Electrode foil fault thickness measuring method based on microscopic image segmentation

Abstract

The invention provides an electrode foil fault thickness measurement method based on microscopic image segmentation, and relates to the technical field of digital image processing, and the method comprises the steps: obtaining an electrode foil fault microscopic image through high-resolution microscopic equipment, and carrying out the image enhancement through gray level histogram equalization, so as to improve the image contrast. And then, carrying out line-by-line statistics on pixel gray scale features of the enhanced microscopic image except a label region, determining a minimum gray scale average value position, and cutting a target region according to the minimum gray scale average value position. Then, adaptive threshold segmentation and morphological processing are carried out on the cutting area, and an accurate segmentation result is obtained; meanwhile, fixed threshold segmentation is carried out on a label area, a pixel scale is extracted, and the actual size of a single pixel is calculated. And finally, counting the pixel area of the segmented region, calculating the average pixel thickness, and converting the average pixel thickness into the actual fault average thickness. According to the scheme, the automation degree is high, measurement is accurate, adaptability is high, and the efficiency and accuracy of electrode foil fault thickness measurement are effectively improved.

Description

Technical Field

[0001] The invention relates to the technical field of digital image processing, and in particular to a method for measuring electrode foil fault thickness based on microscopic image segmentation. Background Art

[0002] As the core raw material for the manufacture of aluminum electrolytic capacitors, the development of electrode foil has been significantly driven by the rapid expansion of the electronics industry, especially the communication products, computers and home appliances markets. At the same time, the demand for miniaturization, high performance and sheet-type aluminum electrolytic capacitors is becoming increasingly urgent, which puts higher requirements on the production technology and quality of electrode foil. Among them, the thickness measurement result of the electrode foil fault has become a key indicator for evaluating its quality.

[0003] The measurement of electrode foil fault thickness mainly relies on image segmentation technology, which processes the electrode foil fault microscopic image taken by a microscope, extracts the fault area, and uses the label information in the image (including equipment parameters, shooting time, pixel scale and its corresponding actual size, etc.) to realize the conversion from pixel size to actual size, thereby determining the average thickness of the electrode foil fault (i.e., the residual core thickness), which is used as the evaluation standard for the quality of the electrode foil product.

[0004] However, at the current stage, due to the diversity of electrode foil types and the interference of environmental factors, its tomographic microscopic images often show the characteristics of insufficient contrast, large background brightness changes and much noise interference. Therefore, the existing electrode foil tomographic thickness measurement method mainly relies on manual adjustment of the global threshold and manual region cropping to segment and extract the tomographic area in the microscopic image, and then measure the thickness based on this. This method not only has high manual operation costs, but is also easily affected by the operator's subjective factors, resulting in inaccurate segmentation results, which may misjudge the product quality of the electrode foil, causing economic losses or even safety hazards. Summary of the invention

[0005] To this end, an embodiment of the present invention provides an electrode foil fault thickness measurement method based on microscopic image segmentation, which is used to solve the problems of high manual operation cost, susceptibility to subjective factors and inaccurate segmentation results in existing electrode foil fault thickness measurement methods.

[0006] In order to solve the above problems, an embodiment of the present invention provides a method for measuring electrode foil fault thickness based on microscopic image segmentation, the method comprising:

[0007] The electrode foil tomographic microscopic image is obtained by a high-resolution microscopic device, and the image is enhanced by grayscale histogram equalization to improve the image contrast;

[0008] Based on the electrode foil tomographic microscopic image after image enhancement, the grayscale feature classification statistics of pixels in the area of ​​the microscopic image except the label are performed line by line to determine the position of the minimum grayscale average value;

[0009] According to the production category of electrode foil (different categories of electrode foil have different sizes of fault areas), the size of the cropping area is determined, and the area is cropped based on the position of the determined minimum grayscale average value;

[0010] Perform adaptive threshold segmentation on the cropped area, and perform morphological opening operation on the segmentation result to remove noise and small area interference to obtain the final segmentation result;

[0011] Perform fixed threshold segmentation on the label area of ​​the microscopic image, extract the pixel ruler area, and calculate the actual size corresponding to a single pixel based on the actual size corresponding to the pixel ruler area;

[0012] The pixel area of ​​the segmented area is counted, the average pixel thickness of the fault area is calculated, and the actual size corresponding to the calculated single pixel is converted into the actual average fault thickness.

[0013] Preferably, the step of performing row-by-row pixel grayscale feature classification statistics on the area of ​​the microscopic image excluding the label to determine the position of the minimum grayscale average value specifically includes:

[0014] First, calculate the grayscale average of each row of pixels: Assume that the height of the tomographic microscopy image is H, the width is W, and the image matrix is ​​I, where I[i,j] represents the grayscale value of the pixel in the i-th row and j-th column, then the grayscale average value of the i-th row is G avg [i] is calculated as:

[0015]

[0016] Then, determine the location i where the minimum grayscale average value is located min , that is, the location of the fault, where the minimum grayscale average value is located i min The calculation formula is:

[0017]

[0018] Preferably, the method of determining the size of the cropping area based on the tomographic microscopic image category, wherein the sizes of the tomographic regions of electrode foils of different categories are different, and performing regional cropping based on the position of the determined minimum grayscale average value as a standard, specifically includes:

[0019] First, the size of the cropping area is determined according to the type of the tomographic microscopic image. The size of the cropping area is set to h×w according to specific needs, where h is the height of the cropping area and w is the width of the cropping area.

[0020] Then, the area is clipped based on the position of the minimum grayscale average value, and the starting position of the clipping area is set to if If it is less than 0, the starting position is (0,0), and the height of the cropping area is adjusted accordingly. The ending position of the cropping area is if If w is greater than H, the end position is (H-1, w-1), and the height of the cropped area is adjusted accordingly.

[0021] Preferably, the adaptive threshold segmentation is performed on the cropped area, and morphological operations are performed on the segmentation results to remove noise and small area interference to obtain the final segmentation result, which specifically includes:

[0022] First, calculate the grayscale mean μ and standard deviation σ of the cropped area:

[0023]

[0024] In the formula, i start and i end They are the starting and ending row indexes of the cropping area respectively;

[0025] Then, the segmentation threshold T is calculated based on the grayscale mean and standard deviation of the cropped area:

[0026] t=μ+k·σ;

[0027] In the formula, k is a custom correction coefficient, which needs to be adjusted according to actual conditions;

[0028] Next, the pixel gray value I[i,j] is compared with the segmentation threshold T. If the pixel gray value is greater than the segmentation threshold, the pixel is marked as foreground, otherwise it is marked as background;

[0029] Finally, the segmentation result is subjected to morphological opening operation to remove noise points and small areas to obtain the final segmentation result.

[0030] Preferably, the fixed threshold segmentation is performed on the label area of ​​the electrode foil tomographic microscopy image, the pixel scale area is extracted, and the actual size corresponding to a single pixel is calculated according to the actual size corresponding to the pixel scale area, which specifically includes:

[0031] First, set a fixed threshold T fixed , perform binary segmentation on the label area of ​​the microscopic image;

[0032] Then, the pixel scale area is extracted based on the area and shape characteristics of the connected domain, and the number of pixels N it contains is calculated. pixel ;

[0033] Finally, according to the actual size L corresponding to the pixel ruler area real , calculate the actual size l corresponding to a single pixel:

[0034]

[0035] Preferably, the pixel area statistics of the segmented area are performed, the average pixel thickness of the fault area is calculated, and the actual size corresponding to the calculated single pixel is converted into the actual average thickness of the fault, which specifically includes:

[0036] First, count the pixel area of ​​the segmented area to get the total number of pixels N fault ;

[0037] Then, assuming that the fault area is uniformly distributed within the cropped area, the average pixel thickness d of the fault area is pixel for:

[0038]

[0039] Finally, the actual size l corresponding to a single pixel is converted into the actual average thickness d real :

[0040]

[0041] It can be seen from the above technical solutions that the present invention has the following beneficial effects:

[0042] (1) High degree of automation: The present invention automatically determines the position of the fault area through hierarchical statistics of pixel grayscale features line by line, without the need for manual positioning, which significantly reduces the cost of manual operation and improves the degree of automation of the measurement process.

[0043] (3) High measurement accuracy: The adaptive threshold segmentation algorithm is used to segment the fault area, and the segmentation results are optimized by combining morphological operations. Compared with manually setting the threshold, the fault area can be identified more accurately, which improves the measurement accuracy. At the same time, the fixed threshold segmentation of the pixel ruler area can accurately calculate the actual size corresponding to a single pixel, thereby ensuring the accurate measurement of the fault thickness.

[0044] (3) Strong adaptability: The method of the present invention has good adaptability to electrode foil tomographic microscopic images with unclear contrast and a large range of background brightness variation, and can effectively meet the measurement requirements under complex image conditions, thereby improving the versatility and practicality of the method. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the implementation cases of the present invention or the technical solutions in the prior art, the following is a brief description of the drawings required for use in the embodiments. By referring to the drawings, the features and advantages of the present invention will be more clearly understood. The drawings are schematic and should not be understood as limiting the present invention in any way. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0046] Figure 1 A flow chart of a method for measuring electrode foil fault thickness based on microscopic image segmentation provided in an embodiment;

[0047] Figure 2 Schematic diagram of segmentation results and corresponding thickness measurement results of tomographic microscopic images of different electrode foils in the embodiment; DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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.

[0049] Embodiment 1

[0050] In order to solve the problems of high manual operation cost, susceptibility to subjective factors and inaccurate segmentation results in the existing electrode foil fault thickness measurement method. Figure 1 As shown, an embodiment of the present invention provides a method for measuring the thickness of an electrode foil based on microscopic image segmentation, the method comprising:

[0051] S1: Obtain the electrode foil cross-sectional microscopic image by high-resolution microscopy equipment, and perform image enhancement by grayscale histogram equalization to improve image contrast;

[0052] S2: Based on the image-enhanced electrode foil tomographic microscopic image, the grayscale feature classification statistics of pixels in the area of ​​the microscopic image except the label are performed row by row to determine the position of the minimum grayscale average value;

[0053] S3: Determine the size of the cropping area according to the production category of the electrode foil (different categories of electrode foil have different sizes of fault areas), and perform regional cropping based on the position of the determined minimum grayscale average value;

[0054] S4: Adaptively threshold segment the cropped area, and perform morphological opening operation on the segmentation result to remove noise and small area interference to obtain the final segmentation result;

[0055] S5: performing fixed threshold segmentation on the label area of ​​the microscopic image, extracting the pixel ruler area, and calculating the actual size corresponding to a single pixel according to the actual size corresponding to the pixel ruler area;

[0056] S6: Count the pixel areas of the segmented regions, calculate the average pixel thickness of the fault region, and convert the calculated actual size corresponding to a single pixel into the actual average fault thickness.

[0057] It can be seen from the above technical scheme that the present invention proposes a method for measuring the thickness of an electrode foil fault based on microscopic image segmentation. The method is based on microscopic image segmentation technology, obtains an electrode foil fault microscopic image through a high-resolution microscopic device, and performs image enhancement through grayscale histogram equalization to improve the image contrast. Subsequently, the method performs row-by-row pixel grayscale feature classification statistics on the area of ​​the microscopic image except the label, automatically determines the fault position, and cuts out the target area accordingly. Then, the adaptive threshold segmentation algorithm and morphological opening operation are applied to accurately segment the cropped area without manually setting the threshold, effectively reducing the operating cost and subjective interference. At the same time, the method also extracts the pixel scale by segmenting the label area with a fixed threshold, calculates the actual size corresponding to a single pixel, and then converts the pixel area statistics of the segmented area into the actual average thickness of the fault. This technical scheme not only significantly improves the work efficiency and measurement accuracy, but also has strong adaptability to images with low contrast and large background brightness changes, ensuring the reliability and stability of the measurement results.

[0058] In step S1, a cross-sectional microscopic image of the electrode foil is obtained by a high-resolution microscopic device, and grayscale histogram equalization is performed on the cross-sectional microscopic image of the electrode foil to enhance image contrast and improve image quality, thereby providing a good basis for subsequent processing.

[0059] In step S2, based on the image-enhanced electrode foil tomographic microscopic image, the grayscale feature classification statistics of the pixels in the area of ​​the microscopic image except the label are performed row by row to determine the position of the minimum grayscale average value, specifically including:

[0060] First, calculate the grayscale average of each row of pixels: Assume that the height of the tomographic microscopy image is H, the width is W, and the image matrix is ​​I, where I[i,j] represents the grayscale value of the pixel in the i-th row and j-th column, then the grayscale average value of the i-th row is G avg [i] is calculated as:

[0061]

[0062] Then, determine the location i where the minimum grayscale average value is located min , that is, the location of the fault, where the minimum grayscale average value is located i minThe calculation formula is:

[0063]

[0064] In step S3, the size of the cropping area is determined according to the microscopic image category (different categories of electrode foil fault area sizes are different), and the area is cropped based on the position of the determined minimum grayscale average value as the standard, specifically including:

[0065] First, the size of the cutting area is determined according to the production category of the electrode foil. The size of the cutting area is set to h×w according to specific needs, where h is the height of the cutting area and w is the width of the cutting area.

[0066] Then, the area is clipped based on the position of the minimum grayscale average value, and the starting position of the clipping area is set to if If it is less than 0, the starting position is (0,0), and the height of the cropping area is adjusted accordingly. The ending position of the cropping area is if If w is greater than H, the end position is (H-1, w-1), and the height of the cropped area is adjusted accordingly.

[0067] This step aims to extract the parts of the image that contain key information for more detailed analysis.

[0068] In step S4, the present invention adopts a local adaptive threshold segmentation algorithm based on the Niblack algorithm to perform adaptive threshold segmentation on the cropped area, that is, the grayscale mean and standard deviation of the cropped area are used as the benchmark to calculate the segmentation threshold, so that the segmentation threshold is adaptively adjusted, which can solve the accuracy problem of fault extraction under the condition of contrast and background brightness changes. Finally, the segmentation result is subjected to morphological operations to obtain the final segmentation result, which specifically includes:

[0069] First, calculate the grayscale mean μ and standard deviation σ of the cropped area:

[0070]

[0071] In the formula, i start and i end The starting and ending row indices of the cropping area, respectively.

[0072] Then, the segmentation threshold T is calculated based on the grayscale mean and standard deviation of the cropped area:

[0073] T = μ + k * σ;

[0074] In the formula, k is a custom correction coefficient, which needs to be adjusted according to actual conditions.

[0075] Next, the pixel grayscale value I[i, j] is compared with the segmentation threshold T. If the pixel grayscale value is greater than the segmentation threshold, the pixel is marked as foreground, otherwise it is marked as background.

[0076] Finally, a morphological opening operation is performed on the segmentation result to remove noise and small areas to obtain the final segmentation result, further improving the accuracy of the segmentation result.

[0077] In step S5, the microscopic image label area is segmented with a fixed threshold value, the pixel scale area is extracted, and the actual size corresponding to a single pixel is calculated according to the actual size corresponding to the pixel scale area;

[0078] First, set a fixed threshold T fixed , perform binary segmentation on the label area of ​​the microscopic image.

[0079] Then, the pixel scale area is extracted based on the area and shape characteristics of the connected domain, and the number of pixels N it contains is calculated. pixel .

[0080] Finally, according to the actual size L corresponding to the pixel ruler area real (expressed in length units), calculate the actual size l corresponding to a single pixel:

[0081]

[0082] It should be noted here that l is the length represented by a single pixel in the actual physical world.

[0083] In step S6, pixel area statistics are performed on the segmented area, the average pixel thickness of the fault area is calculated, and the actual size corresponding to the calculated single pixel is converted into the actual fault average thickness, which specifically includes:

[0084] First, count the pixel area of ​​the segmented area to get the total number of pixels N fault ;

[0085] Then, assuming that the fault area is uniformly distributed within the cropped area, the average pixel thickness d of the fault area is pixel for:

[0086]

[0087] Finally, the actual size l corresponding to a single pixel is converted into the actual average thickness d real :

[0088] d real =d pixel ·l.

[0089] The above steps describe a complete process from acquiring the electrode foil tomographic microscopic image to calculating the actual tomographic average thickness. Figure 2 The segmentation results of different electrode foil tomographic microscopy images and the corresponding thickness measurement results are shown.

[0090] Obviously, the above embodiments are merely examples for clear explanation and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from them are still within the protection scope of the invention.

Claims

1. A method for measuring electrode foil fault thickness based on microscopic image segmentation, characterized in that: include: Based on the enhanced electrode foil tomographic microscopic image, the grayscale feature classification statistics of pixels in the area of ​​the microscopic image except the label are performed line by line to determine the position of the minimum grayscale average value; According to the production category of electrode foil, the size of the fault area of ​​different categories of electrode foil is different, the size of the cropping area is determined, and the area is cropped based on the position of the determined minimum grayscale average value; Perform adaptive threshold segmentation on the cropped area, and perform morphological opening operation on the segmentation result to remove noise and small area interference to obtain the final segmentation result; Perform fixed threshold segmentation on the label area of ​​the microscopic image, extract the pixel ruler area, and calculate the actual size corresponding to a single pixel based on the actual size corresponding to the pixel ruler area; The pixel area of ​​the segmented area is counted, the average pixel thickness of the fault area is calculated, and the actual size corresponding to the calculated single pixel is converted into the actual average fault thickness.

2. The electrode foil fault thickness measurement method based on microscopic image segmentation according to claim 1 is characterized in that: The step of performing pixel grayscale feature classification statistics row by row on the area of ​​the microscopic image except the label to determine the position of the minimum grayscale average value specifically includes: First, calculate the grayscale average of each row of pixels: Assume that the height of the tomographic microscopy image is H, the width is W, and the image matrix is ​​I, where I[i,j] represents the grayscale value of the pixel in the i-th row and j-th column, then the grayscale average value of the i-th row is G avg [i] is calculated as: Then, determine the location i where the minimum grayscale average value is located min , that is, the location of the fault, where the minimum grayscale average value is located i min The calculation formula is:

3. The electrode foil fault thickness measurement method based on microscopic image segmentation according to claim 2, characterized in that: The method is based on the production category of the electrode foil, wherein the sizes of the fault areas of the electrode foils of different categories are different, determining the size of the cropping area, and performing regional cropping based on the position of the determined minimum grayscale average value as the standard, specifically including: First, the size of the cropping region is determined according to the type of the tomographic microscopic image. The size of the cropping region is set to h×w according to specific needs, where j is the height of the cropping region and w is the width of the cropping region. Then, the area is clipped based on the position of the minimum grayscale average value, and the starting position of the clipping area is set to if If it is less than 0, the starting position is (0,0), and the height of the cropping area is adjusted accordingly. The ending position of the cropping area is if If w is greater than H, the end position is (H-1, w-1), and the height of the cropped area is adjusted accordingly.

4. The electrode foil fault thickness measurement method based on microscopic image segmentation according to claim 3 is characterized in that: The adaptive threshold segmentation is performed on the cropped area, and the morphological opening operation is performed on the segmentation result to remove noise and small area interference to obtain the final segmentation result, which specifically includes: First, calculate the grayscale mean μ and standard deviation σ of the cropped area: In the formula, i syart and i end They are the starting and ending row indexes of the cropping area respectively; Then, the segmentation threshold T is calculated based on the grayscale mean and standard deviation of the cropped area: T = μ + k * σ; In the formula, k is a custom correction coefficient, which needs to be adjusted according to actual conditions; Next, the pixel gray value I[i,j] is compared with the segmentation threshold T. If the pixel gray value is greater than the segmentation threshold, the pixel is marked as foreground, otherwise it is marked as background; Finally, the segmentation result is subjected to morphological opening operation to remove noise points and small areas to obtain the final segmentation result.

5. The electrode foil fault thickness measurement method based on microscopic image segmentation according to claim 4 is characterized in that: The method of performing fixed threshold segmentation on the label area of ​​the electrode foil tomographic microscopic image, extracting the pixel scale area, and calculating the actual size corresponding to a single pixel according to the actual size corresponding to the pixel scale area, specifically includes: First, set a fixed threshold T fixed , perform binary segmentation on the label area of ​​the microscopic image; Then, the pixel scale area is extracted based on the area and shape characteristics of the connected domain, and the number of pixels N it contains is calculated. pixel ; Finally, according to the actual size L corresponding to the pixel ruler area real , calculate the actual size l corresponding to a single pixel:

6. The electrode foil fault thickness measurement method based on microscopic image segmentation according to claim 5, characterized in that: The pixel area statistics of the segmented area are performed, the average pixel thickness of the fault area is calculated, and the actual size corresponding to the calculated single pixel is converted into the actual fault average thickness, which specifically includes: First, count the pixel area of ​​the segmented area to get the total number of pixels N fault ; Then, assuming that the fault area is uniformly distributed within the cropped area, the average pixel thickness d of the fault area is pixel for: Finally, the actual size l corresponding to a single pixel is converted into the actual average thickness d real : d real =d pixel ·l。

Images