Automatic alignment method and device for hollow-core optical fiber

Abstract

The invention relates to the technical field of hollow-core optical fiber alignment, in particular to a hollow-core optical fiber automatic alignment method and device, and the method comprises the steps: obtaining a first end face image of a first hollow-core optical fiber and a second end face image of a second hollow-core optical fiber; filtering the first end face image to obtain a first filtered image, and filtering the second end face image to obtain a second filtered image; the first filtering image is rotated angle by angle, and the rotated first filtering image and the second filtering image are subjected to cross-correlation calculation until the optimal alignment angle corresponding to the maximum value of the cross correlation coefficient is obtained; rotating the first hollow-core optical fiber to an optimal alignment angle so as to align the first hollow-core optical fiber with the second hollow-core optical fiber; according to the scheme, based on the end face image of the hollow-core optical fiber, the end face information of the hollow-core optical fiber can be efficiently extracted, the image processing process is simplified, angle matching is optimized through a self-adaptive rotation adjustment algorithm, and fiber cores and anti-resonance structures of the two hollow-core optical fibers can be perfectly aligned.

Description

Technical Field

[0001] The present invention relates to the technical field of hollow-core optical fiber alignment, and in particular to a hollow-core optical fiber automatic alignment method and device. Background Art

[0002] Due to its unique physical properties, hollow-core optical fiber has demonstrated broad application prospects in cutting-edge technologies such as optical communications, high-power laser transmission, nonlinear optics, and biomedical sensing. Unlike traditional solid-core optical fibers, hollow-core optical fibers have a core composed of microstructured channels filled with air or other gases, designed to minimize the absorption and scattering of optical signals by the material, thereby achieving ultra-low loss and high-speed data transmission. This design not only significantly reduces nonlinear effects and dispersion, but also increases its high insensitivity to environmental conditions, enabling it to maintain excellent performance under extreme temperature, humidity, and pressure conditions. Therefore, hollow-core optical fibers offer irreplaceable advantages in applications requiring high performance, stability, and reliability.

[0003] In this context, achieving high-quality splicing of hollow-core fibers is particularly critical and directly related to the actual performance and reliability of these applications. Currently, the most important type of hollow-core fiber is antiresonant hollow-core fiber. The antiresonant cavity structure inside it further reduces the loss and distortion of the optical signal by precisely controlling the propagation path of light. However, the alignment of the antiresonant cavity is extremely important during the splicing process. Any slight deviation will cause the optical signal to leak or be lost between the fiber cores, causing crosstalk or signal attenuation, seriously affecting the quality and efficiency of data transmission. In order to ensure that each channel can effectively transmit signals, precise alignment between the antiresonant cavities must be ensured during splicing. Any slight deviation may cause reflection, refraction or scattering of the optical signal, thereby increasing insertion loss and return loss.

[0004] In the prior art, the end face of the hollow core optical fiber is extracted in a conventional manner and then rotated and docked, and the part beyond the boundary is automatically reset to zero, such as Figure 1 As shown in the figure, the horizontal axis is the rotation angle and the vertical axis is the correlation coefficient. The angle corresponding to the correlation peak indicates that the right-end image must be rotated by the corresponding angle to achieve alignment with the left-end image. However, since the portion of the rotated image that exceeds the original image range is padded with zeros by default, a black background appears at the four corners of the image when the rotation angle is not a multiple of 90 degrees. Since the correlation is calculated on the pixels of the entire image, this black background is included in the correlation coefficient calculation, resulting in the suppression of correlation peaks at angles that are not integer multiples of 90°. After rotations of 90°, 180°, 270°, and 360°, the images are completely free of zeros. Regardless of the left and right images, the correlation peaks are located near these four angles. However, in reality, there are five antiresonant rings inside the hollow-core fiber, so there should be five more prominent correlation peaks, not four, which will seriously affect the accuracy of subsequent hollow-core fiber alignment.

[0005] In view of this, overcoming the defects of the prior art is an urgent problem to be solved in this technical field. Summary of the Invention

[0006] The technical problem to be solved by the present invention is how to provide an automatic alignment method between hollow-core optical fibers.

[0007] The present invention adopts the following technical solutions:

[0008] In a first aspect, a hollow-core optical fiber automatic alignment method is provided, comprising:

[0009] Acquire a first end face image of the first hollow core optical fiber and a second end face image of the second hollow core optical fiber;

[0010] filtering the first end face image to obtain a first filtered image, and filtering the second end face image to obtain a second filtered image;

[0011] Rotating the first filtered image angle by angle, and performing cross-correlation calculation on the rotated first filtered image and the second filtered image respectively, until an optimal alignment angle corresponding to a maximum cross-correlation coefficient is obtained;

[0012] The first hollow-core optical fiber is rotated to the optimal alignment angle to align the first hollow-core optical fiber with the second hollow-core optical fiber.

[0013] Preferably, acquiring a first end face image of the first hollow-core optical fiber and a second end face image of the second hollow-core optical fiber specifically includes:

[0014] Stripping the coating layers of the fusion-spliced ends of the first hollow-core optical fiber and the second hollow-core optical fiber, and neatly cutting the fusion-spliced end faces of the first hollow-core optical fiber and the second hollow-core optical fiber, respectively placing the first hollow-core optical fiber and the second hollow-core optical fiber into the clamps at both ends of the fusion splicer and fixing them;

[0015] Adjusting the positions of the fusion-spliced end faces of the first hollow-core optical fiber and the second hollow-core optical fiber, and obtaining a first intermediate image corresponding to the first hollow-core optical fiber and a second intermediate image corresponding to the second hollow-core optical fiber through end face imaging of a fusion splicer;

[0016] performing grayscale conversion on the first intermediate image and the second intermediate image and adjusting them to the same size to obtain a first grayscale image and a second grayscale image, respectively;

[0017] The cladding information in the first grayscale image and the second grayscale image is removed to obtain the first end face image and the second end face image, respectively.

[0018] Preferably, performing grayscale conversion on the first intermediate image and the second intermediate image and adjusting them to the same size to obtain a first grayscale image and a second grayscale image, respectively, specifically includes:

[0019] Finding a circular edge of the first intermediate image, and obtaining a first center coordinate and a first fiber pixel size of the first hollow-core optical fiber;

[0020] Finding a circular edge of the second intermediate image, and obtaining a second center coordinate and a second fiber pixel size of the second hollow-core optical fiber;

[0021] Cut out a first square image with the first center coordinate at the center of the image according to the first center coordinate and the first optical fiber pixel size;

[0022] Cut out a second square image with the second center coordinate at the center of the image according to the second center coordinate and the second optical fiber pixel size;

[0023] The sizes of the first square image and the second square image are adjusted to be consistent, so as to obtain the first grayscale image and the second grayscale image.

[0024] Preferably, removing the cladding information from the first grayscale image and the second grayscale image to obtain the first end face image and the second end face image, respectively, specifically includes:

[0025] Creating a circular mask; wherein the size of the circular mask is consistent with the first grayscale image and the second grayscale image, the center of the circular mask is at the center of the image, and the radius of the circular mask is consistent with the first fiber pixel size and the second fiber pixel size;

[0026] Perform dot multiplication of the circular mask with the first grayscale image and the second grayscale image to obtain the first end surface image and the second end surface image respectively.

[0027] Preferably, the filtering process of the first end face image and the second end face image is the same, and filtering the first end face image to obtain the first filtered image specifically includes:

[0028] Creating a two-dimensional matrix as a convolution kernel, using the convolution kernel as a sliding window, and moving the sliding window pixel by pixel on the first end-face image; wherein, during each movement, the center of the sliding window is aligned with a pixel point in the first end-face image;

[0029] When the sliding window is aligned with a pixel point, each element in the convolution kernel is multiplied by the corresponding pixel value in the area covered by the sliding window on the first end face image, and all multiplication results are added together to obtain the updated value of the corresponding pixel point after filtering, until the update of all pixel points in the first end face image is completed to obtain the first filtered image.

[0030] Preferably, when processing pixel points at the boundary of the first end face image, zero padding is performed outside the boundary of the first end face image, or the boundary pixel points are copied according to preset requirements so that the sliding window can also perform a complete convolution calculation when it moves to the boundary pixel points.

[0031] Preferably, rotating the first filtered image angle by angle and performing cross-correlation calculation on the rotated first filtered image and the second filtered image respectively until obtaining the optimal alignment angle corresponding to the maximum cross-correlation coefficient specifically includes:

[0032] Setting a rotation angle range and a rotation angle step, and rotating the first filtered image angle by angle according to the rotation angle range and the rotation angle step;

[0033] Cross-correlation calculation is performed on the rotated plurality of first filtered images and the second filtered image to obtain a cross-correlation coefficient, and an angle corresponding to a maximum cross-correlation coefficient is selected as an optimal alignment angle.

[0034] Preferably, the calculation formula of the mutual correlation coefficient is:

[0035]

[0036] Wherein, R is the cross-correlation coefficient, M and N are the number of rows and columns of the image matrix A(x, y) of the first filtered image and the image matrix B(x, y) of the second filtered image, respectively. is the average value of A, is the average value of B, x is the horizontal coordinate, and y is the vertical coordinate.

[0037] In a second aspect, a hollow-core optical fiber automatic alignment device is provided, the hollow-core optical fiber automatic alignment device comprising: a processor and a memory for storing instructions executable by the processor;

[0038] Wherein, the processor is configured to execute the hollow-core optical fiber automatic alignment method described in the first aspect.

[0039] In a third aspect, a non-volatile computer storage medium is provided, wherein the computer storage medium stores computer-executable instructions, and the computer-executable instructions are executed by one or more processors to complete the hollow-core optical fiber automatic alignment method described in the first aspect.

[0040] In a fourth aspect, a chip is provided, comprising: a processor and an interface, for calling and running a computer program stored in a memory to execute the hollow-core optical fiber automatic alignment method as described in the first aspect.

[0041] In a fifth aspect, a computer program product comprising instructions is provided, which, when executed on a computer or a processor, enables the computer or the processor to execute the hollow-core optical fiber automatic alignment method as described in the first to fourth aspects and any one of the above.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] The present invention obtains a first end face image of a first hollow-core optical fiber and a second end face image of a second hollow-core optical fiber; and filters the first end face image to obtain a first filtered image, and filters the second end face image to obtain a second filtered image; then rotates the first filtered image angle by angle, and performs cross-correlation calculation on the rotated first filtered image and the second filtered image respectively, until an optimal alignment angle corresponding to the maximum value of the cross-correlation coefficient is obtained; finally, the first hollow-core optical fiber is rotated to the optimal alignment angle to align the first hollow-core optical fiber with the second hollow-core optical fiber; the automatic alignment scheme proposed by the present invention is based on the end face image of the hollow-core optical fiber, can efficiently extract the end face information of the hollow-core optical fiber, simplifies the image processing process, optimizes the angle matching through an adaptive rotation adjustment algorithm, so that the cores and anti-resonance structures of the two hollow-core optical fibers can be perfectly aligned, and can achieve precise automatic alignment while maintaining high-speed processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0045] Figure 1 This is a schematic diagram of alignment of a hollow-core optical fiber in the prior art provided by an embodiment of the present invention;

[0046] Figure 2 1 is a flow chart of a hollow-core optical fiber automatic alignment method provided by an embodiment of the present invention;

[0047] Figure 3 1 is a schematic diagram of a process for obtaining a first end face image and a second end face image provided by an embodiment of the present invention;

[0048] Figure 4is a schematic diagram of a first intermediate image provided by an embodiment of the present invention;

[0049] Figure 5 is a schematic diagram of a second intermediate image provided by an embodiment of the present invention;

[0050] Figure 6 This is a schematic diagram of a process for obtaining a first grayscale image and a second end face image provided by an embodiment of the present invention;

[0051] Figure 7 is another schematic diagram of a first intermediate image provided by an embodiment of the present invention;

[0052] Figure 8 is another schematic diagram of a second intermediate image provided by an embodiment of the present invention;

[0053] Figure 9 is a schematic diagram of a first square image provided by an embodiment of the present invention;

[0054] Figure 10 is a schematic diagram of a second square image provided by an embodiment of the present invention;

[0055] Figure 11 is a schematic diagram of a first grayscale image provided by an embodiment of the present invention;

[0056] Figure 12 is a schematic diagram of a second grayscale image provided by an embodiment of the present invention;

[0057] Figure 13 is another schematic diagram of a process for obtaining a first end face image and a second end face image provided by an embodiment of the present invention;

[0058] Figure 14 is a schematic diagram of a circular mask provided by an embodiment of the present invention;

[0059] Figure 15 is a schematic diagram of a first end face image provided by an embodiment of the present invention;

[0060] Figure 16 is a schematic diagram of a second end face image provided by an embodiment of the present invention;

[0061] Figure 17 is a schematic diagram of a process for obtaining a first filtered image provided by an embodiment of the present invention;

[0062] Figure 18 is a schematic diagram of a first filtered image provided by an embodiment of the present invention;

[0063] Figure 19 is a schematic diagram of a second filtered image provided by an embodiment of the present invention;

[0064] Figure 20 Schematic diagram of alignment results of a hollow-core optical fiber provided by an embodiment of the present invention;

[0065] Figure 21 It is a structural schematic diagram of a hollow-core optical fiber automatic alignment device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0066] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0067] Unless the context requires otherwise, throughout the specification and claims, the term "including" is to be interpreted as meaning open inclusion, that is, "including, but not limited to". In the description of the specification, the terms "one embodiment", "some embodiments", "exemplary embodiments", "example", "specific example" or "some examples" and the like are intended to indicate that the specific features, structures, materials or characteristics associated with the embodiment or example are included in at least one embodiment or example of the present disclosure. The schematic representation of the above terms does not necessarily refer to the same embodiment or example. In addition, the specific features, structures, materials or characteristics may be included in any one or more embodiments or examples in any appropriate manner, that is, although they may be carried in the embodiments or examples of the above terms due to reasons such as the order and position of appearance, it is not limited to that they can be carried in combination by one embodiment or example.

[0068] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present disclosure, unless otherwise specified, "multiple" means two or more. In addition, for example, the description may also use the method of adding "A" and "B" at the end to describe the same type of nouns as two independent individuals. In this case, the corresponding features defined as "A" and "B" are only used to distinguish the description purposes of the same type of individuals, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated.

[0069] When describing some embodiments, the expressions “coupled”, “coupled” and “connected” and their derivatives may be used. For example, when describing some embodiments, the term “connected” may be used to indicate that two or more components are in direct physical or electrical contact with each other. For another example, when describing some embodiments, the term “coupled” may be used to indicate that two or more components are in direct physical or electrical contact. However, the term “connected” or “coupled” may also mean that two or more components are not in direct contact with each other, but still cooperate or interact with each other, such as “optical coupling”, “wireless connection”, etc. The embodiments disclosed herein are not necessarily limited to the contents of the present invention.

[0070] In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0071] Example 1:

[0072] In order to improve the docking accuracy between hollow-core fibers, in one embodiment, Figure 2 As described above, this embodiment proposes a hollow-core optical fiber automatic alignment method, comprising:

[0073] Step 101: Acquire a first end face image of a first hollow-core optical fiber and a second end face image of a second hollow-core optical fiber.

[0074] The first end-face image is an image of the surface of the first hollow-core fiber to be fused, and the second end-face image is an image of the surface of the second hollow-core fiber to be fused. The two hollow-core fibers are fused through the surface to be fused. It is worth noting that both the first end-face image and the second end-face image are obtained through image processing. The specific image processing process will be described in detail below.

[0075] Step 102: Filter the first end face image to obtain a first filtered image, and filter the second end face image to obtain a second filtered image.

[0076] The purpose of filtering the first and second end-face images is to remove noise from the images and improve image quality and clarity. Noise may arise from factors such as the imaging device and environmental interference. The specific filtering method is described below.

[0077] Step 103: Rotate the first filtered image angle by angle, and perform cross-correlation calculation on the rotated first filtered image and the second filtered image respectively, until an optimal alignment angle corresponding to a maximum cross-correlation coefficient is obtained.

[0078] Among them, angle-by-angle rotation means first determining the rotation angle range (from 0° to 360°), starting from the minimum angle in the rotation angle range, and rotating by a preset angle each time until the rotation angle reaches the maximum angle in the rotation angle range. In one embodiment, the preset rotation angle range can be 0.1° to 0.5°. That is, the preset rotation angle can be 0.1° each time, 0.5° each time, or any angle between 0.1° and 0.5° each time. The smaller the preset rotation angle, the higher the accuracy of the result obtained later.

[0079] For the first filtered image, the image obtained after each rotation by a preset angle is cross-correlated with the second filtered image. Each calculation yields a cross-correlation coefficient, and the rotation angle corresponding to the maximum cross-correlation coefficient is determined as the optimal alignment angle. In this embodiment, the second filtered image is fixed while the first filtered image is rotated. In other embodiments, the first filtered image can be kept stationary while the second filtered image is rotated angle by angle. The calculation method is the same as described above and will not be further described here. The specific steps for the cross-correlation calculation are described below.

[0080] Step 104: Rotate the first hollow-core optical fiber to the optimal alignment angle to align the first hollow-core optical fiber with the second hollow-core optical fiber.

[0081] In one embodiment, after obtaining the optimal alignment angle through the above steps, the optimal alignment angle is input into a fusion splicer, and the fusion splicer controls the first hollow-core optical fiber to rotate to the optimal alignment angle to align the first hollow-core optical fiber with the second hollow-core optical fiber, and then performs a fusion splicing operation.

[0082] In one embodiment, Figure 3 As shown, obtaining a first end face image of the first hollow-core optical fiber and a second end face image of the second hollow-core optical fiber specifically includes:

[0083] Step 1011: Strip the coating layer of the fusion splicing end of the first hollow-core optical fiber and the second hollow-core optical fiber, cut the fusion splicing end faces of the first hollow-core optical fiber and the second hollow-core optical fiber neatly, and place the first hollow-core optical fiber and the second hollow-core optical fiber into the clamps at both ends of the fusion splicer respectively and fix them.

[0084] First, remove the coating from the spliced ends of the first and second hollow-core fibers. The coating protects the fibers, but it can affect splice quality during fusion, so removal is necessary. Then, cleanly cleave the spliced ends. This helps improve splice accuracy and quality and reduces optical signal loss at the splice. For example, use a specialized fiber cleaver to ensure a smooth, burr-free end face.

[0085] Step 1012: Adjust the positions of the fusion end faces of the first hollow-core fiber and the second hollow-core fiber, and obtain a first intermediate image corresponding to the first hollow-core fiber and a second intermediate image corresponding to the second hollow-core fiber through end face imaging of a fusion splicer.

[0086] Place the processed first and second hollow-core fibers into the fixtures at each end of the fusion splicer. The fixtures secure the fibers to prevent movement during the splicing process. Ensure the fibers are positioned accurately and securely in the fixtures to prevent shifting during subsequent operations, which could affect imaging and splice quality.

[0087] like Figure 4 and Figure 5 As shown, the end-face imaging function of the fusion splicer is used to obtain a first intermediate image of the first hollow-core fiber and a second intermediate image of the second hollow-core fiber. End-face imaging typically uses an optical system and an image sensor to capture the fiber end face to obtain a clear image.

[0088] Step 1013: Perform grayscale conversion on the first intermediate image and the second intermediate image and adjust them to the same size to obtain a first grayscale image and a second grayscale image respectively.

[0089] The purpose of grayscale conversion is to simplify image information and reduce the computational effort required for subsequent processing. Grayscale conversion can be achieved by calculating the color value of each pixel in a color image to obtain a grayscale value. For example, a weighted average method can be used to calculate the grayscale value by taking the weighted average of the red, green, and blue channels based on the human eye's sensitivity to different colors.

[0090] In one embodiment, the grayscale conversion process includes the following: For each pixel in a color image (a first intermediate image and a second intermediate image), its color information is typically represented by the values of three channels: red (R), green (G), and blue (B). During grayscale conversion, the values of these three channels are assigned different weights and then added together to obtain the grayscale value of the pixel. Commonly used weights are 0.299, 0.587, and 0.114, and the calculation formula is Gray = 0.299R + 0.587G + 0.114B. For example, for a pixel with R = 100, G = 120, and B = 80, its grayscale value is Gray = 0.299x100 + 0.587x120 + 0.114x80 ≈ 107.4. This weighting is based on the human eye's varying sensitivity to different colors: the human eye is most sensitive to green, followed by red, and least sensitive to blue.

[0091] The first and second intermediate images are grayscale converted according to the above method. After obtaining the grayscale images, subsequent image processing and analysis requires comparison and calculation between the two images. If the images are of different sizes, this increases the difficulty and complexity of the processing. In one embodiment, the first and second grayscale images are resized to the same size using the following method.

[0092] In one embodiment, Figure 6 As shown, the grayscale conversion of the first intermediate image and the second intermediate image and adjusting them to the same size to obtain a first grayscale image and a second grayscale image respectively includes:

[0093] Step 10131: Find the circular edge of the first intermediate image, obtain the first center coordinates and the first fiber pixel size of the first hollow-core fiber; find the circular edge of the second intermediate image, obtain the second center coordinates and the second fiber pixel size of the second hollow-core fiber.

[0094] Among them, such as Figure 7 and Figure 8 As shown in the figure, since the end face of a hollow core fiber is usually circular, the position and size of the fiber can be determined by detecting the circular edge. The specific detection process includes: When detecting a circle, the equation of the circle is (xa) 2 +(yb) 2 =r 2 , where (a, b) is the coordinate of the center of the circle and r is the radius. The three parameters (a, b, r) constitute a three-dimensional space.

[0095] First, perform edge detection on the image to obtain a set of edge pixels. For each edge pixel (x, y), traverse the possible radius r (usually based on prior knowledge or a certain range to determine the radius value range), and then calculate the corresponding circle center coordinates (a, b). According to the circle equation:

[0096] and

[0097] The ± here means there are two possible solutions, and we vote for the corresponding (a, b, r) combination in the parameter space.

[0098] For example, assuming that the coordinates of an edge pixel are (50, 50) and the radius r = 10, the possible coordinates of the circle center can be calculated, and then the counters of these possible (a, b, r) combinations in the parameter space are increased by 1, indicating that this combination has received one vote.

[0099] In the parameter space, after voting for all edge pixels, a three-dimensional "voting space" is formed. The peak in this space is detected, that is, the (a, b, r) combination with the most votes. Usually, a threshold can be set. When the number of votes exceeds this threshold, it is considered that a circle has been found. The setting of this threshold can be adjusted according to factors such as the noise level of the image and the expected size of the circle. For example, if the threshold is set to 50, then the circle corresponding to the (a, b, r) combination with a vote count greater than 50 in the parameter space is considered to be the detected circle. After finding the (a, b, r) combination corresponding to the peak, the center coordinates (a, b) and radius r of the circle in the image are determined, thus completing the detection of the circular edge.

[0100] Step 10132: Cropping out a first square image with the first center coordinate at the center of the image using the first center coordinate and the first fiber pixel size; cropping out a second square image with the second center coordinate at the center of the image using the second center coordinate and the second fiber pixel size.

[0101] like Figure 9 and Figure 10 As shown, once the circular edge is found, the center coordinates and pixel size of the optical fiber can be determined by calculating the center coordinates and radius of the circle. The specific method includes: for each pixel point on the circular edge, a least squares method or other methods can be used to fit a circle to obtain the center coordinates and radius of the circle.

[0102] The center coordinates can represent the position of the fiber in the image, while the radius can represent the pixel size of the fiber.

[0103] For example, suppose there are n pixels on the edge of the circle detected by the above method, and their coordinates are (x1, y1), (x2, y2) ... (xn, yn). The least squares method can be used to fit a circle, and its equation is (xa) 2 +(yb) 2 =r 2 , where (a, b) is the center coordinate of the circle and r is the radius. Using the least squares method, the values of a, b, and r can be solved to obtain the center coordinates and pixel size of the optical fiber.

[0104] According to the first center coordinates of the first hollow-core optical fiber and the first optical fiber pixel size, a first square image with the first center coordinates at the center of the first end-face image can be cropped. The specific process includes: first determining the side length of the square image. The side length can be determined according to a certain multiple of the optical fiber pixel size. For example, the side length can be set to twice the optical fiber pixel size. Then, with the first center coordinate as the center, a square image with a side length of a determined value is cropped. For example, if the first optical fiber pixel size is d, the side length of the square image can be set to 2d. With the first center coordinate (x0, y0) as the center, a square image with a side length of 2d is cropped, and its upper left corner coordinates are (x0-d, y0-d), and the lower right corner coordinates are (x0+d, y0+d).

[0105] Similarly, based on the second center coordinates of the second hollow-core fiber and the second fiber pixel size, a second square image with the second center coordinates at the center of the second end-face image can be cropped. The cropping method is similar to that of the first square image and will not be repeated here.

[0106] Step 10133: Adjust the sizes of the first square image and the second square image to be consistent, so as to obtain the first grayscale image and the second grayscale image.

[0107] Among them, Figure 11 and Figure 12 As shown, the specific process of adjusting the sizes of the first square image and the second square image to be consistent includes:

[0108] First, the size of the adjusted image needs to be determined. A fixed size can be specified, such as a size of 500 pixels wide and 500 pixels high. Alternatively, the sizes of the two images can be determined and then the average is calculated as the corresponding size. Using the size of one image as a reference, the other image is adjusted to the same size. In one embodiment, for the first square image, the resize function in the image adjustment tool can be used to scale the image according to the specified target size. If the image is enlarged, new pixels will be inserted into the image through an interpolation algorithm; if the image is reduced, pixels will be merged or discarded. For example, the first image is adjusted to a size of 500 pixels wide and 500 pixels high. After processing, the size of the first square image becomes the target size. The same operation is performed on the second square image. After processing, the second square image also reaches the target size. According to the above operation, the first grayscale image and the second grayscale image of the same size are finally obtained.

[0109] Step 1014: remove the cladding information from the first grayscale image and the second grayscale image to obtain the first end face image and the second end face image respectively.

[0110] Since the cropped first grayscale image and the second grayscale image both contain cladding brightness information, it will affect the subsequent rotation steps. In one embodiment, Figure 13 As shown, step 1014 specifically includes:

[0111] Step 10141: Create a circular mask; wherein the size of the circular mask is consistent with the first grayscale image and the second grayscale image, the center of the circular mask is at the center of the image, and the radius of the circular mask is consistent with the first fiber pixel size and the second fiber pixel size.

[0112] Among them, Figure 14 As shown, the center of the circular mask produced as required is located at the center of the image. In one embodiment, the center coordinates can be calculated using the image width and height. Assuming the image width is w and the height is h, the abscissa x of the center is x = w / 2, and the ordinate y = h / 2. The radius r of the circular mask is determined based on the first and second fiber pixel sizes obtained in step 10131.

[0113] By looping through each pixel in the image, the distance formula from the point to the circle center is used to determine whether the point is inside or on the circle. If it is, the pixel value at the corresponding position in the mask is set to 1 (selected); otherwise, it is set to 0 (unselected). This process creates a circular mask that is the same size as the first and second grayscale images, centered at the image center, and has the same radius.

[0114] Step 10142: Perform dot multiplication of the circular mask with the first grayscale image and the second grayscale image to obtain the first end face image and the second end face image respectively.

[0115] Among them, Figure 15 and Figure 16 As shown, a dot product operation is performed on the created circular mask and the first grayscale image. During this dot product, pixel values of 1 in the circular mask will retain the corresponding pixel values in the first grayscale image; pixel values of 0 in the circular mask will be set to 0. This operation ultimately yields the first end-face image. In one embodiment, the same method is used to perform a dot product operation on the circular mask and the second grayscale image to yield the second end-face image.

[0116] When processing images generated by fusion splicers, insufficient resolution is a common problem, especially when dealing with objects such as hollow-core fiber antiresonance structures, which are rich in detail but relatively small in size. Due to the pixel limitations of the fusion splicer itself, the first and second end-face images obtained through the above process often exhibit low resolution and lack clarity, significantly impacting subsequent analysis and processing. To improve this situation, efficient image upsampling technology is employed.

[0117] In one embodiment, pixels of the first end surface image and the second end surface image may be amplified, and the process includes:

[0118] Determine the magnification ratio, such as magnifying to k times the original value; for each pixel point of the target image (i.e., the end face image after pixel magnification), calculate its corresponding coordinates in the original image (i.e., the first end face image and the second end face image), and determine each pixel point in a 4x4 neighborhood around the coordinate (16 pixels in the original image); for each pixel point in the 4x4 neighborhood of the coordinate, calculate the distance of each pixel point in the x direction and the y direction, and use an interpolation function to calculate the weight in each direction, and then combine them to obtain a two-dimensional weight; multiply the value of each color channel of the 16 pixels by the corresponding two-dimensional weight and accumulate them to obtain the value of the pixel point in the target image; repeat the above steps for each pixel point in the target image, and finally obtain the magnified first end face image and second end face image.

[0119] In summary, by considering the information of the surrounding 16 pixels to calculate the value of each pixel in the target image, a smoother and more natural magnification effect can be obtained, which not only better preserves the sharpness of the image edge, but also effectively reduces aliasing and blur, and provides a higher quality magnified image. In one embodiment, the first end face image and the second end face image can be upsampled by 1.5 (that is, the k value). The choice of this multiple is based on the consideration of balancing the improvement of image quality and the consumption of computing resources. Too high a magnification factor may introduce unnecessary computational complexity and potential distortion problems; too low a magnification factor may not achieve the expected clarity improvement effect. In one embodiment, the pixel size of the first end face image and the second end face image is increased from 101×101 to 152×152.

[0120] In order to ensure the accuracy of the subsequent cross-correlation calculation, it is also necessary to remove the noise effect on the first end face image and the second end face image after pixel amplification. In one embodiment, Figure 17 As shown, the filtering process of the first end face image and the second end face image is the same, and filtering the first end face image to obtain the first filtered image specifically includes:

[0121] Step 1021: Create a two-dimensional matrix as a convolution kernel, use the convolution kernel as a sliding window, and move the sliding window pixel by pixel on the first end-face image; wherein, each time the sliding window moves, the center of the sliding window is aligned with a pixel point in the first end-face image.

[0122] In one embodiment, a 5x5 sized two-dimensional matrix is defined as the convolution kernel, and filter

[0123] =[-1,-1,-1,-1,-1;

[0124] -1,1,1,1,-1;

[0125] -1,1,9,1,-1;

[0126] -1,1,1,1,-1;

[0127] -1,-1,-1,-1,-1];

[0128] The center value of the convolution kernel filter is 9, the surrounding values are -1, and the center value is 1. When the convolution kernel slides over the end-face image to perform the convolution operation, the center value of 9 will have a greater impact on the pixels at the corresponding position in the image. If the position is at the edge or detail area of the image, the surrounding pixel values will change relatively significantly. After multiplying and summing the values with the corresponding position of the convolution kernel, the pixel values at that position will change significantly, thereby highlighting the edges and details. For example, at the edge of an image, the pixel values on one side of the edge are different from those on the other side. When the convolution kernel is applied, the center value 9 is multiplied by the edge pixels, and the surrounding values are multiplied by the adjacent pixels. Through the summation calculation, the features of the edge pixels will be enhanced.

[0129] Noise typically manifests as random, localized, sudden changes in pixel values in an image. Because of the negative values (-1) surrounding the convolution kernel, when the convolution kernel is applied to a noise point, the surrounding pixel values multiplied by the negative value offset some of the noise point's impact, thereby reducing the noise's impact on the image to a certain extent. For example, for an isolated noise point, the pixel values surrounding it are relatively stable. Multiplying them by the -1 value surrounding the convolution kernel weakens the noise point's intensity.

[0130] Step 1022: When the sliding window is aligned with a pixel point, each element in the convolution kernel is multiplied by the corresponding pixel value in the area covered by the sliding window on the first end face image, and all multiplication results are added together to obtain the updated value of the corresponding pixel point after filtering, until the update of all pixel points in the first end face image is completed to obtain the first filtered image.

[0131] In one embodiment, a defined 5x5 convolution kernel is applied to the first end-face image. Starting from the upper left corner of the first end-face image, the center of the convolution kernel is aligned with each pixel of the first end-face image in turn. For each pixel, the function takes out an image area centered on the pixel and of the same size as the convolution kernel (5x5), and then multiplies the pixel values in the area with the elements of the convolution kernel one by one, and adds all the products to obtain the result as the new filtered value of the pixel. As the convolution kernel slides pixel by pixel on the image, all the pixels of the entire first end-face image will be updated, such as Figure 18 As shown, the first filtered image is finally obtained. Figure 19 As shown, the second filtered image is obtained according to the same operation as above.

[0132] In one embodiment, when processing pixel points at the boundary of the first end face image, zero padding is performed outside the boundary of the first end face image, or the boundary pixel points are copied according to preset requirements so that the sliding window can also perform a complete convolution calculation when it moves to the boundary pixel points.

[0133] Among them, there are usually two ways to perform convolution calculations on edge pixels. The first way is: before performing convolution processing on the first end face image, add one or more circles of pixels with a pixel value of 0 outside the boundary of the image. The number of circles added depends on the size of the convolution kernel. Taking a 5x5 convolution kernel as an example, it is usually necessary to add two circles of pixels with a pixel value of 0 on the top, bottom, left and right of the image. In this way, when the sliding window moves to the pixel point at the boundary of the image, the part of the sliding window that exceeds the image boundary corresponds to the area with these newly added pixel values of 0. When performing convolution calculations, for the part of the sliding window that exceeds the image boundary, since it corresponds to a filling pixel with a value of 0, the result after multiplying with the element at the corresponding position of the convolution kernel is still 0, and does not affect the sum of the results of multiplying other normal pixels with the convolution kernel elements. In this way, it can be ensured that the sliding window can also perform a complete convolution calculation at the boundary pixel points.

[0134] The second method involves replicating boundary pixels based on pre-defined requirements. This involves copying pixels at the image boundary to extend the image boundary. Specific replication methods can be used in a variety of ways, such as replicating the outermost rows or columns. For the upper boundary of an image, for example, the pixel values of the top row are copied and added to the upper portion of the image as the extended pixels. For the lower boundary, the pixel values of the bottom row are copied and added to the lower portion of the image. The same applies to the left and right boundaries. This way, when the sliding window moves to a boundary pixel, the portion beyond the image boundary corresponds to the replicated boundary pixels. During the convolution calculation, the replicated pixels corresponding to the portion of the sliding window beyond the image boundary are multiplied by the corresponding position element of the convolution kernel, and the result is included in the final convolution calculation. Because the boundary pixels are replicated, feature continuity at the image boundary is maintained to a certain extent, allowing the convolution calculation to proceed properly at the boundary pixels. This also avoids the loss of features or unnatural effects at the boundary that may be caused by zero-padding. Different replication methods can be selected based on different pre-defined requirements to suit the specific image processing task and boundary processing requirements.

[0135] The operation on the second end face image is similar to the above process. The same 5x5 convolution kernel is applied to the input image according to the above convolution operation method to obtain the filtered second filtered image.

[0136] By applying the above-mentioned filtering operation of a specific convolution kernel to the two end-face images, the final filtered images are enhanced in terms of details and edges, while the influence of noise is also reduced, thereby improving the quality and analyzability of the images.

[0137] After obtaining the first filtered image and the second filtered image, rotating the first filtered image angle by angle, and performing cross-correlation calculation on the rotated first filtered image and the second filtered image respectively until an optimal alignment angle corresponding to a maximum cross-correlation coefficient is obtained, specifically includes:

[0138] A rotation angle range and a rotation angle step are set, and the first filtered image is rotated angle by angle according to the rotation angle range and the rotation angle step; a plurality of first filtered images after rotation are cross-correlated with the second filtered image to obtain a cross-correlation coefficient, and the angle corresponding to the maximum cross-correlation coefficient is selected as the optimal alignment angle.

[0139] In one embodiment, the first filtered image can be rotated multiple times to obtain multiple rotated first filtered images, and then the multiple rotated first filtered images are cross-correlated with the second filtered image respectively; or, the first filtered image after one rotation is cross-correlated with the second filtered image, and then a second and third rotation is performed until the rotation angle reaches the maximum angle value, and finally the size of the cross-correlation coefficient corresponding to each angle is compared.

[0140] First, determine the rotation range, typically from 0 to 360 degrees. However, this angle range can be narrowed based on the fiber's actual conditions, such as physical limitations or a known approximate alignment range. Also, determine the rotation step size—the angular increment for each rotation. A smaller step size improves calculation accuracy, but also increases the computational complexity. For example, the step size can be set to 0.1 or 0.5 degrees.

[0141] The first filtered image is rotated starting from an initial angle (usually 0 degrees) according to a set step size. The rotation operation can be implemented by image processing software or related functions in a programming language. During the rotation process, the pixels of the image are rearranged according to the rotation matrix. For example, for a pixel point (x, y) in a two-dimensional image (the first filtered image), when the rotation angle is θ, its new coordinates (x', y') can be obtained by the matrix rotation formula, which is:

[0142] x′=xcosθ-ysinθ and y′=x sinθ+ycosθ

[0143] For the rotated first filtered image A and the second filtered image B, the cross-correlation calculation is performed by calculating the sum of the products of the corresponding pixels of A and B within a certain displacement range to measure their similarity. The calculation formula of the cross-correlation coefficient is:

[0144]

[0145] Wherein, R is the cross-correlation coefficient, M and N are the number of rows and columns of the image matrix A(x, y) of the first filtered image and the image matrix B(x, y) of the second filtered image, respectively. is the average value of A, is the average value of B, x is the horizontal coordinate, and y is the vertical coordinate.

[0146] After each rotation and completion of the cross-correlation calculation, the obtained cross-correlation coefficient is recorded. The cross-correlation coefficient reflects the similarity between the two filtered images at the current rotation angle and displacement. After completing the rotation and cross-correlation calculations for all angles (within the set rotation range), compare all the recorded cross-correlation coefficients. Find the maximum value among them, and the rotation angle corresponding to this maximum value is the optimal alignment angle. Because at this angle, the similarity between the two filtered images is the highest, that is, the degree of alignment between the cores of the two hollow-core optical fibers and the anti-resonance structure is the best. In one embodiment, as Figure 20 As shown in the figure, the horizontal axis is the rotation angle, the vertical axis is the correlation coefficient, the left is the first filtered image after rotation, and the right is the second filtered image. It can be seen that when the rotation angle is 46.9°, the two images have the best correspondence.

[0147] The automatic alignment scheme proposed in this embodiment is based on the end face image of the hollow-core fiber, can efficiently extract the end face information of the hollow-core fiber, simplify the image processing process, and optimize the angle matching through an adaptive rotation adjustment algorithm, so that the cores and antiresonance structures of the two hollow-core fibers can be perfectly aligned, and can achieve precise automatic alignment while maintaining high-speed processing.

[0148] Example 2:

[0149] In Example 1, a hollow-core optical fiber automatic alignment method is provided. In this embodiment, a hollow-core optical fiber automatic alignment device is proposed. The hollow-core optical fiber automatic alignment device includes: a processor and a memory for storing processor-executable instructions; wherein the processor is configured to execute the hollow-core optical fiber automatic alignment method described in Example 1.

[0150] like Figure 21 As shown, the hollow-core optical fiber automatic alignment device includes a processor 21 and a memory 22 , wherein the processor 21 and the memory 22 can be connected via a bus or other means.

[0151] The processor 21 may be a central processing unit (CPU). The processor 21 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of these chips.

[0152] The memory 22, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the hollow-core optical fiber automatic alignment method in Example 1 of the present invention. The processor executes the non-transitory software programs, instructions, and modules stored in the memory to perform various functional applications and training processes of the processor.

[0153] The memory 22 may include a program storage area and a training storage area, wherein the program storage area may store an operating system and applications required for at least one function; the training storage area may store training created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 22 may optionally include a memory remotely located relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned networks include but are not limited to the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The one or more modules are stored in the memory 22, and when executed by the processor 21, execution is performed as follows. Figure 1 The hollow core optical fiber automatic alignment method in the embodiment 1 is shown. The specific details of the above hollow core optical fiber automatic alignment method can be referred to the corresponding Figure 1 、 Figure 2 and Figure 3 The corresponding descriptions and effects in the embodiments shown in the figure can be understood and will not be repeated here.

[0154] This embodiment further provides a computer storage medium, wherein the storage medium stores a computer program. The computer program can be executed by a processor to implement the hollow-core optical fiber automatic alignment method described in Embodiment 1.

[0155] The computer storage medium stores computer-executable instructions that can execute the hollow-core optical fiber automatic alignment method of any of the above-described method embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD); the storage medium can also include a combination of the above-mentioned types of memory.

[0156] The specific steps of the hollow-core optical fiber automatic alignment method are shown in Example 1 and will not be repeated in this embodiment.

[0157] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A hollow-core optical fiber automatic alignment method, characterized in that: include: Acquire a first end face image of the first hollow core optical fiber and a second end face image of the second hollow core optical fiber; filtering the first end face image to obtain a first filtered image, and filtering the second end face image to obtain a second filtered image; Rotating the first filtered image angle by angle, and performing cross-correlation calculation on the rotated first filtered image and the second filtered image respectively, until an optimal alignment angle corresponding to a maximum cross-correlation coefficient is obtained; The first hollow-core optical fiber is rotated to the optimal alignment angle to align the first hollow-core optical fiber with the second hollow-core optical fiber.

2. The hollow-core optical fiber automatic alignment method according to claim 1, characterized in that: The method further comprises: Stripping the coating layers of the fusion-spliced ends of the first hollow-core optical fiber and the second hollow-core optical fiber, and neatly cutting the fusion-spliced end faces of the first hollow-core optical fiber and the second hollow-core optical fiber, respectively placing the first hollow-core optical fiber and the second hollow-core optical fiber into the clamps at both ends of the fusion splicer and fixing them; Adjusting the positions of the fusion-spliced end faces of the first hollow-core optical fiber and the second hollow-core optical fiber, and obtaining a first intermediate image corresponding to the first hollow-core optical fiber and a second intermediate image corresponding to the second hollow-core optical fiber through end face imaging of a fusion splicer; performing grayscale conversion on the first intermediate image and the second intermediate image and adjusting them to the same size to obtain a first grayscale image and a second grayscale image, respectively; The cladding information in the first grayscale image and the second grayscale image is removed to obtain the first end face image and the second end face image, respectively.

3. The hollow-core optical fiber automatic alignment method according to claim 2, characterized in that: The method further comprises: Finding a circular edge of the first intermediate image, and obtaining a first center coordinate and a first fiber pixel size of the first hollow-core optical fiber; Finding a circular edge of the second intermediate image, and obtaining a second center coordinate and a second fiber pixel size of the second hollow-core optical fiber; Cut out a first square image with the first center coordinate at the center of the image according to the first center coordinate and the first optical fiber pixel size; Cut out a second square image with the second center coordinate at the center of the image according to the second center coordinate and the second optical fiber pixel size; The sizes of the first square image and the second square image are adjusted to be consistent, so as to obtain the first grayscale image and the second grayscale image.

4. The hollow-core optical fiber automatic alignment method according to claim 3, characterized in that: The method further comprises: Creating a circular mask; wherein the size of the circular mask is consistent with the first grayscale image and the second grayscale image, the center of the circular mask is at the center of the image, and the radius of the circular mask is consistent with the first fiber pixel size and the second fiber pixel size; Perform dot multiplication of the circular mask with the first grayscale image and the second grayscale image to obtain the first end surface image and the second end surface image respectively.

5. The hollow-core optical fiber automatic alignment method according to claim 1, wherein: The method further comprises: Creating a two-dimensional matrix as a convolution kernel, using the convolution kernel as a sliding window, and moving the sliding window pixel by pixel on the first end-face image; wherein, during each movement, the center of the sliding window is aligned with a pixel point in the first end-face image; When the sliding window is aligned with a pixel point, each element in the convolution kernel is multiplied by the corresponding pixel value in the area covered by the sliding window on the first end face image, and all multiplication results are added together to obtain the updated value of the corresponding pixel point after filtering, until the update of all pixel points in the first end face image is completed to obtain the first filtered image.

6. The hollow-core optical fiber automatic alignment method according to claim 5, characterized in that: When processing pixel points at the boundary of the first end face image, zero padding is performed outside the boundary of the first end face image, or the boundary pixel points are copied according to preset requirements so that the sliding window can also perform a complete convolution calculation when it moves to the boundary pixel points.

7. The hollow-core optical fiber automatic alignment method according to claim 1, wherein: The method further comprises: Setting a rotation angle range and a rotation angle step, and rotating the first filtered image angle by angle according to the rotation angle range and the rotation angle step; Cross-correlation calculation is performed on the rotated plurality of first filtered images and the second filtered image to obtain a cross-correlation coefficient, and an angle corresponding to a maximum cross-correlation coefficient is selected as an optimal alignment angle.

8. The hollow-core optical fiber automatic alignment method according to claim 1, wherein: The calculation formula of the mutual correlation coefficient is: Wherein, R is the cross-correlation coefficient, M and N are the number of rows and columns of the image matrix A(x, y) of the first filtered image and the image matrix B(x, y) of the second filtered image, respectively. is the average value of A, is the average value of B, x is the horizontal coordinate, and y is the vertical coordinate.

9. A hollow-core optical fiber automatic alignment device, characterized in that: The hollow core optical fiber automatic alignment device includes: a processor and a memory for storing instructions executable by the processor; The processor is configured to execute the hollow-core optical fiber automatic alignment method according to any one of claims 1 to 8.

10. A non-volatile computer storage medium, characterized in that The computer storage medium stores computer-executable instructions, and the computer-executable instructions are executed by one or more processors to implement the hollow-core optical fiber automatic alignment method according to any one of claims 1 to 8.

Images