CN103543151B - A kind of measuring method of the printed circuit board (PCB) based on microsection bianry image - Google Patents

Abstract

The invention is a method for measuring a printed circuit board based on a binary image of a metallographic slice. The present invention aims at inputting the bitmap image of the binary image of the metallographic slice of a certain hole in the printed circuit board, that is, the PCB image. The hole divides the white area of the PCB image into two single-connected areas, and proposes to calculate the corresponding hole first. The number of printed circuit board layers in the upper half of the area, the thickness of the outer layer wires, and the thickness of each inner layer wire; then vertically flip the input PCB image to obtain a vertical flip image of the input image; then use it again before calculating the upper half area The calculation method calculates the vertically flipped image of the input image, and obtains the number of PCB layers in the lower half area of the corresponding hole, the thickness of the outer layer wire, and the thickness of each inner layer wire; finally, the calculation results of the upper and lower areas are combined to obtain the PCB of the entire hole. Measurements of the number of layers, outer conductor thickness, and thickness of each inner conductor. The invention can obviously improve the measurement accuracy and efficiency.

Description

A Measurement Method of Printed Circuit Board Based on Binary Image of Metallographic Section

technical field

The invention belongs to the quality detection technology of printed circuit boards, and in particular relates to a method for measuring the number of layers of printed circuit boards (PCB, PCB) and the thickness of conducting wires for detecting the internal quality of printed circuit boards based on binary images of metallographic slices. The invention belongs to the innovative technology of the measurement method of the printed circuit board based on the binary image of the metallographic section.

Background technique

PCB product quality inspection methods based on image processing technology have received more and more attention. The existing PCB quality inspection based on image processing technology mainly focuses on the detection and identification of PCB surface components (Su Mingming. Detection and identification of circuit board components [D]. Dalian University of Technology, 2005; Zhu Yan. Based on machine vision technology Circuit detection and implementation of [D]. Northwestern Polytechnical University, 2007) and PCB surface defect detection (Zheng Wei. Image method to detect printed circuit board defects [D]. Xi'an University of Technology, 2002). At present, the internal quality inspection of PCB based on metallographic slice images mainly adopts manual observation method, and there are few related patents and research reports. The manual observation method is inefficient, the stability of the observation results is poor, and the accuracy cannot be guaranteed.

Contents of the invention

The object of the present invention is to provide a method for measuring printed circuit boards based on binary images of metallographic slices that can significantly improve measurement accuracy and efficiency in consideration of the above problems.

The present invention is based on the measuring method of the printed circuit board of metallographic slice binary image, comprises the following steps:

1) Input the bitmap image of the binary image of the metallographic slice of a hole in the printed circuit board, that is, the PCB image. The hole divides the white area of the PCB image into two single-connected areas, where the horizontal direction is the direction of the hole wall , white stands for conductors, black stands for non-conductors;

2) Calculate the number of layers of the printed circuit board in the upper half area of the corresponding hole, the thickness of the outer layer wire, and the thickness of each inner layer wire. The specific calculation steps are as follows:

21) Calculate the coordinates of the starting and ending pixel points of the wire segment, and the column numbers are marked as B _ij and E _ij ; the wire segment is reflected in the PCB image as a horizontal white line, which is a part of a certain layer of wire; according to the input The bitmap image of the binary image of the metallographic slice of a certain hole in the printed circuit board stipulates that white represents wires and black represents non-wires. It can be found that the wire segment has properties: the initial pixel point is white but the pixel point The pixel whose column number is 1 smaller is black, and the terminal pixel is white but the pixel whose column number is 1 larger than the pixel is black;

22) Determine the maximum pixel row number R ₀ of the PCB image of the area involved when calculating the output parameters;

23) wire segment clustering;

24) Calculate the number of PCB layers, the thickness of the outer layer wires, and the thickness of each inner layer wire;

3) Calculate the number of PCB layers in the lower half area of the corresponding hole, the thickness of the outer layer wire, and the thickness of each inner layer wire. The specific method is:

31) Vertical flip step 1) input PCB image, obtain its vertical flip image and use this image as a new input image;

32) Use the previous calculation method for calculating the upper half area to calculate the vertical flip image of the input image, and obtain the number of PCB layers in the lower half area of the corresponding hole, the thickness of the outer layer wires, and the thickness of each inner layer wire (corresponding to the original input PCB layer number, left outer layer wire thickness, each inner layer wire thickness, right outer layer wire thickness) in the lower half of the PCB image;

4) Combining the calculation results of step 2) and step 3), output the number of layers of the printed circuit board corresponding to the hole, the thickness of the outer layer wires, and the thickness of each inner layer wire.

The present invention aims at inputting the bitmap image of the binary image of the metallographic slice of a certain hole in the printed circuit board, that is, the PCB image. The hole divides the white area of the PCB image into two single-connected areas, and proposes to calculate the corresponding hole first. The number of printed circuit board layers in the upper half of the area, the thickness of the outer layer wires, and the thickness of each inner layer wire; then vertically flip the input PCB image to obtain a vertical flip image of the input image; then use it again before calculating the upper half area The calculation method calculates the vertically flipped image of the input image, and obtains the number of PCB layers in the lower half area of the corresponding hole, the thickness of the outer layer wire, and the thickness of each inner layer wire; finally, the calculation results of the upper and lower areas are combined to obtain the PCB of the entire hole. Measurements of the number of layers, outer conductor thickness, and thickness of each inner conductor. The number of PCB layers and the thickness of wires are important parameters to measure the internal quality of PCB. The invention proposes a method for measuring the number of layers of a printed circuit board and the thickness of conductors based on the binary image of metallographic slices by observing and analyzing the relationship law between the binary image of the PCB metallographic slice and the number of layers of the PCB and the thickness of the conductors. The invention is applied to PCB internal quality detection, and can obviously improve measurement accuracy and measurement efficiency. The invention is a convenient and practical method for measuring printed circuit boards based on binary images of metallographic slices.

Description of drawings

Figure 1 is an example of an input PCB metallographic slice binary bitmap image.

Figure 2 is a schematic illustration of the wire segment and some parameters.

Figure 3 is a vertical flip of the PCB image.

Fig. 4 is a flowchart of the present invention.

detailed description

The present invention is based on the measuring method of the printed circuit board of metallographic slice binary image, comprises the following steps:

1) Input the bitmap image of the binary image of the metallographic slice of a hole in the printed circuit board, that is, the PCB image. The hole divides the white area of the PCB image into two single-connected areas, where the horizontal direction is the direction of the hole wall , white stands for conductors, black stands for non-conductors;

2) Calculate the number of layers of the printed circuit board in the upper half area of the corresponding hole, the thickness of the outer layer wire, and the thickness of each inner layer wire. The specific calculation steps are as follows:

21) Calculate the coordinates of the starting and ending pixel points of the wire segment, and the column numbers are marked as B _ij and E _ij ; the wire segment is reflected in the PCB image as a horizontal white line, which is a part of a certain layer of wire; according to the input The bitmap image of the binary image of the metallographic slice of a certain hole in the printed circuit board stipulates that white represents wires and black represents non-wires. It can be found that the wire segment has properties: the initial pixel point is white but the pixel point The pixel whose column number is 1 smaller is black, and the terminal pixel is white but the pixel whose column number is 1 larger than the pixel is black;

22) Determine the maximum pixel row number R ₀ of the PCB image of the area involved when calculating the output parameters;

23) wire segment clustering;

24) Calculate the number of PCB layers, the thickness of the outer layer wires, and the thickness of each inner layer wire;

3) Calculate the number of PCB layers in the lower half area of the corresponding hole, the thickness of the outer layer wire, and the thickness of each inner layer wire. The specific method is:

31) Vertical flip step 1) input PCB image, obtain its vertical flip image and use this image as a new input image;

32) Use the previous calculation method for calculating the upper half area to calculate the vertical flip image of the input image, and obtain the number of PCB layers in the lower half area of the corresponding hole, the thickness of the outer layer wires, and the thickness of each inner layer wire (corresponding to the original input PCB layer number, left outer layer wire thickness, each inner layer wire thickness, right outer layer wire thickness) in the lower half of the PCB image;

4) Combining the calculation results of step 2) and step 3), output the number of layers of the printed circuit board corresponding to the hole, the thickness of the outer layer wires, and the thickness of each inner layer wire.

The concrete method of above-mentioned step 21) is as follows:

Read the height and width of the image in units of pixels, which are recorded as height H and width W respectively; starting from pixel (0,0), identify the wire segment from left to right and from top to bottom of the input image, and set the i-th pixel The column numbers of the starting and ending pixel points of the jth wire segment of the row are respectively recorded as B _ij and E _ij , i represents the row number where the pixel is located, and j means that _Bij and E _ij belong to the jth wire segment, set The maximum possible value of j is L, that is, the maximum possible number of layers of the PCB, and this patent sets L=30.

The concrete method of above-mentioned step 22) is as follows:

221) Calculate B _min =min{B _ij ; i=0,1,2,..,H, j=1,2,...,L}, that is, B _min is B _ij (i=0,1,2, .., H, j=1,2,..., L) the minimum value of B _ij , the following similar explanation; calculation E _max =max{E _ij ; i=0,1,2,...,H, j= 1,2,...,L};

222) According to the order of i=0,1,2,..,H, j=1,2,...,L, for B _ij , judge that |B _ij -B _min |≤P ₁ and |E _ij -E _max Whether |≤P ₁ is established, once B _ij is found to make it established, set R ₀ =iR _c , and go to step 23); P ₁ is the auxiliary parameter for hole wall position identification, unit: pixel; 100 times PCB image P ₁ =80 , 50 times PCB image P ₁ =[(50/100)*80]=40, the values of P ₁ of other magnification images are similarly set in proportion; R _c is the number of pixel lines intercepted at both ends of the wire measurement; 100 PCB image R _c = 40, 50 times PCB image R _c = [(50/100)*40] = 20, the R _c values of other magnification images are set similarly in proportion; if there is no B that satisfies the above conditions _ij , then output "the PCB may have quality problems, or the image is abnormal", and end;

223) If R ₀ ≤ 0, then output "the PCB may have quality problems, or the image is abnormal", and end.

The concrete method of above-mentioned step 23) is as follows:

231) Calculate the minimum value B _j =min{B _ij , i=0,1,2,...,R ₀ } of the B _ij attribute of the jth type of data when j=1,2,...,L;

232) Calculate the minimum value E _j =min{E _ij ; i=0,1,2,...,R ₀ } of the E _ij attribute of the jth type of data when j=1,2,...,L;

233) Clustering the data pairs B _ij and E _ij , so that each data pair in each class after clustering corresponds to the column number of the left and right borders of the white area of the image where a certain layer of possible wires is located; the specific method is as follows : For i=0,1,2,..,R ₀ , j=1,2,…,L, if B _ij , E _ij are valid data and there is B _m or E _m (1≤m≤L) such that |B _ij -B _m | _≤ P ₂ and | E _{ij -E m |} For B _ij and E _ij : where P ₂ is the clustering auxiliary parameter, unit: pixel; 100 times PCB image P ₂ =80, 50 times PCB image P ₂ =[(50/100)*80]=40, other zoom The P ₂ value of the multiple image is set similarly in proportion:

234) If at least one of B _ij and E _ij does not belong to any category, then output "there may be a quality problem in the PCB, or the image is abnormal", and end.

The concrete method of above-mentioned step 24) is as follows:

241) Set the initial value of the current traverse start pixel row number R ₁ and the current traverse termination pixel row number R ₂ to 0, and the initial value of the number of PCB layers F is set to 0;

242) Assuming that the width of the j-th wire segment in the i-th pixel row is C _ij , then for i=0,1,2,...,R ₀ , j=1,2,...,L, if B _ij , E _ij is valid data, let C _ij =E _ij -B _ij +1;

243) Calculate the number of PCB layers F, the specific method: according to the order of j=1,2,...,L, if there is i (i=0,1,2,..,R ₀ ) so that C _ij ≠0 then F= F+1;

244) Calculate the thickness C _j of the j-th layer conductor, the specific method is as follows: for each j (j=1,2,...,F), take i=0,1,2,...,R ₀ sequentially, if any i makes C _ij ≠0, then set R ₁ =i+R _c , R ₂ =R ₁ +P ₃ -1, calculate C _j =min{C _ij ; R ₁ ≤i≤R ₂ }, P ₃ is used for The length of the measured wire segment (unit: pixel), 100 times PCB image P ₃ =43, 50 times PCB image P ₃ =[(50/100)*43]=21, the value of P ₃ of other magnification images is proportional similar settings;

245) Output the number of PCB layers F, the thickness of the left outer layer conductor C ₀ , the thickness of the right outer layer conductor C _F ; the thickness of the jth inner layer conductor from left to right is C _j , j=2,3,...,F -1, the output is valid only if it is not the initial value.

The above-mentioned 100 times PCB image P ₁ =80, 50 times PCB image P ₁ =[(50/100)*80]=40, the values of P ₁ of other magnification images are set similarly in proportion; R _c is the wire measurement The number of pixel lines intercepted at both ends, 100 times PCB image R _c = 40, 50 times PCB image R _c = [(50/100)*40] = 20, the R _c value of other magnification images is set similarly in proportion .

The 100 times PCB image P ₂ =80, the 50 times PCB image P ₂ =[(50/100)*80]=40, and the values of P ₂ for other magnification images are set similarly in proportion.

The 100 times PCB image P ₃ =43, the 50 times PCB image P ₃ =[(50/100)*43]=21, and the values of P ₃ of other magnification images are set similarly in proportion.

The implementation flowchart of the method for measuring the number of layers of a printed circuit board and the thickness of a wire based on a metallographic slice binary image of the present invention is shown in FIG. 4 . The specific implementation is as follows:

Step 1: Input the bitmap image of the binary image of the metallographic slice of a certain hole in the printed circuit board, that is, the PCB image. The hole divides the white area of the PCB image into two single-connected areas, where the horizontal direction is the hole wall Direction, white for wires, black for non-wires. As shown in Figure 1.

Step 2: Calculate the number of layers of the printed circuit board in the upper half area of the corresponding hole, the thickness of the outer layer wire, and the thickness of each inner layer wire. The specific calculation steps are as follows:

Step 2.1: Calculate the coordinates of the starting and ending pixel points of the wire segment, and the column numbers are marked as B _ij and E _ij . The wire segment is reflected in the PCB image as a horizontal white line, which is a part of a certain layer of wire; according to the bitmap image of the binary image of the metallographic slice of a hole in the input printed circuit board, the white represents the wire , black represents non-wire, and it can be found that the wire segment has properties: the starting pixel point is white but the pixel point smaller than the pixel column number of the pixel point is black, and the ending pixel point is white but smaller than the pixel column number of the pixel point Pixels larger than 1 are black. Specific method: read the height and width of the image in units of pixels, and record them as height H and width W respectively. Starting from pixel (0,0), identify the wire segment from left to right and from top to bottom of the input image, and record the column numbers of the start and end pixel points of the jth wire segment in the i-th pixel row as B _ij and E _ij , i indicates the row number where the pixel is located, j indicates that B _ij and E _ij belong to the jth wire segment, and the maximum possible value of j is set to L, which is the maximum possible number of layers of the PCB. This patent sets L=30.

Step 2.2: Determine the maximum pixel row number R ₀ of the PCB image of the area involved in calculating the output parameters. specific method:

Step 2.2.1: Calculate B _min =min{B _ij ; i=0,1,2,...,H,j=1,2,...,L}, that is, B _min is B _ij (i=0,1 , 2, . . . , H, _j =1, 2, . Calculate E _max =max{E _ij ; i=0, 1, 2, . . . , H, j=1, 2, . . . , L}.

Step 2.2.2: According to the order of i=0,1,2,..,H, j=1,2,...,L, for B _ij , determine that |B _ij -B _min |≤P ₁ and |E _ij Whether -E _max |≤P ₁ is true, once B _ij is found to make it true, set R ₀ =iR _c and go to step 2.3. P ₁ is an auxiliary parameter for hole wall position identification (unit: pixel). 100 times PCB image P ₁ =80, 50 times PCB image P ₁ =[(50/100)*80]=40, the value of P ₁ of other magnification images is set similarly in proportion; R _c is the two The number of pixel rows intercepted by the terminal. The 100x PCB image R _c =40, the 50x PCB image R _c =[(50/100)*40]=20, and the R _c values of other magnification images are set similarly in proportion. If there is no B _ij satisfying the above conditions, output "the PCB may have a quality problem, or the image is abnormal" and end.

Step 2.2.3: If R ₀ ≤ 0, output "the PCB may have quality problems, or the image is abnormal", and end.

Step 2.3: Clustering of wire segments. specific method:

Step 2.3.1: Calculate the minimum value B _j =min{B _ij , i=0,1,2,...,R ₀ } of the B _ij attribute of the jth type of data when j=1,2,...,L .

Step 2.3.2: Calculate the minimum value E _j =min{E _ij ; i=0,1,2,...,R ₀ } of the E _ij attribute of the jth type of data when j=1,2,...,L .

Step 2.3.3: clustering the data pairs B _ij and E _ij , so that each data pair in each class after clustering corresponds to the column number of the left and right borders of the white area of the image where the possible wires of a certain layer are located; Specific method: for i=0,1,2,...,R ₀ , j=1,2,...,L, if B _ij and E _ij are valid data and there is B _m or E _m (1≤m≤L ) so that at least one of |B _ij -B _m |≤P ₂ and |E _ij -E _m |≤P ₂ is established, then the data pair B _ij and E _ij are classified into m categories), so as to obtain a new clustered The data pair B _ij , E _ij . Among them, P ₂ is a clustering auxiliary parameter (unit: pixel). The 100 times PCB image P ₂ =80, the 50 times PCB image P ₂ =[(50/100)*80]=40, and the values of P ₂ of other magnification images are similarly set in proportion.

Step 2.3.4: If at least one of B _ij and E _ij does not belong to any category, then output "there may be a quality problem in the PCB, or the image is abnormal", and end.

Step 2.4: Calculate the number of PCB layers, the thickness of the outer layer wires, and the thickness of each inner layer wire. specific method:

Step 2.4.1: Set the initial values of the current traverse start pixel row number R ₁ and the current traverse end pixel row number R ₂ to 0. The initial value of PCB layer F is set to 0.

Step 2.4.2: Let the width of the j-th wire segment of the i-th pixel row be C _ij , then for i=0,1,2,...,R ₀ , j=1,2,...,L, if B _ij and E _ij are valid data, let C _ij =E _ij -B _ij +1.

Step 2.4.3: Calculate the number of PCB layers F. The specific method: according to the order of j=1, 2, ..., L, if there exists i (i=0, 1, 2, ..., R ₀ ) such that C _ij ≠0, then F=F+1.

Step 2.4.4: Calculate the thickness C _j of the jth layer wire. Specific method: for each j (j=1,2,...,F), take i=0,1,2,...,R ₀ sequentially, if there is i so that C _ij ≠0, then set R ₁ =i +R _c , R ₂ =R ₁ +P ₃ -1, calculate C _j =min{C _ij ; R ₁ ≤i≤R ₂ }. P ₃ Length of wire segment used for measurement (unit: pixel). The 100x PCB image P ₃ =43, the 50x PCB image P ₃ =[(50/100)*43]=21, and the values of P ₃ for other magnification images are set similarly in proportion.

Step 2.4.5: Output the number of PCB layers F, the thickness of the left outer layer conductor C ₀ , and the thickness of the right outer layer conductor C _F ; the thickness of the jth inner layer conductor from left to right is C _j , j=2,3, ..., F-1. The output is valid only if it is not the initial value.

In this way, the number of PCB layers in the upper half area of the corresponding hole, the thickness of the outer layer wires, and the thickness of each inner layer wire are calculated.

Step 3: Calculate the number of PCB layers in the lower half area of the corresponding hole, the thickness of the outer layer wires, and the thickness of each inner layer wire. The specific method is:

Step 3.1: Vertically flip the PCB image input in step 1, get its vertically flipped image and use this image as a new input image.

Step 3.2: Use the previous calculation method for calculating the upper half area to calculate the vertically flipped image of the input image, and obtain the number of PCB layers in the lower half area of the corresponding hole, the thickness of the outer layer wires, and the thickness of each inner layer wire (corresponding to the original Enter the number of PCB layers in the lower half of the PCB image, the thickness of the left outer layer wire, the thickness of each inner layer wire, and the thickness of the right outer layer wire).

Step 4: Combining the calculation results of Step 2 and Step 3, output the number of layers of the printed circuit board corresponding to the hole, the thickness of the outer layer wires, and the thickness of each inner layer wire.

Applying the method of the present invention, the measurement results of the upper half region of the PCB image shown in Figure 1 are as follows: the number of PCB layers is 5, and the thickness of each layer is 72 pixels (left outer layer) and 21 pixels (the first inner layer) from left to right. ), 21 pixels (the second inner layer), 21 pixels (the third inner layer), and 71 pixels (the outer right layer); the measurement results of the lower half area are as follows: the number of PCB layers is 5, and the thickness of each layer is in order from left to right 74 pixels (left outer layer), 21 pixels (1st inner layer), 21 pixels (2nd inner layer), 21 pixels (3rd inner layer), 71 pixels (right outer layer). According to the magnification when the image is collected and the actual size corresponding to each pixel, it is easy to obtain the actual thickness of each layer of wire by using this patent.

According to the method of the invention, the measurement of the number of PCB layers and the thickness of the wires can be realized in a short time and with high precision.

Claims (1)

1. A method for measuring a printed circuit board based on a metallographic section binary image, characterized in that it comprises the steps: 1) Input the bitmap image of the binary image of the metallographic slice of a hole in the printed circuit board, that is, the PCB image. The hole divides the white area of the PCB image into two single-connected areas, where the horizontal direction is the direction of the hole wall , white stands for conductors, black stands for non-conductors; 2) Calculate the number of layers of the printed circuit board in the upper half area of the corresponding hole, the thickness of the outer layer wire, and the thickness of each inner layer wire. The specific calculation steps are as follows: 21) Calculate the coordinates of the starting and ending pixel points of the wire segment, and the column numbers are marked as B _ij and E _ij ; the wire segment is reflected in the PCB image as a horizontal white line, which is a part of a certain layer of wire; according to the input The bitmap image of the binary image of the metallographic slice of a certain hole in the printed circuit board stipulates that white represents wires and black represents non-wires. It can be found that the wire segment has properties: the initial pixel point is white but the pixel point The pixel whose column number is 1 smaller is black, and the terminal pixel is white but the pixel whose column number is 1 larger than the pixel is black; 22) Determine the maximum pixel row number R ₀ of the PCB image of the area involved when calculating the output parameters; 23) wire segment clustering; 24) Calculate the number of PCB layers, the thickness of the outer layer wires, and the thickness of each inner layer wire; 3) Calculate the number of PCB layers in the lower half area of the corresponding hole, the thickness of the outer layer wire, and the thickness of each inner layer wire. The specific method is: 31) Vertical flip step 1) input PCB image, obtain its vertical flip image and use this image as a new input image; 32) Use the calculation method for calculating the upper half area again to calculate the vertical flip image of the input image, and obtain the number of PCB layers, the thickness of the outer layer wires, and the thickness of each inner layer wire in the lower half area of the corresponding hole; 4) Comprehensive step 2) and the calculation result of step 3), output the layer number of the printed circuit board of corresponding hole, outer layer wire thickness, the thickness of every inner layer wire; The concrete method of above-mentioned step 21) is as follows: Read the height and width of the image in units of pixels, which are recorded as height H and width W respectively; starting from pixel (0,0), identify the wire segment from left to right and top to bottom of the input image, and set the i-th pixel The column numbers of the starting and ending pixel points of the jth wire segment of the row are recorded as B _ij and E _ij respectively, i represents the row number where the pixel is located, j means that _Bij and E _ij belong to the jth wire segment, set The maximum possible value of j is L, that is, the maximum possible number of layers of the PCB, set L=30; The concrete method of above-mentioned step 22) is as follows: 221) Calculate B _min =min{B _ij ; i=0,1,2,..,H, j=1,2,...,L}, that is, B _min is B _ij (i=0,1,2, .., H, j=1,2,..., L) the minimum value of B _ij , the following similar explanation; calculation E _max =max{E _ij ; i=0,1,2,...,H, j= 1,2,...,L}; 222) According to the order of i=0,1,2,..,H, j=1,2,...,L, for B _ij , judge that |B _ij -B _min |≤P ₁ and |E _ij -E _max Whether |≤P ₁ is established, once B _ij is found to make it established, set R ₀ =iR _c , and go to step 23); P ₁ is the auxiliary parameter for hole wall position identification, unit: pixel; 100 times PCB image P ₁ =80 , 50 times PCB image P ₁ =[(50/100)*80]=40, the values of P ₁ of other magnification images are similarly set in proportion; R _c is the number of pixel lines intercepted at both ends of the wire measurement; 100 PCB image R _c = 40, 50 times PCB image R _c = [(50/100)*40] = 20, the R _c values of other magnification images are set similarly in proportion; if there is no B that satisfies the above conditions _ij , then output "the PCB may have quality problems, or the image is abnormal", and end; 223) If R ₀ ≤ 0, then output "the PCB may have quality problems, or the image is abnormal", and end; The concrete method of above-mentioned step 23) is as follows: 231) Calculate the minimum value B _j =min{B _ij , i=0,1,2,...,R ₀ } of the B _ij attribute of the jth type of data when j=1,2,...,L; 232) Calculate the minimum value E _j =min{E _ij ; i=0,1,2,...,R ₀ } of the E _ij attribute of the jth type of data when j=1,2,...,L; 233) Clustering the data pairs B _ij and E _ij , so that each data pair in each class after clustering corresponds to the column number of the left and right borders of the white area of the image where a certain layer of possible wires is located; the specific method is as follows : For i=0,1,2,..,R ₀ , j=1,2,…,L, if B _ij , E _ij are valid data and there is B _m or E _m (1≤m≤L) such that _At least one of |B _ij -B _m | _≤ P ₂ and |E _{ij -E m} | B _ij , E _ij : where P ₂ is the clustering auxiliary parameter, unit: pixel; 100 times PCB image P ₂ =80, 50 times PCB image P ₂ =[(50/100)*80]=40, other magnifications _The P2 value of the image is set similarly in proportion: 234) If at least one of B _ij and E _ij does not belong to any category, then output "there may be a quality problem in the PCB, or the image is abnormal", and end; The concrete method of above-mentioned step 24) is as follows: 241) Set the initial value of the current traverse start pixel row number R ₁ and the current traverse termination pixel row number R ₂ to 0, and the initial value of the number of PCB layers F is set to 0; 242) Assuming that the width of the j-th wire segment in the i-th pixel row is C _ij , then for i=0,1,2,...,R ₀ , j=1,2,...,L, if B _ij , E _ij is valid data, let C _ij =E _ij -B _ij +1; 243) Calculate the number of PCB layers F, the specific method: according to the order of j=1,2,...,L, if there is i (i=0,1,2,..,R ₀ ) so that C _ij ≠0 then F= F+1; 244) Calculate the thickness C _j of the j-th layer conductor, the specific method is as follows: for each j (j=1,2,...,F), take i=0,1,2,...,R ₀ sequentially, if any i makes C _ij ≠0, then set R ₁ =i+R _c , R ₂ =R ₁ +P ₃ -1, calculate C _j =min{C _ij ; R ₁ ≤i≤R ₂ }, P ₃ is used for The length of the measured wire segment, 100 times PCB image P ₃ =43, 50 times PCB image P ₃ =[(50/100)*43]=21, the value of P ₃ of other magnification images is set similarly in proportion; 245) Output the number of PCB layers F, the thickness of the left outer layer conductor C ₀ , the thickness of the right outer layer conductor C _F ; the thickness of the jth inner layer conductor from left to right is C _j , j=2,3,...,F -1, the output is valid only if it is not the initial value.