A method, device, system and storage medium for detecting optical fiber bundle defects

Abstract

The present application discloses a method, device, system and storage medium for optical fiber bundle defect detection, which are used to improve the accuracy of optical fiber bundle defect detection. The present application includes: using a shading strip to perform petal shielding on part of the split ends, leaving only one split end; performing image acquisition on the combined end of the optical fiber bundle to be tested, and generating a combined end image; performing a first binarization process on the combined end image, and segmenting the optical fiber petal end face on the combined end image; performing noise removal process on the non-optical fiber area of ​​the optical fiber petal end face on the combined end image; performing a second binarization process on the optical fiber petal end face on the combined end image; using morphological corrosion process and area screening process on the optical fiber petal end face on the combined end image to extract the area of ​​each optical fiber in the optical fiber petal end face; performing dark defect analysis and abnormal quantity analysis on the area of ​​each optical fiber, and generating the petal optical fiber analysis result of the combined end.

Description

Technical Field

[0001] Embodiments of the present application relate to the field of optical fiber bundle detection, and in particular, to a method, device, system, and storage medium for optical fiber bundle defect detection. Background Art

[0002] Fiber bundle quality and quantity detection is a crucial part of the fiber acceptance process. Each fiber bundle is composed of several integrated optical fibers, forming multiple fan-shaped petals in a fiber bundle. In the prior art, the fiber quantity detection and damage identification method is manual detection. The main problems of manual detection are: low efficiency, strong human subjective factors, and inability to quickly and accurately identify the quantity and quality of fiber bundles.

[0003] In order to overcome the above problems, in the prior art, the cross-section of the optical fiber bundle is imaged and quickly inspected through recognition networks and machine vision. However, in the existing machine vision inspection, the number and quality of optical fibers are usually directly inspected for the entire optical fiber bundle end face, and each petal of the optical fiber bundle end face is not counted separately. In addition, the optical fiber is mainly judged to be qualified by the overall luminous profile of the optical fiber, and the light transmission quality of the overall complete profile optical fiber cannot be inspected. That is, although the overall luminous profile shows that it is qualified, the light transmission uniformity of the optical fiber may still be poor, that is, the light transmission uniformity of the optical fiber cannot be evaluated, resulting in a reduction in the accuracy of optical fiber bundle defect detection. Summary of the invention

[0004] The present application discloses a method, device, system and storage medium for detecting optical fiber bundle defects, which are used to improve the detection accuracy of optical fiber bundle defects.

[0005] The first aspect of the present application discloses a method for detecting optical fiber bundle defects, comprising:

[0006] Place the split ends of the optical fiber bundle to be tested on a light homogenizing plate, and use a light shielding strip to block part of the split ends, leaving only one split end. The optical fiber bundle to be tested includes N split ends and one combined end, where N is an integer greater than 2.

[0007] The image acquisition device is used to acquire an image of the combined end of the optical fiber bundle to be tested, and an image of the combined end is generated. The combined end has two symmetrically distributed optical fiber petals.

[0008] The first binarization process is performed on the beam-combined end image according to the preset split end face grayscale threshold, and the fiber split end face is segmented on the beam-combined end image;

[0009] Perform noise removal processing on the non-fiber area of ​​the fiber split end face on the beam combining end image;

[0010] A second binarization process is performed on the fiber split end face on the beam combining end image according to a preset fiber grayscale threshold;

[0011] The fiber split end face on the beam combining end image after the second binarization process is subjected to morphological corrosion processing and area screening processing to extract the area of ​​each optical fiber in the fiber split end face;

[0012] The dark defect analysis and abnormal quantity analysis are performed on the area of ​​each optical fiber to generate the analysis results of the split optical fiber at the combined end.

[0013] Optionally, noise removal processing is performed on the non-fiber region of the fiber split end face on the beam combining end image, including:

[0014] Performing optical fiber feature screening on the optical fiber splitting end face to determine an optical fiber feature area, where the optical fiber feature area is an area where only the optical fiber splitting is retained;

[0015] The fiber split end face on the beam combining end image is cropped according to the fiber feature area to remove the noise area generated during the binarization process.

[0016] Optionally, after performing dark defect analysis and abnormal quantity analysis on the region of each optical fiber to generate the first split optical fiber analysis result, the method further includes:

[0017] Use the shading strip to re-block the split end, retaining only one split end each time, and take the image of the combined end to perform defect detection on the corresponding fiber split, generating a total of N split fiber analysis results;

[0018] Perform fiber bundle uniformity analysis on the N split optical fiber analysis results to generate uniformity analysis results.

[0019] Optionally, after performing image acquisition on the combined end of the optical fiber bundle to be tested by an image acquisition device to generate a combined end image, performing a first binarization process on the combined end image according to a preset petal end face grayscale threshold, and before segmenting the optical fiber petal end face on the combined end image, the method further includes:

[0020] Calculate the attention matrix for each pixel on the combined end image;

[0021] According to the attention value of the attention matrix, the fiber region of the beam-combining end image is roughly divided to determine two petal regions;

[0022] Determine the optical fiber center coordinates of the beam combining end image;

[0023] Calculate the standard deviation and grayscale mean corresponding to the two petal areas in the combined end image;

[0024] Calculate the convolution kernel adjustment coefficient according to the grayscale mean;

[0025] Calculate the petal-splitting optical fiber filter kernel according to the optical fiber center coordinates, standard deviation, convolution kernel adjustment coefficient and preset convolution kernel function;

[0026] The image at the beam combining end is filtered by using the split fiber filter core.

[0027] Optionally, determining the optical fiber center coordinates of the beam combining end image includes:

[0028] Acquire a photographic image of the optical fiber bundle to be tested that is not shielded by the light shielding strip;

[0029] Determine the circular fiber profile of the captured image according to the edge inspection algorithm, and calibrate the center coordinates of the circular fiber profile;

[0030] The optical fiber center coordinates of the beam combining end image are determined according to the optical fiber profile of the captured image and the calibrated center coordinates.

[0031] Optionally, determining the optical fiber center coordinates of the beam combining end image includes:

[0032] Determine the pixel coordinates of the four corner points of the two petal areas on the beam-combined end image, as well as the coordinates of the two tips with the shortest distance between the two petal areas;

[0033] Cross-connect the pixel coordinates of the four corner points to generate the first center coordinates;

[0034] Take the center point of the line connecting the two tip coordinates as the second center coordinate;

[0035] The center point of the line connecting the first center coordinate and the second center coordinate is taken as the center coordinate of the optical fiber.

[0036] The second aspect of the present application discloses a device for detecting optical fiber bundle defects, comprising:

[0037] The preprocessing unit is used to place the split ends of the optical fiber bundle to be tested on the light homogenizing plate, and use the shading strip to block part of the split ends, leaving only one split end, and the optical fiber bundle to be tested includes N split ends and one beam combining end, where N is an integer greater than 2;

[0038] The first generating unit is used to collect images of the combined end of the optical fiber bundle to be tested through an image acquisition device to generate an image of the combined end, where the combined end has two symmetrically distributed optical fiber petals;

[0039] A segmentation unit, used for performing a first binarization process on the beam-combined end image according to a preset split end face grayscale threshold, and segmenting the optical fiber split end face on the beam-combined end image;

[0040] A denoising unit is used to remove noise in non-fiber regions of the fiber split end face on the beam combining end image;

[0041] A binarization unit, used for performing a second binarization process on the fiber split end face on the beam combining end image according to a preset fiber grayscale threshold;

[0042] An extraction unit, used for performing morphological corrosion processing and area screening processing on the fiber split end face on the beam combining end image after the second binarization processing, so as to extract the area of ​​each optical fiber in the fiber split end face;

[0043] The second generating unit is used to perform dark defect analysis and abnormal quantity analysis on the area of ​​each optical fiber, and generate the analysis result of the split optical fiber at the combined end.

[0044] Optional, denoising unit, including:

[0045] Performing optical fiber feature screening on the optical fiber splitting end face to determine an optical fiber feature area, where the optical fiber feature area is an area where only the optical fiber splitting is retained;

[0046] The fiber split end face on the beam combining end image is cropped according to the fiber feature area to remove the noise area generated during the binarization process.

[0047] Optionally, after the second generating unit, the device further includes:

[0048] The third generation unit uses a shading strip to re-shield the split end, retaining only one split end each time, and takes an image of the beam combining end to perform defect detection on the corresponding optical fiber petals, generating a total of N split optical fiber analysis results;

[0049] The fourth generating unit is used to perform optical fiber bundle uniformity analysis on the N split optical fiber analysis results to generate a uniformity analysis result.

[0050] Optionally, after the first generating unit and before the segmenting unit, the device further includes:

[0051] A first calculation unit is used to calculate the attention matrix for each pixel point on the combined end image;

[0052] A first determination unit is used to roughly divide the optical fiber area of ​​the beam combining end image according to the size of the attention value of the attention matrix to determine two petal areas;

[0053] A second determining unit, used to determine the optical fiber center coordinates of the beam combining end image;

[0054] The second calculation unit is used to calculate the standard deviation and grayscale mean corresponding to the two petal areas in the beam-combined end image;

[0055] A third calculation unit is used to calculate a convolution kernel adjustment coefficient according to the grayscale mean;

[0056] A fourth calculation unit, used for calculating the petal-split optical fiber filter kernel according to the optical fiber center coordinates, the standard deviation, the convolution kernel adjustment coefficient and the preset convolution kernel function;

[0057] The filtering unit is used to filter the image at the beam combining end through the split fiber filter core.

[0058] Optionally, the second determining unit includes:

[0059] Acquire a photographic image of the optical fiber bundle to be tested that is not shielded by the light shielding strip;

[0060] Determine the circular fiber profile of the captured image according to the edge inspection algorithm, and calibrate the center coordinates of the circular fiber profile;

[0061] The optical fiber center coordinates of the beam combining end image are determined according to the optical fiber profile of the captured image and the calibrated center coordinates.

[0062] Optionally, the second determining unit includes:

[0063] Determine the pixel coordinates of the four corner points of the two petal areas on the beam-combined end image, as well as the coordinates of the two tips with the shortest distance between the two petal areas;

[0064] Cross-connect the pixel coordinates of the four corner points to generate the first center coordinates;

[0065] Take the center point of the line connecting the two tip coordinates as the second center coordinate;

[0066] The center point of the line connecting the first center coordinate and the second center coordinate is taken as the center coordinate of the optical fiber.

[0067] A third aspect of the present application provides a system for detecting optical fiber bundle defects, comprising:

[0068] Processor, memory, input-output unit, and bus;

[0069] The processor is connected to the memory, the input and output unit, and the bus;

[0070] The memory stores a program, and the processor calls the program to execute the first aspect and any optional method of the first aspect.

[0071] A fourth aspect of the present application provides a computer-readable storage medium, on which a program is stored. When the program is executed on a computer, the program executes the first aspect and any optional method of the first aspect.

[0072] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0073] In the present application, the beam splitting end of the optical fiber bundle to be tested is first placed on a light homogenizing plate, and a shading strip is used to block part of the beam splitting end, leaving only one beam splitting end, wherein the optical fiber bundle to be tested includes N beam splitting ends and one beam combining end, and N is an integer greater than 2. A beam splitting end forms two symmetrical fan-shaped lobes on the beam combining end, and the two stacked optical fiber lobes are first detected. The beam combining end of the optical fiber bundle to be tested is imaged by an image acquisition device to generate an image of the beam combining end, and the beam combining end has two symmetrically distributed optical fiber lobes. The beam combining end image is first binarized according to a preset grayscale threshold of the lobed end face, and the optical fiber lobed end face is segmented on the beam combining end image. That is, the other parts except the two optical fiber lobes are removed, because the other beam splitting ends are blocked, so the optical fiber lobes corresponding to these blocked beam splitting ends on the beam combining end are not lit, and the background area and other unlit optical fiber lobes are effectively removed by binarization through the preset grayscale threshold of the lobed end face. Next, the fiber flap end face on the beam-combining end image is subjected to noise removal processing in the non-fiber area. The noise at the edge of the fiber bundle is removed. The fiber flap end face on the beam-combining end image is subjected to a second binarization processing according to the preset fiber grayscale threshold. Each fiber in the fiber flap is initially visualized. The fiber flap end face on the beam-combining end image after the second binarization processing is subjected to morphological corrosion processing to disconnect the connection between the fibers and remove the edge burrs, and then the area screening processing is performed to extract the area of ​​each fiber in the fiber flap end face. Dark defect analysis and abnormal number analysis are performed on the area of ​​each fiber to generate the analysis results of the flap fiber at the beam-combining end. Not only can the number of abnormal fibers in the two symmetrical fiber flaps be determined, but also whether the uniformity of the symmetrical fiber flap is qualified can be determined, so as to realize the identification of the number of flaps on the fiber bundle end face, and the optical fiber transmittance can be judged. Finally, each flap of the fiber bundle can be accurately counted and the optical fiber with poor transmittance can be detected, thereby improving the accuracy of fiber bundle defect detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0075] Figure 1 A schematic diagram of an embodiment of a method for detecting optical fiber bundle defects of the present application;

[0076] Figure 2 A schematic diagram of an embodiment of a method for removing noise from a non-fiber region of a fiber split end face of the present application;

[0077] Figure 3 A schematic diagram of an embodiment of a method for analyzing uniformity of an optical fiber bundle of the present application;

[0078] Figure 4 A schematic diagram of an embodiment of a method for filtering a beam-combining end image in the present application;

[0079] Figure 5 A schematic diagram of an embodiment of a method for determining the optical fiber center coordinates of a beam combining end image according to the present application;

[0080] Figure 6 A schematic diagram of another embodiment of the method for determining the optical fiber center coordinates of the beam combining end image of the present application;

[0081] Figure 7 A schematic diagram of another embodiment of the optical fiber bundle defect detection device of the present application;

[0082] Figure 8 A schematic diagram of a system for optical fiber bundle defect detection in the present application;

[0083] Fig. 9 The principle diagram of morphological corrosion operation using 3*3 structural elements in this application;

[0084] Fig.10 This is the effect diagram of using 3*3 structural elements to perform morphological corrosion operations and area screening in this application;

[0085] Fig.11 A schematic diagram of an image before optical fiber feature screening for this application;

[0086] Fig.12 A schematic diagram of an image after noise removal in this application. DETAILED DESCRIPTION

[0087] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0088] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0089] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0090] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0091] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0092] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0093] In the prior art, the cross-section of the optical fiber bundle is imaged and quickly inspected through recognition networks and machine vision. However, in the existing machine vision inspection, the optical fiber quantity and quality are usually directly inspected for the entire optical fiber bundle end face, and each petal of the optical fiber bundle end face is not counted separately. In addition, the optical fiber is mainly judged to be qualified by the overall luminous profile of the optical fiber, and the light transmission quality of the overall complete profile optical fiber cannot be inspected. That is, although the overall luminous profile shows that it is qualified, the light transmission uniformity of the optical fiber may still be poor, that is, the light transmission uniformity of the optical fiber cannot be evaluated, resulting in a reduction in the accuracy of optical fiber bundle defect detection.

[0094] Based on this, the present application discloses a method, device, system and storage medium for detecting optical fiber bundle defects, which are used to improve the accuracy of optical fiber bundle defect detection.

[0095] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0096] The method of the present application can be applied to a server, a device, a terminal or other devices with logic processing capabilities, and the present application does not limit this. For the convenience of description, the following description is made by taking the execution subject as an example of a terminal.

[0097] See also Figure 1 The present application provides an embodiment of a method for detecting optical fiber bundle defects, comprising:

[0098] 101. Place the split ends of the optical fiber bundle to be tested on a light homogenizing plate, and use a shading strip to block part of the split ends, leaving only one split end. The optical fiber bundle to be tested includes N split ends and one combined end, where N is an integer greater than 2.

[0099] In this embodiment, the splitting end of the optical fiber bundle to be tested is placed on a light homogenizing plate, with the cross section of the optical fiber bundle to be tested facing upward, and the splitting end is selectively shielded by a shading strip, and a black and white camera and a telecentric lens are used to take pictures of the combined end of the optical fiber bundle to be tested from top to bottom. Specifically, the cross section of the optical fiber bundle to be tested is circular, including 4 splitting ends and one combined end, and there are 4 optical fibers of the same number at the 4 splitting ends, and there are 8 proportionally distributed fan-shaped areas at the combined end, and each splitting end occupies two symmetrically distributed fan-shaped areas on the combined end, forming symmetrical optical fiber petals.

[0100] 102. Capture an image of the combined end of the optical fiber bundle to be tested by an image acquisition device to generate an image of the combined end, where the combined end has two symmetrically distributed optical fiber petals;

[0101] The terminal uses an image acquisition device to capture images of the combined end of the optical fiber bundle to be tested and generates an image of the combined end. Because of the occlusion, the two symmetrically distributed optical fiber petals at the combined end are the most conspicuous.

[0102] 103. Performing a first binarization process on the beam-combined end image according to a preset split end face grayscale threshold, and segmenting the optical fiber split end faces on the beam-combined end image;

[0103] The terminal performs the first binarization processing on the beam-combined end image according to the preset split end face grayscale threshold, so as to segment the fiber split end face on the beam-combined end image. For example, the split end face grayscale threshold is set to 30 grayscales, as shown in the following formula;

[0104]

[0105] The pixels with grayscale greater than 30 on the image are set to 255, and the others are set to 0. After the fiber splitting area is obtained, the area is selected on the original image of the combined end of the beam to obtain the fiber splitting end face with satisfactory grayscale. The fiber splitting end face is roughly shaped like two fan-shaped areas.

[0106] 104. Perform noise removal processing on the non-fiber region of the fiber split end face on the beam combining end image;

[0107] In step 103, the extracted optical fiber split end face may contain some background areas or other noise-generating areas that need to be cleared. In this embodiment, since it is known that the optical fiber split is a 45-degree fan-shaped area and the size parameters can be determined, noise removal processing in non-optical fiber areas can be performed in a variety of ways, which are not limited here.

[0108] 105. Perform a second binarization process on the fiber split end face on the beam combining end image according to a preset fiber grayscale threshold;

[0109] The terminal performs a second binarization process on the fiber lob end face on the beam-combining end image according to the preset fiber grayscale threshold. The binarization process here is slightly different from the binarization process in the previous step. The binarization process in the previous step is used to determine the area of ​​the two fiber lobes, while the binarization process in the current step is to determine the area of ​​each fiber according to the preset fiber grayscale threshold. Because some fibers are abnormal and do not emit light obviously, some are completely damaged and do not emit light, so through the binarization process, normal fibers and abnormal fibers can be determined, and then the number of fibers in each fiber lob can be used to determine the completely damaged fibers. The fiber grayscale threshold is used to distinguish the grayscale value of the fiber and other parts in the image, which can be set according to the empirical value. When the fiber with the split end is placed on the light-diffusing plate, due to the obstruction of the lobes, the obscured lobes have low light transmittance, which makes the difference between the light-transmitting fiber and the background area (including the obscured part) large, and the grayscale difference in the image is amplified, which can better screen the fiber area (second binarization process).

[0110] 106. Perform morphological corrosion processing and area screening processing on the fiber split end face on the beam combining end image after the second binarization processing to extract the area of ​​each optical fiber in the fiber split end face;

[0111] The terminal uses morphological corrosion processing and area screening processing on the fiber split end face on the beam-combining end image after the second binarization processing. Morphological processing can disconnect some subtle connections and remove some burrs that may exist on the edge, which is conducive to subsequent feature extraction. Area screening processing is to better select the area that meets the fiber area, and finally extract the area of ​​each fiber in the fiber split end face.

[0112] The morphological erosion process in this embodiment is to use the structure element B to slide on the second binarized image A. When the structure element B covers the pixels of the binary image A, the structure element B is used to perform a logical "AND" operation with the binary image A. If the area covered by the structure element is all 1, the binary image pixel is 1, otherwise it is 0. Although erosion can disconnect some connection points in the image, it will cause the area of ​​the image to be smaller than the original area. Fig. 9 As shown, Fig. 9 This is the principle diagram of using 3*3 structure element to perform morphological corrosion operation. Please check Fig.10 This is the effect diagram of using 3*3 structural elements for morphological corrosion operation and area screening.

[0113] 107. Perform dark defect analysis and abnormal quantity analysis on the area of ​​each optical fiber to generate the analysis results of the split optical fiber at the combined end.

[0114] First, after morphological corrosion processing and area screening processing, the terminal compares the grayscale of each extracted optical fiber with that of the normal optical fiber. If the difference reaches 30%, the optical fiber needs to be determined as an abnormal optical fiber. Then, each optical fiber is segmented and counted to generate the number of abnormal optical fibers and the number of normal optical fibers.

[0115] In this embodiment, the beam splitting end of the optical fiber bundle to be tested is first placed on a light homogenizing plate, and a light shielding strip is used to block part of the beam splitting end, leaving only one beam splitting end, wherein the optical fiber bundle to be tested includes N beam splitting ends and one beam combining end, and N is an integer greater than 2. One beam splitting end forms two symmetrical fan-shaped lobes on the beam combining end, and the two stacked optical fiber lobes are first detected. The beam combining end of the optical fiber bundle to be tested is imaged by an image acquisition device to generate an image of the beam combining end, and the beam combining end has two symmetrically distributed optical fiber lobes. The beam combining end image is first binarized according to a preset grayscale threshold of the lobed end face, and the optical fiber lobed end face is segmented on the beam combining end image. That is, the other parts except the two optical fiber lobes are removed, because the other beam splitting ends are blocked, so the optical fiber lobes corresponding to these blocked beam splitting ends on the beam combining end are not lit, and the background area and other unlit optical fiber lobes are effectively removed by binarization through the preset grayscale threshold of the lobed end face. Next, the fiber flap end face on the beam-combining end image is subjected to noise removal processing in the non-fiber area. The noise at the edge of the fiber bundle is removed. The fiber flap end face on the beam-combining end image is subjected to a second binarization processing according to the preset fiber grayscale threshold. Each fiber in the fiber flap is initially visualized. The fiber flap end face on the beam-combining end image after the second binarization processing is subjected to morphological corrosion processing to disconnect the connection between the fibers and remove the edge burrs, and then the area screening processing is performed to extract the area of ​​each fiber in the fiber flap end face. Dark defect analysis and abnormal number analysis are performed on the area of ​​each fiber to generate the analysis results of the flap fiber at the beam-combining end. Not only can the number of abnormal fibers in the two symmetrical fiber flaps be determined, but also whether the uniformity of the symmetrical fiber flap is qualified can be determined, so as to realize the identification of the number of flaps on the fiber bundle end face, and the optical fiber transmittance can be judged. Finally, each flap of the fiber bundle can be accurately counted and the optical fiber with poor transmittance can be detected, thereby improving the accuracy of fiber bundle defect detection.

[0116] See also Figure 2 The present application provides an embodiment of a method for noise removal processing of a non-fiber region of a fiber split end face, comprising:

[0117] 201. Performing optical fiber feature screening on the optical fiber splitting end face to determine an optical fiber feature region, where the optical fiber feature region is a region where only the optical fiber splitting is retained;

[0118] 202. The fiber split end face on the beam combining end image is cropped according to the fiber feature region to remove the noise region generated during the binarization process.

[0119] In this embodiment, since the area of ​​the fiber lobes is the largest after binary segmentation, feature screening is performed based on the area of ​​the segmented area, and each lobe in the fiber bundle is segmented. At the same time, the convex hull (sector-shaped area) of the segmented shape is calculated, that is, the minimum convex polygon of the shape (minimum sector-shaped area) is calculated, and then the fiber lobes on the fiber end face image are cropped based on the minimum convex polygon area. Please see Fig.11 and Fig.12 , Fig.11 This is the image before fiber feature screening (area screening). Fig.12 It is the image after noise removal, which is also the minimum convex polygon image and the cropped image.

[0120] See also Figure 3 The present application provides an embodiment of a method for analyzing uniformity of an optical fiber bundle, comprising:

[0121] 301. Use a shading strip to re-block the splitting end, retain only one splitting end each time, take an image of the beam combining end to perform defect detection of the corresponding optical fiber petal, and generate N split optical fiber analysis results in total;

[0122] 302. Perform fiber bundle uniformity analysis on the N split optical fiber analysis results to generate a uniformity analysis result.

[0123] In this embodiment, the terminal uses a shading strip to re-shield the beam splitting end, retaining only one beam splitting end at a time, and captures the image of the beam combining end to perform defect detection on the corresponding optical fiber petals, generating a total of N petal optical fiber analysis results, that is, detecting the two symmetrical optical fiber petals formed at each beam splitting end. According to the analysis structure, the uniformity of the overall light transmittance of the optical fiber bundle can be analyzed.

[0124] See also Figure 4 The present application provides an embodiment of a method for filtering a beam-combining end image, comprising:

[0125] 401. Calculate the attention matrix for each pixel on the combined end image;

[0126] After the optical fiber is set to be equally split, and each split end has a symmetrical fan-shaped area, the uniformity of the optical fiber's light transmission can be well detected, but special noise will be generated during the detection process. When the optical fiber bundle to be tested is placed on the light-homogenizing plate and the other split ends are shielded by the shading strip, the shielding may not be tight, resulting in some optical fibers at some split ends being able to transmit some light. If such optical fibers happen to be located at the edge of the lit optical fiber split at the combined end, the subsequent positioning of the optical fiber split will be inaccurate. Specifically, when clearing noise in non-optical areas, the selected fan-shaped area will be slightly enlarged, and the part that does not originally belong to the optical fiber split will be increased, resulting in inaccurate subsequent detection. Therefore, it is necessary to remove the noise generated by such edge optical fibers that are not completely shielded.

[0127] In this embodiment, because the optical fiber brightness of the light-transmitting optical fiber obtained by the loose shielding generally does not exceed the optical fiber brightness of the normal optical fiber on the lit optical fiber petal, the terminal can construct a filter kernel for filtering through the current beam-combined end image. First, the terminal calculates the attention matrix for each pixel point on the beam-combined end image, generates a corresponding attention value for the pixel point, and then generates an attention matrix.

[0128] 402. Roughly divide the optical fiber region of the beam-combining end image according to the attention value of the attention matrix to determine two petal regions;

[0129] The terminal selects pixels whose attention is greater than a preset threshold, and fits the fan-shaped area from these pixels through the edge points. Each petal area forms two fitted straight lines and one fitted arc. The petal area is generated based on these fitted lines and their intersection points.

[0130] 403. Determine the optical fiber center coordinates of the beam combining end image;

[0131] The terminal determines the fiber center coordinates of the beam combining end image, where the fiber center coordinates refer to the positions of the symmetrical points of the two fiber splitting points.

[0132] 404. Calculate the standard deviation and grayscale mean corresponding to the two petal regions in the beam-combined end image;

[0133] The terminal first calculates the standard deviation and grayscale mean based on the pixel points corresponding to the two petal areas in the combined end image. The standard deviation of the first petal area is , the gray mean is , the standard deviation of the first petal area is , the gray mean is .

[0134] 405. Calculate the convolution kernel adjustment coefficient according to the grayscale mean;

[0135] Next, the terminal calculates the grayscale mean and Calculate the convolution kernel adjustment coefficient. Specifically, the terminal prepares a reference fiber bundle with the same specifications as the fiber bundle to be tested in advance, removes multiple split ends in the reference fiber bundle, and leaves only one split end, which is placed on a light-diffusing plate. When the light-diffusing plate parameters and other environmental parameters and fiber parameters are the same, the combined end has two illuminated and symmetrical fiber lobes. Take an image, and calculate the reference grayscale mean through all the pixels in the two reference split fiber areas. The convolution kernel adjustment coefficient is the grayscale mean. and Divide by the reference grayscale mean , the reference grayscale mean is the grayscale mean of the entire pixel points in the two petal areas, and the convolution kernel adjustment coefficient of the two areas, that is, the grayscale weight, is obtained. The formula is as follows:

[0136]

[0137]

[0138] 406. Calculate the petal-splitting optical fiber filter kernel according to the optical fiber center coordinates, the standard deviation, the convolution kernel adjustment coefficient and the preset convolution kernel function;

[0139] Next, the terminal calculates the center coordinates of the optical fiber. , standard deviation and , convolution kernel adjustment coefficient and , the preset convolution kernel function is to calculate the fiber filter kernel for the two fiber lobes respectively and The formula is as follows:

[0140]

[0141]

[0142] in, is the pixel coordinate of the first petal area, The pixel coordinates of the first petal area.

[0143] Next, the two convolution kernels are integrated to generate the final split-fiber filter kernel.

[0144]

[0145] in, is the final split fiber filter core, is the reference area of ​​the petal area, which can be calculated in advance. is the actual area of ​​the first petal region, is the actual area of ​​the second petal area, that is, the area weight of each petal area is calculated.

[0146] 407. Perform filtering processing on the beam combining end image by using the split fiber filter core.

[0147] Finally, the terminal filters the entire beam-combined end image through the final split fiber filter kernel, which can effectively remove the edge fiber noise caused by loose occlusion. In this embodiment, the split area is determined from the two symmetrically distributed fiber splits according to the attention matrix, and then the fiber center coordinates are calculated. Then, the respective split fiber filter kernel convolution kernels are calculated according to the standard variance and grayscale mean in the split area, and then integrated to form a final split fiber filter kernel, and then the beam-combined end image is filtered.

[0148] See also Figure 5 The present application provides an embodiment of a method for determining the optical fiber center coordinates of a beam combining end image, comprising:

[0149] 501. Acquire a photographic image of the optical fiber bundle to be tested which is not shielded by the shading strip;

[0150] The terminal first obtains a photographic image of the optical fiber bundle to be tested that is not shielded by the shading strip, that is, an image that is lit as a whole.

[0151] 502. Determine the circular optical fiber profile of the captured image according to an edge inspection algorithm, and calibrate the center coordinates of the circular optical fiber profile;

[0152] After the terminal obtains a photographed image of the optical fiber bundle to be tested that is not shielded by the shading strip, the terminal determines the circular optical fiber contour of the photographed image according to the edge inspection algorithm. The contour is a circle, and the center coordinates are calibrated by the circle.

[0153] 503. Determine the optical fiber center coordinates of the beam combining end image according to the optical fiber profile of the captured image and the calibrated center coordinates.

[0154] The terminal scales and matches the fiber profile of the captured image with the image at the beam-combining end, and marks the center coordinates on the beam-combining end image through the calibrated center coordinates. The calculation method of the center coordinates is relatively simple, the amount of calculation is relatively small, but the accuracy is also high.

[0155] See also Figure 6 The present application provides another embodiment of a method for determining the optical fiber center coordinates of a beam combining end image, comprising:

[0156] 601. Determine the coordinates of the four corner pixel points of the two petal areas on the beam-combined end image, and the coordinates of the two tips of the two petal areas with the shortest distance;

[0157] The terminal first determines the pixel coordinates of the four corner points of the two petal areas on the beam-combined end image, as well as the coordinates of the two tips with the shortest distance between the two petal areas. Because the terminal fits the fan-shaped area through the edge points, each petal area forms two fitted straight lines and one fitted arc. The petal area is generated based on these fitted lines and their intersections. At this time, each petal area (sector) has three points, two corner points and one tip coordinate point.

[0158] 602. Cross-connect the pixel coordinates of the four corner points to generate the first center coordinates;

[0159] The terminal first cross-connects the coordinates of the four corner pixel points to generate the first center coordinates, because the four corner pixel points may have edge noise, which is the edge area expansion caused by loose occlusion, and its accuracy is low. Therefore, the next step of adjustment is required.

[0160] 603. Take the center point of the line connecting the two tip coordinates as the second center coordinate;

[0161] At this time, the terminal takes the center point of the line connecting the two tip coordinates as the second center coordinate, and the tip coordinate is less affected by edge noise.

[0162] 604. Take the center point of the line connecting the first center coordinate and the second center coordinate as the center coordinate of the optical fiber.

[0163] Although the tip coordinates are less affected by edge noise. However, if the tip coordinates are detected with deviations, the second center coordinates will also be offset. Therefore, the second center coordinates and the second center coordinates need to be adjusted. Specifically, the terminal takes the center point of the line connecting the first center coordinate and the second center coordinate as the optical fiber center coordinate. This method performs center calculation based on the actual petal area, and can determine the center point according to the actual situation of the optical fiber to be tested. Because the size and shape of each optical fiber petal of the optical fiber to be tested are not very perfect, and the symmetry is not very perfect, there may be a situation where the optical fiber center coordinates and the contour center are unknown, so this method can deal with this problem well.

[0164] See also Figure 7 The present application provides an embodiment of a device for detecting optical fiber bundle defects, comprising:

[0165] The preprocessing unit 701 is used to place the split ends of the optical fiber bundle to be tested on the light homogenizing plate, and use the shading strip to block part of the split ends, leaving only one split end, and the optical fiber bundle to be tested includes N split ends and one combined end, where N is an integer greater than 2;

[0166] The first generating unit 702 is used to collect images of the combined end of the optical fiber bundle to be tested through an image acquisition device to generate an image of the combined end, where the combined end has two symmetrically distributed optical fiber petals;

[0167] A first calculation unit 703 is used to calculate an attention matrix for each pixel point on the combined end image;

[0168] A first determining unit 704 is used to roughly divide the optical fiber area of ​​the beam combining end image according to the attention value of the attention matrix to determine two petal areas;

[0169] The second determining unit 705 is used to determine the optical fiber center coordinates of the beam combining end image;

[0170] Optionally, the second determining unit 705 includes:

[0171] Acquire a photographic image of the optical fiber bundle to be tested that is not shielded by the light shielding strip;

[0172] Determine the circular fiber profile of the captured image according to the edge inspection algorithm, and calibrate the center coordinates of the circular fiber profile;

[0173] The optical fiber center coordinates of the beam combining end image are determined according to the optical fiber profile of the captured image and the calibrated center coordinates.

[0174] Optionally, the second determining unit 705 includes:

[0175] Determine the pixel coordinates of the four corner points of the two petal areas on the beam-combined end image, as well as the coordinates of the two tips with the shortest distance between the two petal areas;

[0176] Cross-connect the pixel coordinates of the four corner points to generate the first center coordinates;

[0177] Take the center point of the line connecting the two tip coordinates as the second center coordinate;

[0178] The center point of the line connecting the first center coordinate and the second center coordinate is taken as the center coordinate of the optical fiber.

[0179] The second calculation unit 706 is used to calculate the standard deviation and grayscale mean corresponding to the two petal regions in the beam-combined end image;

[0180] The third calculation unit 707 is used to calculate the convolution kernel adjustment coefficient according to the grayscale mean;

[0181] The fourth calculation unit 708 is used to calculate the petal-split optical fiber filter kernel according to the optical fiber center coordinates, the standard deviation, the convolution kernel adjustment coefficient and the preset convolution kernel function;

[0182] A filtering unit 709, configured to filter the image at the combined end through a split fiber filter core;

[0183] The segmentation unit 710 is used to perform a first binarization process on the beam-combined end image according to a preset split end face grayscale threshold, and segment the fiber split end faces on the beam-combined end image;

[0184] A denoising unit 711 performs noise removal processing on the non-fiber region of the fiber split end face on the beam combining end image;

[0185] Optionally, the denoising unit 711 includes:

[0186] Performing optical fiber feature screening on the optical fiber splitting end face to determine an optical fiber feature area, where the optical fiber feature area is an area where only the optical fiber splitting is retained;

[0187] The fiber split end face on the beam combining end image is cropped according to the fiber feature area to remove the noise area generated during the binarization process.

[0188] A binarization unit 712, configured to perform a second binarization process on the fiber split end face on the beam combining end image according to a preset fiber grayscale threshold;

[0189] The extraction unit 713 is used to perform morphological corrosion processing and area screening processing on the fiber split end face on the beam combining end image after the second binarization processing, so as to extract the area of ​​each optical fiber in the fiber split end face;

[0190] The second generating unit 714 is used to perform dark defect analysis and abnormal quantity analysis on the area of ​​each optical fiber, and generate the analysis result of the split optical fiber at the combined end;

[0191] The third generating unit 715 uses a shading strip to re-shield the splitting end, retaining only one splitting end each time, and taking an image of the beam combining end to perform defect detection of the corresponding optical fiber petal, thereby generating N split optical fiber analysis results in total;

[0192] The fourth generating unit 716 is used to perform fiber bundle uniformity analysis on the N split optical fiber analysis results to generate a uniformity analysis result.

[0193] See also Figure 8 , the present application provides a system for detecting optical fiber bundle defects, comprising:

[0194] Processor 801 , memory 802 , input-output unit 803 , and bus 804 .

[0195] The processor 801 is connected to the memory 802 , the input and output unit 803 , and the bus 804 .

[0196] The memory 802 stores a program, and the processor 801 calls the program to execute the following steps: Figure 1 , Figure 2 and Figure 3 , Figure 4 , Figure 5 and Figure 6 The method in .

[0197] The present application provides a computer-readable storage medium, on which a program is stored, and when the program is executed on a computer, the program performs the following steps: Figure 1 , Figure 2 and Figure 3 , Figure 4 , Figure 5 and Figure 6 The method in .

[0198] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0199] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0200] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0201] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0202] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, read-only memory), random access memory (RAM, random access memory), disk or optical disk and other media that can store program code.

Claims

1. A method for detecting optical fiber bundle defects, characterized in that: include: The split ends of the optical fiber bundle to be tested are placed on a light homogenizing plate, and a light shielding strip is used to block part of the split ends, leaving only one split end, the optical fiber bundle to be tested includes N split ends and one combined end, where N is an integer greater than 2; Capturing an image of the combined end of the optical fiber bundle to be tested by an image acquisition device to generate an image of the combined end, wherein the combined end has two symmetrically distributed optical fiber petals; Calculating an attention matrix for each pixel on the combined end image; According to the attention value of the attention matrix, the optical fiber region of the combined end image is roughly divided to determine two petal regions; Determining the optical fiber center coordinates of the beam combining end image; Calculating the standard deviation and grayscale mean corresponding to the two petal regions in the combined end image; Calculate the convolution kernel adjustment coefficient according to the grayscale mean; Calculating a petal-split optical fiber filter kernel according to the optical fiber center coordinates, the standard deviation, the convolution kernel adjustment coefficient and a preset convolution kernel function; Performing filtering processing on the beam-combining end image by using the split-fiber optical fiber filter core; Performing a first binarization process on the beam-combined end image according to a preset split end face grayscale threshold, segmenting the fiber split end faces on the beam-combined end image; Performing noise removal processing on the non-fiber region of the fiber split end face on the beam combining end image; Performing a second binarization process on the fiber split end face on the beam combining end image according to a preset fiber grayscale threshold; Applying morphological corrosion processing and area screening processing to the fiber split end face on the beam combining end image after the second binarization processing to extract the area of ​​each optical fiber in the fiber split end face; A dark defect analysis and an abnormal quantity analysis are performed on the region of each optical fiber to generate an analysis result of the split optical fiber at the combined end.

2. The method according to claim 1, characterized in that The noise removal process of the non-fiber region is performed on the fiber split end face on the beam combining end image, including: Performing optical fiber feature screening on the optical fiber splitting end face to determine an optical fiber feature area, wherein the optical fiber feature area is an area where only the optical fiber splitting is retained; The fiber split end face on the beam combining end image is cropped according to the fiber feature region to remove the noise region generated during the binarization process.

3. The method according to claim 1, characterized in that After performing dark defect analysis and abnormal quantity analysis on the region of each optical fiber to generate a first split optical fiber analysis result, the method further includes: Use the shading strip to re-block the split end, retaining only one split end each time, and take the image of the combined end to perform defect detection on the corresponding fiber split, generating a total of N split fiber analysis results; Perform fiber bundle uniformity analysis on the N split optical fiber analysis results to generate uniformity analysis results.

4. The method according to claim 1, characterized in that: Determining the optical fiber center coordinates of the beam combining end image includes: Acquire a photographic image of the optical fiber bundle to be tested that is not shielded by the shading strip; Determine the circular optical fiber profile of the captured image according to an edge inspection algorithm, and calibrate the center coordinates of the circular optical fiber profile; The optical fiber center coordinates of the beam combining end image are determined according to the optical fiber profile of the captured image and the calibrated center coordinates.

5. The method according to claim 1, characterized in that Determining the optical fiber center coordinates of the beam combining end image includes: Determine the pixel coordinates of four corner points of the two petal areas on the combined end image, and the coordinates of two tips with the shortest distance between the two petal areas; Cross-connect the pixel coordinates of the four corner points to generate the first center coordinates; Take the center point of the line connecting the two tip coordinates as the second center coordinate; The center point of the line connecting the first center coordinate and the second center coordinate is taken as the center coordinate of the optical fiber.

6. A device for detecting optical fiber bundle defects, characterized in that: include: A preprocessing unit is used to place the split ends of the optical fiber bundle to be tested on the light homogenizing plate, and use the shading strip to block part of the split ends, leaving only one split end, the optical fiber bundle to be tested includes N split ends and one combined end, N is an integer greater than 2; A first generating unit is used to collect images of the combined end of the optical fiber bundle to be tested by an image acquisition device to generate an image of the combined end, wherein the combined end has two symmetrically distributed optical fiber petals; A first calculation unit is used to calculate the attention matrix for each pixel point on the combined end image; A first determination unit is used to roughly divide the optical fiber area of ​​the beam combining end image according to the size of the attention value of the attention matrix to determine two petal areas; A second determining unit, used to determine the optical fiber center coordinates of the beam combining end image; The second calculation unit is used to calculate the standard deviation and grayscale mean corresponding to the two petal areas in the beam-combined end image; A third calculation unit is used to calculate a convolution kernel adjustment coefficient according to the grayscale mean; A fourth calculation unit, used for calculating the petal-split optical fiber filter kernel according to the optical fiber center coordinates, the standard deviation, the convolution kernel adjustment coefficient and the preset convolution kernel function; A filtering unit, used for filtering the image at the beam combining end through a split fiber filter core; A segmentation unit, used for performing a first binarization process on the beam-combined end image according to a preset split end face grayscale threshold, and segmenting the optical fiber split end face on the beam-combined end image; A denoising unit, for performing noise removal processing on the non-fiber region of the fiber split end face on the beam combining end image; A binarization unit, used for performing a second binarization process on the fiber split end face on the beam combining end image according to a preset fiber grayscale threshold; An extraction unit, used for performing morphological corrosion processing and area screening processing on the fiber split end face on the beam combining end image after the second binarization processing, so as to extract the area of ​​each optical fiber in the fiber split end face; The second generating unit is used to perform dark defect analysis and abnormal quantity analysis on the area of ​​each optical fiber to generate the analysis result of the split optical fiber at the combined end.

7. The device according to claim 6, characterized in that Denoising unit, including: Performing optical fiber feature screening on the optical fiber splitting end face to determine an optical fiber feature area, wherein the optical fiber feature area is an area where only the optical fiber splitting is retained; The fiber split end face on the beam combining end image is cropped according to the fiber feature region to remove the noise region generated during the binarization process.

8. A system for detecting optical fiber bundle defects, characterized in that: include: Processor, memory, input-output unit, and bus; The processor is connected to the memory, the input and output unit, and the bus; The memory stores a program, and the processor calls the program to execute the method according to any one of claims 1 to 5.

9. A computer-readable storage medium having a program stored thereon, wherein the program, when executed on a computer, performs the method according to any one of claims 1 to 5.

Images