A method and system for automatic polarity detection for MT-FA

Abstract

The present invention discloses a method for automatic polarity detection of MT-FA, including: S1: sequentially emitting optical signals through a laser light source module at the MT end, and the optical signals are transmitted from the MT end to the FA end through an optical fiber; S2: the dual cameras at the FA end perform real-time image capture; S3: an image processing module analyzes the image data collected by the dual cameras; S4: determines whether the polarity at the FA end is correct; S5: when it is determined that the polarity is incorrect, prompts "polarity error", and extracts and reports the error information points. The present invention adopts a detection method of dual cameras and multi-feature fusion, which not only comprehensively captures the light output situation of the optical fiber from different angles, but also comprehensively analyzes various features such as the light intensity, centroid, and gradient of the optical signal, making up for the limitations of the existing single-camera and single-feature detection methods. The automatic detection can ensure stable and consistent detection results under various operating conditions, further improving the reliability of the system.

Description

Technical Field

[0001] The present invention relates to the technical field of MT-FA polarity testing, and in particular to a method and system for automatic polarity detection of MT-FA. Background Art

[0002] When detecting the polarity of optical fiber components in the field of optical communication, the existing test schemes usually emit light from the MT end to the FA end after the products are docked, and then manually check to determine whether the light sequence is correct. However, in the case where the fiber spacing at the FA end is very small (such as 127 μm), manual detection is prone to misjudgment or missed detection. This not only reduces the accuracy of detection, but may also lead to signal transmission errors in actual applications, seriously affecting the communication quality. In addition, the operation efficiency of manual detection is low. The existing manual detection efficiency is about 120 pcs / H, which is difficult to meet the requirements of high-efficiency production in a large-scale production environment. During the manual detection process, due to factors such as human subjective judgment and non-standard operation, it is easy to cause misjudgment, and there is a lack of an effective anti-fooling mechanism, further increasing the risk of errors. Currently, some automated detection methods usually use a single camera or a single feature for polarity detection, and these methods also have obvious limitations, such as a single perspective, limited information, and a single feature being easily affected, making it difficult to cope with a complex optical signal environment. Summary of the Invention

[0003] In view of this, the present invention provides a method and system for automatic polarity detection of MT-FA, aiming to solve the problems of easy misjudgment, missed detection, and low efficiency in manual detection in the prior art.

[0004] To achieve the above object, a method and system for automatic polarity detection of MT-FA provided by the present invention includes the following steps:

[0005] S1: The laser light source module at the MT end sequentially emits optical signals, and the optical signals are transmitted from the MT end to the FA end through optical fibers;

[0006] S2: The dual cameras at the FA end perform real-time image capture;

[0007] S3: The image processing module analyzes the image data collected by the dual cameras;

[0008] S4: Determine whether the polarity at the FA end is correct;

[0009] S5: When it is determined that the polarity is incorrect, prompt "Polarity Error", and extract and report the error information points.

[0010] As a further improvement method of the present invention:

[0011] Optionally, in step S1, the laser light source module at the MT end sequentially emits optical signals, and the optical signals are transmitted from the MT end to the FA end through optical fibers, including:

[0012] The laser light source module at the MT end sequentially emits optical signals according to the emission sequence and time interval preset by the control unit. The emission sequence is set according to the arrangement sequence of the optical fiber channels to ensure that each optical fiber channel can emit optical signals in sequence. The optical signals are transmitted through the optical fibers and gradually transmitted to the FA end along each optical fiber channel at the MT end. Each beam of optical signal is transmitted from the MT end to the FA end through the optical fiber path corresponding to the optical fiber channel.

[0013] In step S1, through the emission sequence and time interval preset by the control unit, it is ensured that the optical signals are emitted sequentially according to the optical fiber channel sequence, avoiding signal interference and confusion. Through precise time interval control, the system can effectively manage the transmission of optical signals, ensure consistent timing, and ensure that the signals of each optical fiber channel are recognizable. During the transmission process, the optical signals are transmitted along their respective optical fiber paths, avoiding path confusion and ensuring the correct polarity between the MT end and the FA end. Sequentially emitting signals helps subsequent image processing and polarity detection, simplifies the analysis process, and improves the accuracy and reliability of detection.

[0014] Optionally, in step S2, the dual cameras at the FA end perform real-time image capture, including:

[0015] The two cameras are respectively placed at angular positions of 8 degrees and 42.5 degrees to capture the light emission situation at the optical fiber end of the FA end in real time. Camera 1 is located at the 8-degree angular position to capture the front optical signal of the optical fiber end, and camera 2 is located at the 42.5-degree angular position to capture the light emission state of the optical fiber from the side.

[0016] This dual-view capture method in step S2 can more accurately reflect the light emission characteristics of the optical fiber, improving the accuracy and robustness of polarity determination.

[0017] Optionally, in step S3, image processing analyzes the image data collected by the cameras, including:

[0018] S31: The image captured by camera 1 is filtered by Gaussian filter to obtain image P1;

[0019] S32: The image captured by camera 2 is filtered by median filter to obtain image P2;

[0020] S33: Canny edge detection is applied to both image P1 and P2 to obtain edge images B1 and B2 respectively;

[0021] S34: The edge images B1 and B2 are synthesized to obtain a comprehensive edge image B combined 。

[0022] Step S3 integrates a variety of image processing techniques to ensure high-quality input data in fiber polarity determination. The images of Camera 1 and Camera 2 are processed by Gaussian filtering and median filtering respectively to reduce noise while retaining edge details. Then, Canny edge detection is applied to accurately extract edge information. Finally, the two edge images are synthesized to generate the comprehensive edge image B. combined This multi-step processing method not only improves the robustness of the algorithm but also optimizes the overall processing flow, thus significantly enhancing the reliability and accuracy of the system in practical applications.

[0023] Optionally, in step S34, the edge images B1 and B2 are synthesized to obtain the comprehensive edge image B, combined including:

[0024] S341: Calculate the average optical signal intensities G1 and G2 of the edge images B1 and B2 respectively and assign weights, specifically:

[0025]

[0026] where N1 and N2 are the total number of pixels in the edge images respectively; and are the gray values of the edge images B1 and B2 at the pixel (x, y) respectively; and are the weights of B1 and B2 based on the optical signal intensity respectively.

[0027] S3411: Calculate the centroid positions Q1 and Q2 of the edge images B1 and B2 respectively, specifically:

[0028]

[0029] The x coordinate of the centroid position is the sum of the products of the coordinate x of each pixel position (x, y) and the gray value of the edge image at that position, and then divided by the sum of all pixel gray values; the y coordinate is the sum of the products of the coordinate y of each pixel position (x, y) and the gray value of the edge image at that position, and then divided by the sum of all pixel gray values.

[0030] S3412: Calculate the position deviation ΔQ according to the centroid positions Q1 and Q2, specifically:

[0031]

[0032] Step S341 provides a comprehensive and accurate basis for polarity determination by calculating the optical signal intensity of each edge image and assigning weights according to its distribution, and at the same time determining the centroid position. This comprehensive method not only improves the accuracy of polarity determination but also enhances the adaptability of the system to different optical signal conditions.

[0033] S342: Calculate the mean of the gradient magnitudes of edge images B1 and B2, and assign weights, specifically:

[0034]

[0035] Where and are the gray values of edge images B1 and B2 at pixel (x + m, y + n) respectively; H x1 (x, y) and H x2 (x, y) are the horizontal gradients, H y1 (x, y) or H y2 (x, y) are the vertical gradients, which are obtained through convolution operations with Sobel convolution kernels K x (m, n) and K y (m, n) respectively, where m and n are the index ranges of the convolution kernels (taking values -1, 0, 1); the gradient magnitude of each pixel is obtained by calculating the square root of the sum of the squares of the horizontal and vertical gradients, and then, the average of the gradient magnitudes of all pixels is calculated to obtain H1 and H2 respectively; and are the weights of B1 and B2 based on the gradient magnitudes respectively.

[0036] S3421: Calculate the average directions θ1 and θ2 of edge images B1 and B2, specifically:

[0037]

[0038] The average directions θ1 and θ2 are calculated by averaging the gradient directions of all pixels. The gradient direction of each pixel is calculated by the ratio of its vertical gradient H y1 (x, y) or H y2 (x, y) and the horizontal gradient H x1 (x, y) or H x2 (x, y) and using the arctangent function;

[0039] S3422: Calculate the direction difference Δθ according to the average directions θ1 and θ2, specifically:

[0040] Δθ = θ1 - θ2

[0041] Step S342 provides rich directional and gradient features for polarity determination through the analysis of gradient magnitudes, average directions, and direction differences. By integrating the information from two perspectives, accurately calculating and comparing gradient-related attributes further enhances the robustness of the determination process.

[0042] S343: Based on the signal strength and the average gradient, obtain the comprehensive weights W1 and W2 of the edge images B1 and B2 respectively, specifically as follows:

[0043]

[0044] Step S343 fuses the signal strength and the gradient information together, assigns weights to each edge image, makes up for the deficiencies of single features, and improves the accuracy of polarity determination.

[0045] S344: Use non-linear fusion of edge information to obtain the comprehensive edge image B combined , specifically as follows:

[0046]

[0047] For each pixel, calculate the product of and , and then take the maximum value of these two values; This fusion method ensures that the most significant edge information is highlighted in the comprehensive edge image;

[0048] S3441: Calculate the centroid position Q combined of the edge image B combined , specifically as follows:

[0049]

[0050] The x coordinate of the centroid position Q combined is the sum of the products of the coordinate x of each pixel position (x, y) and the gray value B combined (x, y) of the comprehensive edge image at this position, and then divided by the sum of all pixel gray values; The y coordinate is the sum of the products of the coordinate y of each pixel position (x, y) and the gray value B combined (x, y) of the comprehensive edge image at this position, and then divided by the sum of all pixel gray values.

[0051] S3442: Calculate the average direction θ combined of the edge image B combined , specifically as follows:

[0052]

[0053] For each pixel (x, y), by calculating the ratio of its vertical gradient and horizontal gradient , which is partial differential calculation, and use the arctangent function to determine the gradient direction of this pixel. Average the gradient directions of all pixels, and N in the formula combinedIndicates the total number of pixels in the comprehensive edge image. The directions of all pixels are averaged to obtain the average direction θ of the entire comprehensive edge image combined .

[0054] Step S344 provides a comprehensive and accurate analysis basis for fiber polarity determination by means of non - linear fusion, extraction of centroid position and direction information. The non - linear fusion method ensures the quality and representativeness of the edge image, while the centroid and direction information of the comprehensive edge image provide key global features for polarity determination, thus significantly improving the accuracy and robustness of the determination.

[0055] Optionally, determining whether the polarity of the FA end is correct in step S4 includes:

[0056] S41: Using the centroid position Q combined of the edge image B combined to perform similarity detection with the centroid positions Q1 and Q2 of the edge images B1 and B2. Specifically:

[0057]

[0058] Calculate the absolute difference between the centroid Q combined of the comprehensive edge image and the average value of the centroid positions of the edge images B1 and B2, and normalize it. The denominator for normalization is the larger absolute value among Q1 and Q2;

[0059] S42: Using the average direction θ combined of the edge image B combined to perform similarity detection with the average directions θ1 and θ2 in the edge images B1 and B2. Specifically:

[0060]

[0061] Calculate the absolute difference between the centroid θ combined of the comprehensive edge image and the average value of the average directions of the edge images B1 and B2, and normalize it. The denominator for normalization is the larger absolute value among the average directions θ1 and θ2;

[0062] S43: Calculate the total deviation Δ total :

[0063] △ total = W1·(△Q + △Q sim ) + W2·(△θ + △θ sim )

[0064] The total deviation is a weighted combination of the position and direction deviations. By integrating the deviations of position and direction into a total quantity, the polarity state of the optical fiber can be evaluated more comprehensively.

[0065] S44: Utilize the total deviation Δ total Compare it with a set threshold value to determine the polarity. Specifically:

[0066] If Δ total is less than a threshold value T, the polarity is correct; otherwise, the polarity is incorrect. The setting of the threshold value T needs to be based on a large amount of experimental data;

[0067] Step S4 greatly improves the accuracy and efficiency of determination by combining the comparison of multiple image features, the comprehensive calculation of errors, and simple threshold determination. By detecting the centroid and direction similarity between the comprehensive edge image and the individual edge image, and then calculating the total deviation and comparing it with the set threshold value, the polarity of the optical fiber can be effectively determined whether it is correct.

[0068] Optionally, when it is determined that the polarity is incorrect in step S5, prompt "Polarity incorrect" and provide the specific error information. Specifically:

[0069] Inform the user that a polarity error is detected, and display the specific parameters causing the error, such as the position deviation ΔQ, the direction difference Δθ and other values, to help the user understand the problem;

[0070] Step S5 improves the transparency, professionalism and user experience of the system by prompting "Polarity incorrect" and providing specific parameter information, enabling the user to more effectively diagnose and correct problems, thereby ensuring the accuracy of the optical fiber polarity determination and the efficiency of the overall operation.

[0071] The present invention also provides an automatic polarity detection system for MT-FA, including:

[0072] Optical signal transmission module: Sequentially transmit optical signals at the MT end, and transmit the optical signals from the MT end to the FA end through the optical fiber;

[0073] Image acquisition module: Shoot the light output situation of the optical fiber at the FA end through two cameras at different angles;

[0074] Image processing module: Process the images captured by the cameras to obtain edge images and comprehensive edge images for polarity detection;

[0075] Feature extraction and analysis module: Extract the optical signal intensity, centroid position, and gradient features from the edge images and comprehensive edge images;

[0076] Polarity determination module: Determine whether the polarity at the FA end is correct;

[0077] Error prompt and information feedback module: When the polarity determination is incorrect, give a prompt and provide error information.

[0078] The beneficial effects of the present invention are:

[0079] By introducing an automated detection system, the present invention uses cameras and image processing algorithms to replace traditional manual inspection methods, overcoming the problems of misjudgment and missed detection caused by subjective judgment and non-standard operations in manual detection. Especially in the case where the fiber spacing at the FA end is very small (such as 127μm). Compared with the efficiency of about 120 pcs / H in traditional manual detection, the present invention improves the efficiency by 30% through the automated detection system, reaching 180 pcs / H. This efficiency improvement not only meets the requirements of mass production for rapid detection, but also reduces labor costs and optimizes the production process. The present invention adopts a detection method of dual cameras and multi-feature fusion, which not only comprehensively captures the light output of the optical fiber from different angles, but also comprehensively analyzes various features such as the light intensity, centroid, and gradient of the optical signal, making up for the limitations of existing single-camera and single-feature detection methods. In addition, an anti-fooling mechanism is built into the automated detection process, eliminating human operation errors. Different from the lack of an anti-fooling mechanism in manual detection, automated detection can ensure stable and consistent detection results under various operating conditions, further improving the reliability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 It is a flowchart of a method for automatic polarity detection of MT-FA according to the present invention.

[0081] Figure 2 It is a schematic diagram of a method for automatic polarity detection of MT-FA according to the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0082] The present invention will be further described below with reference to the accompanying drawings, but the present invention is not limited in any way. Any transformation or replacement based on the teachings of the present invention falls within the protection scope of the present invention.

[0083] Example 1: As Figure 1 and Figure 2 shown, the present invention provides a method for automatic polarity detection of MT-FA, including the following steps:

[0084] S1: The laser light source module at the MT end sequentially emits optical signals, and the optical signals are transmitted from the MT end to the FA end through the optical fiber; in this embodiment, the laser light source module emits light with a wavelength of 850 nm, the MT end is connected to the FA end through a 12-core MPO connector, and the fiber spacing at the FA end is 127μm; the laser light source module sequentially activates each channel according to a preset order, the output power is set to 5 mW, and the continuous emission time is 500 ms / channel.

[0085] S2: The dual cameras at the FA end perform real-time image capture:

[0086] The camera 1 is located at an 8-degree angle position to capture the front optical signal at the end of the optical fiber, and the camera 2 is located at a 42.5-degree angle position to capture the light output state of the optical fiber from the side; the dual cameras in this embodiment both have a resolution of 1920×1080; the two cameras capture the optical signal in real time at a rate of 30 frames per second, and for each optical signal emission cycle, the cameras collect a set of image data.

[0087] S3: Image processing and analysis of the image data captured by the cameras:

[0088] S31: The image captured by camera 1 is processed using Gaussian filtering to obtain image P1; the filter kernel size of the Gaussian filtering in this embodiment is 5×5, and the standard deviation σ = 1.0;

[0089] S32: The image captured by camera 2 is processed using median filtering to obtain image P2; the filter kernel size of the median filtering in this embodiment is 3×3;

[0090] S33: Canny edge detection is applied to both image P1 and P2 to obtain edge images B1 and B2 respectively; the low threshold of the Canny edge detection in this embodiment is 50, and the high threshold is 150;

[0091] S34: The edge images B1 and B2 are synthesized to obtain a comprehensive edge image B combined ;

[0092] S341: Calculate the average optical signal intensities G1 and G2 of the edge images B1 and B2 respectively, and assign weights, specifically:

[0093]

[0094] where N1 and N2 are the total number of pixels in the edge images respectively; and are the gray values of the edge images B1 and B2 at the pixel (x, y) respectively; and are the weights of B1 and B2 based on the optical signal intensity respectively;

[0095] S3411: Calculate the centroid positions Q1 and Q2 of the edge images B1 and B2 respectively, specifically:

[0096]

[0097] S3412: Calculate the position deviation ΔQ according to the centroid positions Q1 and Q2, specifically:

[0098]

[0099] S342: Calculate the mean of the gradient amplitudes of the edge images B1 and B2, and assign weights, specifically:

[0100]

[0101] Among them, and are the gray values of the edge images B1 and B2 at the pixel (x + m, y + n) respectively; H x1 (x, y) and H x2 (x, y) are the gradients in the horizontal direction, H y1 (x, y) or H y2 (x, y) are the gradients in the vertical direction, and are obtained through convolution operations with the Sobel convolution kernels K x (m, n) and K y (m, n) respectively, where m and n are the index ranges of the convolution kernels (taking values -1, 0, 1); and are the weights of B1 and B2 based on the gradient magnitude respectively;

[0102] S3421: Calculate the average directions θ1 and θ2 of the edge images B1 and B2, specifically:

[0103]

[0104] S3422: Calculate the direction difference Δθ according to the average directions θ1 and θ2, specifically:

[0105] △θ = θ1 - θ2;

[0106] S343: Based on the signal strength and the average gradient, obtain the comprehensive weights W1 and W2 of the edge images B1 and B2 respectively, specifically:

[0107]

[0108] S344: Use non - linear fusion of edge information to obtain the comprehensive edge image B combined , specifically:

[0109]

[0110] S3441: Calculate the centroid position Q combined of the edge image B combined , specifically:

[0111]

[0112] S3442: Calculate the average direction θ combined of the edge image B combined , specifically:

[0113]

[0114] Among them, and are the vertical gradient and horizontal gradient of each pixel (x, y), is for partial differential calculation; N combined represents the total number of pixels in the comprehensive edge image.

[0115] S4: Determine whether the polarity of the FA end is correct:

[0116] S41: Use the centroid position Q combined of the edge image B combined to perform similarity detection with the centroid positions Q1 and Q2 of the edge images B1 and B2, specifically:

[0117]

[0118] S42: Use the average direction θ combined of the edge image B combined to perform similarity detection with the average directions θ1 and θ2 in the edge images B1 and B2, specifically:

[0119]

[0120] S43: Calculate the total deviation Δ total :

[0121] △ total = W1·(△Q + △Q sim ) + W2·(△θ + △θ sim )

[0122] S44: Use the total deviation Δ total to compare with the set threshold to determine the polarity, specifically:

[0123] If Δ total is less than a threshold T, the polarity is correct, otherwise the polarity is incorrect; the setting of the threshold T needs to be based on a large amount of experimental data;

[0124] S5: When the polarity is determined to be incorrect, prompt "Polarity error" and provide the specific error information, specifically:

[0125] Inform the user that a polarity error is detected, display the specific parameters causing the error, such as the position deviation ΔQ, direction difference Δθ, etc. values, to help the user understand the problem;

[0126] In this example, the number of effective edge pixels of B1 is 500, the average gray value is 120, the number of effective edge pixels of B2 is 480, and the average gray value is 110. Calculate the optical signal intensity weights W G1 ≈0.522 and W G2≈0.478; The centroid positions of B1 and B2 are Q1 = (150, 200) and Q2 = (155, 195) respectively, and the position deviation ΔQ≈7.07; The average gradients of the edge images B1 and B2 are H1 = 15 and H2 = 14, W H1 ≈0.517 and W H2 ≈0.483; The calculated average directions are θ1 = 30° and θ2 = 28°, then the direction difference Δθ = 2°; The comprehensive weights are W1 = 1.039 and W2 = 0.961; The centroid position Q combined of the comprehensive edge image B combined = (152, 197), the average direction θ combined = 29°, then ΔQ sin ≈0.003, Δθ sin ≈0, the total deviation Δ total = 9.28; Set the threshold T = 10, since 9.28 is less than 10, the polarity determination is correct;

[0127] Example 2:

[0128] The present invention also provides an automatic polarity detection system for MT-FA, including:

[0129] Optical signal transmission module: Sequentially transmit optical signals at the MT end, and transmit the optical signals from the MT end to the FA end through optical fibers;

[0130] Image acquisition module: Shoot the light output situation of the optical fiber at the FA end through 2 cameras at different angles;

[0131] Image processing module: Process the images captured by the cameras to obtain edge images and comprehensive edge images for polarity detection;

[0132] Feature extraction and analysis module: Extract features such as optical signal intensity, centroid position, and gradient from the edge images and comprehensive edge images to provide a basis for polarity determination;

[0133] Polarity determination module: Determine whether the polarity at the FA end is correct;

[0134] Error prompt and information feedback module: When the polarity determination is incorrect, prompt the user through the interface or other means and provide error information to facilitate the user to adjust or process.

[0135] Specifically, each module in the automatic polarity detection system for MT-FA in the embodiments of the present invention adopts the same technical means as those in the above-mentioned Figure 1 automatic polarity detection method for MT-FA described therein, and can produce the same technical effects, which will not be elaborated here.

[0136] It should be understood that the above embodiments are for illustrative purposes only and are not limiting to the scope of the patent application in terms of this structure.

[0137] It should be noted that the serial numbers of the above embodiments of the present invention are only for description and do not represent the superiority or inferiority of the embodiments. Moreover, the term "comprising", "including" or any other variant thereof in this article is intended to cover non-exclusive inclusion, so that a process, device, article or method comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, device, article or method. Without further limitation, an element defined by the statement "comprising one..." does not exclude the existence of additional identical elements in the process, device, article or method comprising that element.

[0138] Through the description of the above embodiments, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product is stored in a storage medium as described above (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to enable a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in various embodiments of the present invention.

[0139] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, are equally included in the patent protection scope of the present invention.

Claims

1. A method for automatic polarity detection of MT-FA, characterized in that: The following steps are involved: S1: The laser light source module at the MT end emits optical signals in sequence, and the optical signals are transmitted from the MT end to the FA end through optical fibers; S2: The dual cameras at the FA end take real-time images; Camera 1 is located at an angle of 8 degrees to capture the front light signal at the end of the optical fiber, and Camera 2 is located at an angle of 42.5 degrees to capture the light output state of the optical fiber from the side; S3: The image processing module analyzes the image data collected by the dual cameras; S4: Determine whether the polarity of the FA terminal is correct; S5: When a polarity error is determined, a "polarity error" prompt is displayed, and the error information points are extracted and reported.

2. The method for automatic polarity detection of MT-FA according to claim 1, characterized in that: The step S3 comprises: S31: The image captured by the camera 1 is subjected to Gaussian filtering to obtain an image P1; S32: The image captured by the camera 2 is subjected to median filtering to obtain an image P2; S33: applying Canny edge detection to the images P1 and P2 to obtain edge images B1 and B2 respectively; S34: Perform synthesis processing on the edge images B1 and B2 to obtain a comprehensive edge image B combined .

3. The method for automatic polarity detection of MT-FA according to claim 2, characterized in that: The step S34 includes: S341: Calculate the average light signal strengths G1 and G2 of the edge images B1 and B2 respectively, and assign weights, specifically: Among them, N1 and N2 are the total number of pixels in the edge image; and are the grayscale values ​​of edge images B1 and B2 at pixel (x, y) respectively; and are the weights of B1 and B2 based on the optical signal intensity; S3411: Calculate the centroid positions Q1 and Q2 of the edge images B1 and B2 respectively, specifically: S3412: Calculate the position deviation ΔQ based on the center of mass positions Q1 and Q2, specifically: S342: Calculate the mean values ​​H1 and H2 of the gradient amplitudes of the edge images B1 and B2, and assign weights, specifically: in, and are the grayscale values ​​of edge images B1 and B2 at pixel (x+m, y+n); H x1 (x,y) and H x2 (x, y) is the gradient in the horizontal direction, H y1 (x,y) or H y2 (x, y) is the vertical gradient, which is respectively convolved with the Sobel convolution kernel K x (m,n) and K y The convolution operation of (m,n) is performed, where m and n are the index ranges of the convolution kernel and their values ​​are -1, 0, and 1. and are the weights of B1 and B2 based on the gradient magnitude; S3421: Calculate the average directions θ1 and θ2 of the edge images B1 and B2, specifically: S3422: Calculate the direction difference Δθ according to the average directions θ1 and θ2, specifically: △θ=θ1-θ2 S343: Based on the signal strength and the gradient mean, the comprehensive weights W1 and W2 of the edge images B1 and B2 are obtained respectively, specifically: S344: Obtain a comprehensive edge image B by nonlinearly fusing edge information combined , specifically: S3441: Calculate edge image B combined The center of mass position Q combined , specifically: S3442: Calculate edge image B combined The average direction θ combined , specifically: in, and are the vertical and horizontal gradients of each pixel (x,y), is partial differential calculation; N combined Represents the total number of pixels in the integrated edge image.

4. The method for automatic polarity detection of MT-FA according to claim 3, characterized in that: The step S4 includes: S41: Using edge image B combined The center of mass position Q combined The similarity detection is performed with the centroid positions Q1 and Q2 of the edge images B1 and B2, specifically: S42: Using edge image B combined The average direction θ combined The similarity detection is performed with the average directions θ1 and θ2 in the edge images B1 and B2, specifically: S43: Calculate the total deviation Δ total : △ total =W1·(△Q+△Q sim )+W2·(△θ+△θ sim ) S44: Using the total deviation Δ total Compare with the set threshold to determine the polarity, specifically: If Δ total If it is less than a threshold value T, the polarity is correct, otherwise the polarity is wrong.

5. The method for automatic polarity detection of MT-FA according to claim 4, characterized in that: The step S5 includes: If a polarity error is detected, the specific parameters causing the error are displayed, including the position deviation ΔQ and the direction difference Δθ values.

6. A system for automatic polarity detection of MT-FA, characterized in that: include: Optical signal transmission module: emits optical signals in sequence at the MT end, and transmits the optical signals from the MT end to the FA end through optical fibers; Image acquisition module: uses two cameras at different angles to capture the light output of the optical fiber at the FA end; Image processing module: processes the image captured by the camera to obtain edge images and comprehensive edge images for polarity detection; Feature extraction and analysis module: extract light signal intensity, centroid position, and gradient features from edge images and integrated edge images; Polarity determination module: determines whether the polarity of the FA terminal is correct; Error prompt and information feedback module: when the polarity judgment is wrong, it will prompt and provide error information; To realize an automatic polarity detection method for MT-FA as described in any one of claims 1-5.

Images