Method for automatic measurement of pore size of a nuclear pore membrane

Abstract

The application discloses a kind of nuclear pore membrane pore size automatic measurement method, by being plated to sample surface conductive film, to the sample after plating film is carried out electron microscope scanning, to scanning image sequentially do USM sharpening processing, corrosion inflation processing, hough circle detection processing, obtain nuclear pore membrane hole inside and outside diameter and its statistical distribution;The method of the application can detect all the pores of nuclear pore membrane, and the concentric circles, intersecting circles are detected by hough circle gradient detection algorithm and concentric circle processing algorithm, with high accuracy, the error is within 5% with artificial statistics;Using algorithm automatically operates, detection time is short, does not need manual operation, greatly improves detection accuracy;And simple architecture can be transplanted and run, improve application range.

Description

Technical Field

[0001] This invention belongs to the field of nuclear device technology, and specifically relates to an automated method for measuring the pore size of nuclear pore membranes. Background Technology

[0002] Nuclear pore membranes, also known as nuclear track-etched membranes, utilize the penetration of heavy ions into organic polymer plastic films, leaving a narrow irradiation damage channel along the path of the heavy ions. This channel, after sensitization, can be etched using appropriate chemical reagents to transform the damage channel into regular cylindrical or conical micropores. By controlling the irradiation and etching conditions of nuclear pores, nuclear pore membranes with different pore densities and pore sizes can be obtained. Precise control of the pore size and subsequent nuclear pore membrane testing are crucial steps in ensuring the quality of nuclear pore membranes. Nuclear pore sizes range from a few tenths of a micrometer to tens of micrometers, and conventional optical microscopy methods cannot accurately and quickly measure their pore size.

[0003] Currently, traditional digital image processing is used to measure the pore size of nuclear pore membranes. An automatic nuclear pore membrane detection method based on super-resolution restored images and circular pore edge criteria is proposed. This method utilizes super-resolution restored image restoration theory to further clarify the pore membrane images obtained from scanning electron microscopy. However, it cannot accurately determine intersecting pores at the edges of circular pores, and the binarization method used during image acquisition cannot guarantee the complete capture of image details.

[0004] Another method for detecting nuclear pore membranes using mathematical morphology is proposed. A binarization method based on grayscale mathematical morphology is constructed, and a layered extraction method based on binary mathematical morphology is proposed to count the total number of pores and their radii. However, while the binarization method effectively preserves the pore membrane image, it loses many details from the original grayscale image, resulting in the accuracy and information breadth of pore size detection failing to meet the required standards, necessitating a certain degree of manual intervention. Summary of the Invention

[0005] In view of the above-mentioned technical problems existing in the prior art, the purpose of the present invention is to provide an automatic detection method for nuclear pore membranes that ensures accurate detection of intersecting and overlapping pores with high accuracy.

[0006] To achieve the above-mentioned objectives, the present invention adopts the following technical solution: an automated method for measuring the pore size of nuclear pore membranes, comprising the following steps:

[0007] (1) Sputter a conductive metal thin film onto the sample surface;

[0008] (2) The sample after coating was scanned by electron microscopy to obtain the original image;

[0009] (3) The original images above were processed sequentially by USM sharpening, erosion and expansion processing, and Hough circle detection to obtain the number and radius distribution of sample holes;

[0010] (4) Statistical analysis was performed on the number and radius distribution of the above sample pores to obtain accurate nuclear pore membrane pore size values.

[0011] Furthermore, in step (1), the sample surface is coated by sputtering at a magnification of 30K or less; the film thickness is proportional to the sputtering coefficient, sputtering current, and sputtering time.

[0012] Further, in step (3), the USM sharpening process includes applying Gaussian blur to the original image, and then subtracting a coefficient from the original image multiplied by the Gaussian blurred image to obtain the USM sharpened image.

[0013] Furthermore, in step (3), the erosion and dilation processing employs a closing operation on the USM sharpening image processing.

[0014] Further, in step (3), the Hough circle detection uses the gradient Hough algorithm to detect concentric circles. The first circle detection is performed on the image after erosion and dilation processing using the gradient Hough algorithm. After a circle is detected, the center (x1, y1) and radius r1 are recorded. This circle is then destroyed to segment the outer diameter of the hole. The gradient Hough transformation algorithm is then used again for the second gradient Hough transformation algorithm. The center (x2, y2) and radius r2 are recorded again. All circles and their radii are detected in sequence.

[0015] The beneficial effects of adopting the technical solution of this invention are as follows: This invention provides an automated method for measuring the pore size of nuclear pore membranes. By depositing a conductive thin film on the sample surface, scanning the coated sample with an electron microscope, and sequentially performing USM sharpening, corrosion expansion processing, and Hough circle detection processing on the scanned image, the inner and outer diameters of the nuclear pore membrane are obtained. The method of this invention can detect all pores of nuclear pore membranes, including concentric circles and intersecting circles. It has a short detection time, high efficiency, and uses an algorithm to automatically perform calculations without manual operation, which greatly improves the detection accuracy. Moreover, the architecture is simple, portable, and expands the application range. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of an automated method for measuring the pore size of nuclear pore membranes according to an embodiment of the present invention;

[0017] Figure 2 This is a statistical distribution diagram of the inner diameter of the nuclear pore membrane detected by the method of this embodiment of the invention;

[0018] Figure 3 This is a statistical distribution diagram of the outer diameter of nuclear pore membranes detected by the method of this embodiment of the invention;

[0019] Figure 4This is a schematic diagram of the true inner and outer diameter distribution of nuclear pore membrane pores detected by the method of this embodiment of the invention. Detailed Implementation

[0020] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] See attached document Figure 1 This invention provides an automated method for measuring the pore size of nuclear pore membranes, comprising the following steps:

[0022] (1) Spraying a conductive metal film onto the sample surface; When the nuclear pore membrane, as a non-metallic sample, is irradiated by an electron beam, the excess electrons do not have a conductive path, which will cause the charge to accumulate continuously and the charging phenomenon will affect the trajectory of the incident electrons, resulting in black and white stripes or distortion and irregular motion in the image; Spraying a layer of conductive metal film onto the sample surface can not only increase the conductivity of the sample and reduce the charging phenomenon, but also protect the sample, reduce radiation damage to the sample, and improve the contrast of the secondary electron image.

[0023] (2) The sample after coating was scanned by electron microscopy to obtain the original image;

[0024] (3) The original images above were processed sequentially by USM sharpening, erosion and expansion processing, and Hough circle detection to obtain the number and radius distribution of sample holes;

[0025] (4) Statistical analysis was performed on the number and radius distribution of the above sample pores to obtain accurate nuclear pore membrane pore size values.

[0026] Preferably, in step (1), the sample surface is coated by sputtering at a magnification of 30K or less; the film thickness is proportional to the sputtering coefficient, sputtering current, and sputtering time. An excessively thick conductive film will affect the imaging under the scanning electron microscope, while an excessively thin film will prevent it from providing conductive and thermal protection. Before each coating operation, a thinner coating thickness is determined using a smaller current and a shorter sputtering time, and then a thicker coating thickness is determined using a larger current and a longer sputtering time. Simultaneously, the imaging effect under the scanning electron microscope is observed, and the sputtering time or sputtering current is dynamically adjusted based on the imaging effect.

[0027] Preferably, in step (1), an Au film is deposited on the sample surface by sputtering at a magnification of 30K or less; the air sputtering coefficient is 0.07, and the argon sputtering coefficient is 0.17. The film layer is uniform and has no obvious features.

[0028] Preferably, in step (3), the original image is converted into a grayscale image; then USM sharpening is performed to create a two-bit Gaussian low-pass filter. By adjusting the blur radius and sigma filtering parameters, a Gaussian blurred image is obtained. Numerical weights from 0 to 1 are introduced, and the processed edge image is obtained by subtracting the original image from the Gaussian filtered image.

[0029] The USM sharpening process includes generating an unsharpened mask and generating a sharpened image; the unsharpened mask is generated using formula (1).

[0030]

[0031] Where I(m,n) represents the original image, and H is a high-pass filter. This represents the convolution operation;

[0032] To reduce the noise generated during the sharpening process, a Gaussian low-pass filter is used instead of a high-pass filter during mask generation. Therefore, the generation of the unsharpened mask is expressed by formula (2).

[0033]

[0034] Among them, G σ This represents a Gaussian low-pass filter with a standard deviation of σ.

[0035] The sharpened image is generated using formula (3).

[0036] I sp (m,n)=I(m,n)+λM G (3)

[0037] Where λ represents the sharpening intensity, which controls the size of the overshoot artifact during the sharpening process; the USM sharpened image is obtained according to formulas (2) and (3), and is expressed by formula (4):

[0038]

[0039] Among them, I sp (m,n) represents the edge image after USM sharpening.

[0040] Preferably, in step (3), the corrosion expansion uses a closed-loop structural element S to perform an expansion operation on the target hole, and then uses S to perform an corrosion operation on the expanded result to obtain the processed hole image. Erosion can eliminate edge points, causing the edges to shrink inward, eliminating object edges, removing objects smaller than the structural element, and breaking the thin connections between two objects; expansion merges the points around the object into the object. When the distance between two objects is relatively close, the expansion element may connect the two objects together. The expansion element is used to fill the voids between objects or between two objects.

[0041] Etching can separate adhered target holes in a scanning electron microscope, while dilation can restore broken target holes. However, after etching, the area of ​​the target hole is smaller than its original area, while after dilation, the target hole is larger than its original area. This embodiment of the invention employs a closing operation. First, a dilation operation is performed on the target hole using a structuring element S, and then an erosion operation is performed on the dilated result using S. This embodiment of the invention uses erosion and dilation to process the image, using a square mask for erosion and dilation. By adjusting the mask radius, the number of erosions, the number of dilations, and the mask radius, an image is obtained that fills small holes within the target, connects adjacent objects, and smooths its edges without significantly changing its area.

[0042] Preferably, in step (3), Hough circle detection uses global image features to connect edge pixels to form a closed boundary of the region, transforms the image space into a parameter space, describes the points in the parameter space, and achieves the purpose of detecting image edges; statistical calculations are performed on all points that may fall on the edge, and the degree to which they belong to the edge is determined based on the statistical results of the data.

[0043] The general equation of a circle is given by: (xa) 2 +(yb) 2 =r 2

[0044] Where (a,b) is the center of the circle, and r is the radius of the circle; in the image plane, the feature point (x,y) is an unknown quantity, and (a,b) is treated as a known quantity; the coordinates of the feature points in the image are scanned and recorded; the circle on the XY plane is transformed into the abr parameter space, and the equation of the circle in the parameter space becomes as follows:

[0045] (ax) 2 +(by) 2 =r 2

[0046] In the parameter space, the detected coordinates (x, y) are known quantities, while the parameters (a, b) and r are treated as unknown quantities. Thus, each feature point (x, y) constitutes a cone in the parameter space abr. Points on the same circle in image space correspond to all three-dimensional cones in the parameter plane, which must intersect at a single point. While there may be multiple intersection points between different cones, all cones must intersect at a specific point. This point, with the highest number of intersections, is the center of the circle. By detecting this point, the parameters of the circle are obtained, and its coordinates (a, b, r) represent the center and radius coordinates of the circle.

[0047] In practical Hough circle detection, feature points are mapped to parameter space, and then the parameter space is discretized. A three-dimensional accumulator (a,b,r) is used to store the positions of each point in the parameter space. When a point appears repeatedly at a certain position in the parameter space, the accumulator increments by 1 at that position. A threshold is set. When the accumulated value at a certain position exceeds this threshold, that point is taken as the center of the circle, thus obtaining the center of the pore and its radius on the nuclear pore membrane.

[0048] When concentric circles or intersecting circles appear, the exact position of the circle's center and its radius cannot be obtained. This embodiment of the invention uses the gradient Hough algorithm to detect concentric circles. The image after erosion and dilation is processed using the gradient Hough algorithm for the first circle detection. After detecting a circle, the center (x1, y1) and radius r1 are recorded. This circle is then disrupted, and the outer diameter of the hole is segmented. The gradient Hough transform algorithm is then used for the second gradient transform algorithm, and the center (x2, y2) and radius r2 are recorded again. This process is repeated to detect all circles (x1, y1, r2 ... n ,y n ), and its radius r n .

[0049] In this embodiment of the invention, a value assignment method is used to detect the outer diameter of a circle in the event of damage. The inner circle area obtained in the first step is assigned a value. By observation, it can be seen that the inside of the hole is usually dark, while the outer diameter is light. Therefore, when the inner diameter area is selected, the inner diameter circle center is used as the center, and the inner diameter circle radius r±X (X is the radius adjustment value) is assigned a gray value. The X is adjusted according to the assignment effect to ensure that the drawn circle is close to the outer diameter, thereby identifying and detecting the outer diameter of the circle.

[0050] See attached document Figure 2 , 3 4. The statistical distribution of the inner and outer diameters of the nuclear pore membrane detected by the method of the present invention, and the actual diameter size is calculated based on the pixel size of the actual measured scale, which is 97.51, i.e. 2μm.

[0051] The total number of pores per unit area of ​​the nuclear pore membrane detected by the method of this embodiment of the invention is compared with the total number of pores per unit area of ​​the nuclear pore membrane counted manually. The results are shown in Table 1.

[0052] Table 1. Comparison of Total Pore Count Detection Results

[0053]

[0054] The method of this invention has an error of less than 5% compared with the traditional manual statistical method, indicating that the automatic detection method of nuclear pore membrane pore size of this invention is accurate and can replace manual operation for detecting nuclear pore membrane pore size.

[0055] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention is also intended to include these modifications and variations.

Claims

1. A method for automated measurement of the pore size of a nuclear pore membrane, characterized by, The method comprises the following steps: (1) spraying a conductive metal film on the surface of the sample; (2) scanning the sample after the film is sprayed by an electron microscope to obtain an original picture; (3) sequentially performing USM sharpening processing, corrosion dilation processing and Hough circle detection processing on the original picture to obtain the number and radius distribution of sample holes; The USM sharpening processing comprises performing Gaussian blurring on the original picture, and then subtracting an image after a Gaussian blurring picture is multiplied by a coefficient from the original image to obtain a USM sharpening processing picture; The corrosion dilation processing adopts a closed operation to process the USM sharpening processing picture; The Hough circle detection process adopts gradient Hough algorithm to detect concentric circles, and the first circle detection is performed on the image after the corrosion expansion processing using the gradient Hough algorithm, a circle is detected, and the center (x1, y1) and the radius r1 are recorded; the circle is destroyed, the outer diameter of the hole is segmented, the second gradient conversion algorithm is performed using the gradient Hough conversion algorithm again, the center (x2, y2) and the radius r2 are recorded again, and all the circles (x n ,y n ) and the radii thereof are detected in turn; (4) performing statistical analysis on the obtained number and radius distribution of sample holes to obtain an accurate pore size value of the nuclear pore membrane.

2. The method of claim 1, wherein the method is characterized by, In the step (1), the sample surface is plated by a sputtering plating method; the film thickness is proportional to the sputtering coefficient, the sputtering current and the sputtering time.

3. The method for automatically measuring the pore size of a nuclear pore membrane according to claim 1 or 2, wherein the method is characterized by, In the step (1), the sample surface is plated by a sputtering plating method under a magnification of 30K or less; the air sputtering coefficient is 0.07, and the argon sputtering coefficient is 0.

17.

4. The method of claim 1, wherein the method is characterized by, In the step (3), the original image is converted into a gray-scale image; then, USM sharpening is performed, a two-bit Gaussian low-pass filter is created, a Gaussian blurred image is obtained by adjusting the blur radius and sigma filter parameters, a numerical weight of 0 to 1 is introduced, the original image and the Gaussian filtered image are subtracted to obtain a processed edge image.

5. The method of claim 4, wherein the step of measuring the pore size of the nanopore membrane is performed automatically. In the step (3), the edge image of the USM sharpening processing is expressed by a formula (4), where I sp (m, n) represents the edge image after USM sharpening processing; I(m, n) represents the original picture, H is a high-pass filter, represents convolution operation, G σ represents a Gaussian low-pass filter with a standard deviation of σ, and λ represents the sharpening intensity.

Images