to output the first coating feature image;

[0073] Input the first coating characteristic image into two 3×3 convolutional layers, each of which has 64 convolution kernels, and output a second coating characteristic image and a third coating characteristic image;

[0074] The first coating characteristic image, the second coating characteristic image and the third coating characteristic image are spliced ​​in the channel dimension, and the spliced ​​characteristic image is input into a 1×1 convolution layer with 32 convolution kernels for convolution, and a fourth coating characteristic image is output;

[0075] Input the fourth coating feature image into two 3×3 convolutional layers, each of which has 80 convolution kernels, and add a CBAM attention mechanism after each convolutional layer to output the fifth coating feature image and the sixth coating feature image;

[0076] Furthermore, the coating feature fusion layer is used to fuse the first coating feature image, the fifth coating feature image and the sixth coating feature image to obtain coating appearance features.

[0077] In this step of the present embodiment, the appearance of the copper-plated sample is inspected, and the copper-plated sample without appearance defects in the coating is obtained through the coating defect detection model, and subsequent thickness uniformity detection, adhesion detection and conductivity detection are performed. This method can avoid inaccurate subsequent detection due to appearance defects in the coating, thereby improving the accuracy of coating quality detection.

[0078] Furthermore, the thickness uniformity detection module is used to detect the uniformity of the thickness of the copper-plated samples with qualified appearance quality detection results. The specific process includes:

[0079] Acquire copper-plated samples with qualified appearance quality test results, and construct a second copper-plated sample set;

[0080] Obtaining the profile of each copper-plated sample in the second copper-plated sample set before copper plating, recorded as a first profile; obtaining the profile of the copper-plated sample, recorded as a second profile; obtaining a plating profile through the first profile and the second profile;

[0081] Selecting N thickness detection points on the coating profile, obtaining the coating thickness of the N thickness detection points, and obtaining a thickness uniformity value according to the coating thickness;

[0082] Compare the thickness uniformity value of each copper-plated sample with the standard thickness uniformity value; when the thickness uniformity value is less than the standard thickness uniformity value, the coating thickness quality test result of the copper-plated sample is unqualified; otherwise, the coating thickness quality test result of the copper-plated sample is qualified; the standard thickness uniformity value is obtained based on experience;

[0083] Further, obtaining the coating profile through the first profile and the second profile includes:

[0084] Selecting M detection points in a copper-plated copper sample; obtaining the coating thickness of the copper-plated copper sample at the M detection points by an X-ray thickness gauge; and obtaining the position of the first contour in the second contour according to the coating thickness;

[0085] The portion between the first contour and the second contour is the plated portion, and the contour of the plated portion is obtained and recorded as the plated contour.

[0086] This step of the present embodiment performs thickness uniformity testing on the copper-plated samples that have passed the appearance quality test, which can avoid the influence of appearance defects of the copper-plated samples on the thickness uniformity test results, and obtains the coating profile by combining the X-ray thickness gauge with the profile changes of the samples before and after copper plating, and can obtain the coating thickness at different parts of the complex profile workpiece, thereby accurately calculating the thickness uniformity value and obtaining the coating thickness quality test results. This method can more accurately obtain the coating thickness quality test results, thereby ensuring the comprehensive quality of the coating.

[0087] Furthermore, the calculation formula of the thickness uniformity value is:

[0088] ;

[0089] in, Expressed as the uniform thickness value of the i-th copper-plated sample; Expressed as the coating thickness at the jth thickness detection point; Indicated as the set standard value of coating thickness; It is expressed as the variance of the coating thickness at N thickness detection points in the i-th copper plating sample; Expressed as the uniformity coefficient of the first coating thickness; Expressed as the uniformity coefficient of the second coating thickness; and The default value is 0.5, and it can be changed according to the specific implementation in actual application.

[0090] In this embodiment, four thickness detection points are selected on each copper-plated sample in the second copper-plated sample set, and the thickness uniformity value is obtained. In this embodiment, four copper-plated samples are selected to obtain thickness uniformity value related data, please refer to Table 1 for details.

[0091] Table 1 Coating thickness quality test table

[0092]

[0093] This step of the present embodiment combines the difference between the coating thickness and the standard value, and also combines the variance of the coating thickness at each thickness detection point to calculate the thickness uniformity value. This method can measure the uniformity of the coating thickness and provide a data basis for judging the coating thickness quality detection results.

[0094] Furthermore, the adhesion detection module is used to perform adhesion detection on the copper-plated samples that have passed the coating thickness quality detection in the second copper-plated sample set to obtain the coating adhesion quality detection results; the specific steps include:

[0095] Acquire copper-plated samples with qualified coating thickness quality test results in the second copper-plated sample set to construct a third copper-plated sample set;

[0096] Selecting M adhesion testing areas of equal area on each of the copper-plated samples;

[0097] Adhere the adhesion test sheet to the M adhesion detection areas, apply tension in opposite directions to the adhesion test sheet and the copper-plated sample, the tension value is a set fixed value, and the direction of the tension is perpendicular to the adhesion test sheet; the material of the adhesion test sheet can be metal, plastic, etc.;

[0098] After the force is applied at time T, the pulling force is cancelled and the adhesion test sheet and the adhesive on the copper-plated sample are cleaned;

[0099] Collecting a coating image of the adhesion detection area and obtaining an appearance quality detection result of the adhesion detection area;

[0100] When the appearance quality inspection results of the M adhesion inspection areas are all qualified in the appearance quality inspection, obtaining the third profile of the copper-plated sample; otherwise, the coating adhesion quality inspection result of the copper-plated sample is unqualified in the coating adhesion quality inspection;

[0101] The degree of overlap between the third contour and the second contour of each copper-plated sample is obtained; when the degree of overlap is greater than a set degree of overlap threshold, the coating adhesion quality test result of the corresponding copper-plated sample is that the coating adhesion quality test is qualified; otherwise, the coating adhesion quality test result is that the coating adhesion quality test is unqualified; the degree of overlap threshold is obtained based on experience.

[0102] This step of the present embodiment performs adhesion testing on the copper-plated samples that have passed the coating thickness quality test, which can avoid inaccurate adhesion testing caused by unqualified coating thickness. At the same time, a non-destructive method is used to test the adhesion of different areas of the coating of the copper-plated sample, and the adhesion test results can be accurately obtained by combining the appearance change of the coating after the tensile force is applied and the qualified contour change, thereby ensuring the comprehensive quality of the coating.

[0103] Furthermore, the conductivity detection module performs conductivity detection on the copper-plated samples that have passed the coating adhesion quality detection in the third copper-plated sample set, and the specific steps include:

[0104] Obtain copper-plated samples with qualified coating adhesion quality test results in the third copper-plated sample set to construct a fourth copper-plated sample set; clean impurities on the surface of each copper-plated sample in the fourth copper-plated sample set to obtain P conductivity test positions of the copper-plated samples;

[0105] Obtain a first resistivity of the copper-plated sample in each of the conductivity detection positions; compare the first resistivity with the second resistivity of the copper-plated sample before copper plating to obtain a relative resistivity difference, and the calculation formula is: ;in, It is represented by the relative difference in resistivity of the kth copper-plated sample in the fourth copper-plated sample set at the fth conductivity detection position; Expressed as The corresponding first resistivity; Expressed as The corresponding second resistivity; the first resistivity and the second resistivity can be obtained by a four-probe method or the like;

[0106] When the relative resistivity differences of the kth copper-plated sample at all conductivity detection positions belong to the set relative resistivity difference standard interval, the variance of the relative resistivity difference of the kth copper-plated sample at all conductivity detection positions is obtained; otherwise, the conductivity quality test result of the conductive plating layer is unqualified;

[0107] When the variance is less than a set variance threshold, the test result of the conductivity quality of the copper-plated layer is qualified; otherwise, the test result of the conductivity quality of the copper-plated layer is unqualified; the variance threshold is obtained based on experience.

[0108] The conductivity of the coating in this step of the present embodiment is not only affected by the thickness of the coating, but also by the composition of the coating. When non-copper components appear in the copper-plated coating, the resistivity of the coating will also be affected. In this embodiment, the conductivity test is performed on the copper-plated samples that are qualified in the appearance test, thickness uniformity test and adhesion test. The resistivity of different positions of the copper-plated samples before and after copper plating is tested to obtain the relative difference in resistivity at different positions, and then obtain the coating conductivity quality test results of the coating. This method can avoid the situation where the appearance test, thickness uniformity test and adhesion test are unqualified, which affects the coating conductivity quality test results. At the same time, the use of multi-point detection can more comprehensively and accurately evaluate the resistivity of the copper-plated sample, avoid the potential errors caused by single-position measurement, and ensure higher quality test results.

[0109] Furthermore, the second resistivity includes: detecting the resistivity of each of the copper-plated samples at each of the conductivity detection positions before copper plating, which is recorded as the second resistivity.

[0110] This step of the present embodiment obtains the resistivity at each conductivity detection position before copper plating. This method can provide a data basis for the conductivity detection of the copper-plated sample after copper plating and obtain the degree of resistivity change of the copper-plated sample after copper plating.

[0111] Reference Figure 1 The output unit is applied to S30 of an intelligent detection method for copper plating quality.

[0112] Furthermore, the output unit is used to compare the number of copper-plated samples that pass the coating thickness quality inspection with the number of copper-plated samples in the first copper-plated sample set to obtain a quality inspection pass rate.

[0113] In this step of the present embodiment, the copper-plated sample is subjected to appearance inspection, thickness uniformity inspection, adhesion inspection and conductivity inspection, and all four inspections are non-destructive inspections. This method can detect not only appearance defects but also intrinsic defects in the copper-plated sample coating area without destroying the copper-plated sample coating. The quality of the copper-plated layer can be fully understood to ensure that its performance in practical applications meets the expected requirements, thereby improving the reliability and durability of the product.

[0114] In this embodiment, the coating quality of 134 copper-plated circuit boards produced by Company A was tested. The test results are shown in Table 2.

[0115] Table 2 Coating quality test results

[0116]

[0117] From Table 2, it can be obtained that the quality inspection pass rate of the coating of 134 copper-plated circuit boards is 97.01%.

[0118] Embodiment 2

[0119] Reference Figure 1 The collection unit in the embodiment is applied to S10 of an intelligent detection method for copper plating quality.

[0120] Furthermore, in this embodiment, 148 copper-plated switches produced by Company B are collected as copper-plated samples to construct a first copper-plated sample set.

[0121] Reference Figure 1 The coating quality detection unit is applied to S20 of an intelligent detection method for coating quality for copper plating.

[0122] Furthermore, the coating quality detection unit includes: an appearance detection module, a thickness uniformity detection module, an adhesion detection module and a conductivity detection module.

[0123] Furthermore, the appearance inspection module is used to obtain the appearance quality inspection results of the coating of the copper-plated samples in the first copper-plated sample set; the appearance quality inspection results include qualified appearance quality inspection and unqualified appearance quality inspection; the specific steps include: collecting the coating image of each copper-plated sample in the copper-plated sample set; inputting the coating image into the coating defect detection model to obtain the appearance quality inspection result;

[0124] The coating defect detection model includes an input layer, a coating feature image acquisition layer, a coating feature fusion layer, a defect recognition layer and an appearance quality detection result layer; the input layer is used to input the coating image into the coating defect detection model;

[0125] The coating feature map acquisition layer is used to perform convolution on the coating image to obtain a plurality of coating feature maps;

[0126] The coating feature fusion layer is used to fuse the multiple coating feature maps to obtain coating appearance features; the coating appearance features include color defect features, bubble defect features, crack defect features, scratch defect features and foreign body defect features; the defect recognition layer is used to recognize defects of the coating based on the appearance features;

[0127] The appearance quality inspection result layer is used to obtain the appearance quality inspection result based on the defect recognition result output by the defect recognition layer; when the defect recognition result is that color defects, bubble defects, crack defects, scratch defects and foreign matter defects do not exist, the appearance quality inspection result is that the appearance quality inspection is qualified; otherwise, the appearance quality inspection result is that the appearance quality inspection is unqualified.

[0128] Furthermore, the step of acquiring a plurality of the coating characteristic maps in the coating characteristic map acquisition layer comprises:

[0129] Input the coating image into a 3×3 convolutional layer with 64 convolution kernels, and add a CBAM attention mechanism after the convolutional layer to output the first coating feature image;

[0130] Input the first coating characteristic image into two 3×3 convolutional layers, each of which has 64 convolution kernels, and output a second coating characteristic image and a third coating characteristic image;

[0131] The first coating characteristic image, the second coating characteristic image and the third coating characteristic image are spliced ​​in the channel dimension, and the spliced ​​characteristic image is input into a 1×1 convolution layer with 32 convolution kernels for convolution, and a fourth coating characteristic image is output; the fourth coating characteristic image is input into two 3×3 convolution layers, and the number of convolution kernels of each convolution layer is 80, and a CBAM attention mechanism is added after each convolution layer to output a fifth coating characteristic image and a sixth coating characteristic image;

[0132] Furthermore, the coating feature fusion layer is used to fuse the first coating feature image, the fifth coating feature image and the sixth coating feature image to obtain coating appearance features.

[0133] Furthermore, the thickness uniformity detection module is used to detect the uniformity of the thickness of the copper-plated samples with qualified appearance quality detection results. The specific process includes:

[0134] Acquire copper-plated samples with qualified appearance quality test results, and construct a second copper-plated sample set;

[0135] Obtaining the profile of each copper-plated sample in the second copper-plated sample set before copper plating, recorded as a first profile; obtaining the profile of the copper-plated sample, recorded as a second profile; obtaining a plating profile through the first profile and the second profile;

[0136] Selecting N thickness detection points on the coating profile, obtaining the coating thickness of the N thickness detection points, and obtaining a thickness uniformity value according to the coating thickness;

[0137] The thickness uniformity value of each copper-plated sample is compared with the standard thickness uniformity value; when the thickness uniformity value is less than the standard thickness uniformity value, the coating thickness quality test result of the copper-plated sample is that the coating thickness quality test fails; otherwise, the coating thickness quality test result of the copper-plated sample is that the coating thickness quality test passes; the standard thickness uniformity value is obtained based on experience.

[0138] Furthermore, the calculation formula of the thickness uniformity value is:

[0139] ;

[0140] in, Expressed as the uniform thickness value of the i-th copper-plated sample; Expressed as the coating thickness at the jth thickness detection point; Indicated as the set standard value of coating thickness; It is expressed as the variance of the coating thickness at N thickness detection points in the i-th copper plating sample; Expressed as the uniformity coefficient of the first coating thickness; Expressed as the uniformity coefficient of the second coating thickness; and The default value is 0.5, and it can be changed according to the specific implementation in actual application.

[0141] In this embodiment, four thickness detection points are selected on each copper-plated sample in the second copper-plated sample set, and the thickness uniformity value is obtained. In this embodiment, four copper-plated samples are selected to obtain thickness uniformity value related data, please refer to Table 3 for details.

[0142] Table 3 Coating thickness quality test table

[0143]

[0144] Furthermore, the adhesion detection module is used to perform adhesion detection on the copper-plated samples that have passed the coating thickness quality detection in the second copper-plated sample set to obtain the coating adhesion quality detection results; the specific steps include:

[0145] Obtain copper-plated samples with qualified coating thickness quality test results from the second copper-plated sample set to construct a third copper-plated sample set; select M adhesion test areas of equal area on each of the copper-plated samples;

[0146] Adhere the adhesion test sheet to the M adhesion detection areas, apply tension in opposite directions to the adhesion test sheet and the copper-plated sample, and the tension value is a set fixed value;

[0147] After the force is applied at time T, the pulling force is cancelled and the adhesion test sheet and the adhesive on the copper-plated sample are cleaned;

[0148] Collecting a coating image of the adhesion detection area and obtaining an appearance quality detection result of the adhesion detection area;

[0149] When the appearance quality inspection results of the M adhesion inspection areas are all qualified in the appearance quality inspection, the third contour of the copper-plated sample is obtained; otherwise, the coating adhesion quality inspection result of the copper-plated sample is that the coating adhesion quality inspection is unqualified; the overlap between the third contour and the second contour of each copper-plated sample is obtained; when the overlap is greater than a set overlap threshold, the coating adhesion quality inspection result of the corresponding copper-plated sample is that the coating adhesion quality inspection is qualified; otherwise, the coating adhesion quality inspection result is that the coating adhesion quality inspection is unqualified.

[0150] Furthermore, the conductivity detection module performs conductivity detection on the copper-plated samples that have passed the coating adhesion quality detection in the third copper-plated sample set, and the specific steps include:

[0151] Obtain copper-plated samples with qualified coating adhesion quality test results in the third copper-plated sample set to construct a fourth copper-plated sample set; clean impurities on the surface of each copper-plated sample in the fourth copper-plated sample set to obtain P conductivity test positions of the copper-plated samples;

[0152] Obtain a first resistivity of the copper-plated sample in each of the conductivity detection positions; compare the first resistivity with the second resistivity of the copper-plated sample before copper plating to obtain a relative resistivity difference, and the calculation formula is: ;in, It is represented by the relative difference in resistivity of the kth copper-plated sample in the fourth copper-plated sample set at the fth conductivity detection position; Expressed as The corresponding first resistivity; Expressed as The corresponding second resistivity;

[0153] When the relative resistivity differences of the kth copper-plated sample at all conductivity detection positions belong to the set relative resistivity difference standard interval, the variance of the relative resistivity difference of the kth copper-plated sample at all conductivity detection positions is obtained; otherwise, the conductivity quality test result of the conductive plating layer is unqualified;

[0154] When the variance is less than a set variance threshold, the test result of the conductivity quality of the copper-plated layer is qualified; otherwise, the test result of the conductivity quality of the copper-plated layer is unqualified.

[0155] Furthermore, the second resistivity includes: detecting the resistivity of each of the copper-plated samples at each of the conductivity detection positions before copper plating, which is recorded as the second resistivity.

[0156] Reference Figure 1 The output unit is applied to S30 of an intelligent detection method for copper plating quality.

[0157] Furthermore, the output unit is used to compare the number of copper-plated samples that pass the coating thickness quality inspection with the number of copper-plated samples in the first copper-plated sample set to obtain a quality inspection pass rate.

[0158] In this embodiment, the coating quality of 148 copper-plated switches produced by Company B was tested. The test results are shown in Table 4.

[0159] Table 4 Plating quality test results

[0160]

[0161] From Table 4, it can be obtained that the quality inspection pass rate of 148 copper-plated switch coatings is 97.97%.

[0162] In this embodiment, the copper-plated sample is subjected to appearance inspection, thickness uniformity inspection, adhesion inspection and conductivity inspection, and all four inspections are non-destructive inspections. This method can detect not only appearance defects, but also intrinsic defects in the copper-plated sample coating area without destroying the copper-plated sample coating. The quality of the copper-plated layer can be fully understood to ensure that its performance in practical applications meets the expected requirements and improve the reliability and durability of the product. In this embodiment, the copper-plated sample that has passed the appearance quality inspection is subjected to thickness uniformity inspection, which can avoid the influence of the appearance defects of the copper-plated sample on the thickness uniformity inspection result, and the contour change of the sample before and after copper plating is used to obtain the coating contour, and then the thickness of the coating at multiple thickness inspection points is obtained to calculate the thickness uniformity value. The thickness uniformity value can not only measure the difference between the coating thickness at each point and the standard thickness, but also measure the variance of the coating thickness at each point. Through this method, the coating thickness quality inspection result can be obtained more accurately. In this embodiment, the copper-plated sample that has passed the appearance inspection, thickness uniformity inspection and adhesion inspection is subjected to conductivity inspection. The resistivity of different positions of the copper-plated sample before and after copper plating is tested to obtain the relative difference in resistivity at different positions, and then obtain the test results of the conductivity quality of the coating. This method can avoid the situation where the appearance test, thickness uniformity test and adhesion test are unqualified, which affects the test results of the conductivity quality of the coating. At the same time, the use of multi-point detection can more comprehensively and accurately evaluate the resistivity of the copper-plated sample, avoid the potential errors caused by single-position measurement, and ensure higher quality test results.

[0163] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent detection system for copper plating quality, characterized in that: include: Collection unit: used for collecting copper-plated samples produced in the same batch and constructing a first copper-plated sample set; A coating quality detection unit: used for performing quality detection on the copper-plated samples in the first copper-plated sample set to obtain quality detection results; The coating quality detection unit includes: an appearance detection module, a thickness uniformity detection module, an adhesion detection module and a conductivity detection module; the thickness uniformity detection module is used to detect the coating thickness uniformity of the copper-plated samples with qualified appearance quality detection results. The specific process includes: Acquire copper-plated samples with qualified appearance quality test results, and construct a second copper-plated sample set; Obtaining the profile of each copper-plated sample in the second copper-plated sample set before copper plating, recorded as a first profile; obtaining the profile of the copper-plated sample, recorded as a second profile; obtaining a plating profile through the first profile and the second profile; Select N thickness detection points on the coating profile, obtain the coating thickness of the N thickness detection points, and obtain the thickness uniformity value according to the coating thickness; the calculation formula of the thickness uniformity value is: Among them, hdjy i It is expressed as the uniform thickness of the i-th copper-plated sample; hd (i,j) Expressed as the coating thickness at the jth thickness detection point; Indicates the set coating thickness standard value; VAR i It is expressed as the variance of the coating thickness at N thickness detection points in the i-th copper plating sample; α1 is expressed as the uniformity coefficient of the first coating thickness; α2 is expressed as the uniformity coefficient of the second coating thickness Compare the thickness uniformity value of each copper-plated sample with the standard thickness uniformity value; when the thickness uniformity value is less than the standard thickness uniformity value, the coating thickness quality test result of the copper-plated sample is unqualified; otherwise, the coating thickness quality test result of the copper-plated sample is qualified; Output unit: used to obtain the quality inspection pass rate according to the quality inspection result.

2. The intelligent detection system for copper plating quality according to claim 1, characterized in that: The appearance inspection module is used to obtain appearance quality inspection results of the coatings of the copper-plated samples in the first copper-plated sample set; the appearance quality inspection results include appearance quality inspection qualified and appearance quality inspection unqualified.

3. The intelligent detection system for copper plating quality according to claim 1, characterized in that: The adhesion detection module is used to perform adhesion detection on the copper-plated samples that have passed the coating thickness quality detection in the second copper-plated sample set, and obtain the coating adhesion quality detection results; the specific steps include: Acquire copper-plated samples with qualified coating thickness quality test results in the second copper-plated sample set to construct a third copper-plated sample set; Selecting M adhesion testing areas of equal area on each of the copper-plated samples; Adhere the adhesion test sheet to the M adhesion detection areas, apply tension in opposite directions to the adhesion test sheet and the copper-plated sample, and the tension value is a set fixed value; After the force is applied at time T, the pulling force is cancelled and the adhesion test sheet and the adhesive on the copper-plated sample are cleaned; Collecting a coating image of the adhesion detection area and obtaining an appearance quality detection result of the adhesion detection area; When the appearance quality inspection results of the M adhesion inspection areas are all qualified in the appearance quality inspection, obtaining the third profile of the copper-plated sample; otherwise, the coating adhesion quality inspection result of the copper-plated sample is unqualified in the coating adhesion quality inspection; Obtaining the degree of overlap between the third profile and the second profile of each of the copper-plated samples; when the degree of overlap is greater than a set degree of overlap threshold, the coating adhesion quality test result of the corresponding copper-plated sample is that the coating adhesion quality test is qualified; otherwise, the coating adhesion quality test result is that the coating adhesion quality test is unqualified.

4. The intelligent detection system for copper plating quality according to claim 1, characterized in that: The conductivity detection module performs conductivity detection on the copper-plated samples that have passed the coating adhesion quality detection in the third copper-plated sample set, and the specific steps include: Obtain copper-plated samples with qualified coating adhesion quality test results in the third copper-plated sample set to construct a fourth copper-plated sample set; clean impurities on the surface of each copper-plated sample in the fourth copper-plated sample set to obtain P conductivity test positions of the copper-plated samples; Obtain a first resistivity of the copper-plated sample in each of the conductivity detection positions; compare the first resistivity with the second resistivity of the copper-plated sample before copper plating to obtain a relative resistivity difference, and the calculation formula is: Among them, dzl (k,f) It is represented by the relative difference in resistivity of the kth copper-plated sample in the fourth copper-plated sample set at the fth conductivity detection position; Represented as dzl (k,f) The corresponding first resistivity; Represented as dzl (k,f) The corresponding second resistivity; When the relative resistivity differences of the kth copper-plated sample at all conductivity detection positions belong to the set relative resistivity difference standard interval, the variance of the relative resistivity difference of the kth copper-plated sample at all conductivity detection positions is obtained; otherwise, the conductivity quality test result of the conductive plating layer is unqualified; When the variance is less than a set variance threshold, the test result of the conductivity quality of the coating of the copper-plated sample is qualified; otherwise, the test result of the conductivity quality of the coating is unqualified.

5. The intelligent detection system for copper plating quality according to claim 4, characterized in that: The second resistivity includes: detecting the resistivity of each of the copper-plated samples at each of the conductivity detection positions before copper plating, which is recorded as the second resistivity.

6. The intelligent detection system for copper plating quality according to claim 1, characterized in that: The output unit includes: comparing the number of copper-plated samples that pass the coating conductivity quality test with the number of copper-plated samples in the first copper-plated sample set to obtain a quality test pass rate.

7. An intelligent detection method for copper plating quality, characterized in that: include: Collect copper-plated samples produced in the same batch to construct a first copper-plated sample set; Performing quality inspection on the copper-plated samples in the first copper-plated sample set to obtain quality inspection results; The quality inspection includes: appearance inspection, thickness uniformity inspection, adhesion inspection and conductivity inspection; the thickness uniformity inspection is to inspect the thickness uniformity of the copper-plated samples with qualified appearance quality inspection results, and the specific process includes: Acquire copper-plated samples with qualified appearance quality test results, and construct a second copper-plated sample set; Obtaining the profile of each copper-plated sample in the second copper-plated sample set before copper plating, recorded as a first profile; obtaining the profile of the copper-plated sample, recorded as a second profile; obtaining a plating profile through the first profile and the second profile; Select N thickness detection points on the coating profile, obtain the coating thickness of the N thickness detection points, and obtain the thickness uniformity value according to the coating thickness; the calculation formula of the thickness uniformity value is: Among them, hdjy i It is expressed as the uniform thickness of the i-th copper-plated sample; hd (i,j) Expressed as the coating thickness at the jth thickness detection point; Indicates the set coating thickness standard value; VAR i It is expressed as the variance of the coating thickness at N thickness detection points in the i-th copper-plated sample; α1 is expressed as the uniformity coefficient of the first coating thickness; α2 is expressed as the uniformity coefficient of the second coating thickness; Compare the thickness uniformity value of each copper-plated sample with the standard thickness uniformity value; when the thickness uniformity value is less than the standard thickness uniformity value, the coating thickness quality test result of the copper-plated sample is unqualified; otherwise, the coating thickness quality test result of the copper-plated sample is qualified; The quality inspection pass rate is obtained according to the quality inspection result.

8. The intelligent detection method for copper plating quality according to claim 7, characterized in that: The appearance inspection is used to obtain appearance quality inspection results of the coatings of the copper-plated samples in the first copper-plated sample set; the appearance quality inspection results include qualified appearance quality inspection and unqualified appearance quality inspection.

Images