Communication optical fiber management method, system and electronic equipment based on color recognition
Through convolutional neural network and optimized feature matrix extraction of sequence evolution constraint factors, the environmental dependence and granularity problems of traditional color recognition methods are solved, and higher precision color recognition and product quality control are achieved.
Patent Information
- Application Number
- CN202210778436.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2022-03-22
- Filing Date
- 2022-07-04
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2042-07-04
AI Technical Summary
Traditional color recognition methods are greatly affected by the external environment, and the detection granularity is rough, making it difficult to accurately identify weak color changes, which affects the accuracy of product quality inspection and control.
Convolutional neural network is used to extract the spectral graph feature matrix of incident and reflected light, and construct the loss function based on the constraint factor of sequence evolution, optimize the correlation extraction of the feature matrix, improve the constraints of the transfer matrix, thereby improving the classification accuracy of color recognition sex.
It improves the accuracy and detection accuracy of color recognition, can better check and control product quality, and enhances sensitivity to weak color changes.
Smart Images

Figure CN114937034B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of communication optical fibers, and more specifically, to a communication optical fiber management method, system and electronic equipment based on color recognition. Background Art
[0002] Communication optical fiber is a sensor that uses optical fiber as a functional material. Compared with other materials, optical fiber has many unique properties. Optical fiber itself is a sensitive component that has certain effects on many parameters. Its raw materials are abundant, cheap, small in size, light in weight, low in loss, excellent in optical characteristics, wide in bandwidth, and not subject to electromagnetic interference. Accordingly, optical fiber sensors have the advantages of high sensitivity, high precision, high speed, high density, and can realize non-contact and non-destructive detection. Color detection and recognition is an important means of product quality inspection and control.
[0003] Different colors of light can be expressed as different wavelengths, and color recognition is essentially wavelength recognition. The basic detection principle of fiber optic sensors is to use external factors to change the optical parameters of light propagating in the optical fiber. The changed light wave signal records the information of the parameter to be measured, so the parameter to be measured can be obtained by modulating and demodulating the light wave signal. The transmitting optical fiber emits a light signal to illuminate the object to be measured, and the receiving optical fiber collects the light signal reflected back by the object to be measured. The light intensity changes with the color of the surface of the object to be measured. Therefore, the change in the light intensity received by the optical fiber accurately reflects the change in color, which can achieve color detection and recognition.
[0004] Traditional color recognition uses a light source to directly illuminate an object and performs color recognition based on the relationship between incident light and reflected light. However, this method is greatly affected by the external environment. In addition, the traditional luminous flux of incident light and reflected light is not sensitive to slight color changes, that is, the detection granularity of the traditional statistical model is relatively coarse. Therefore, a communication fiber management solution based on color recognition is expected. Summary of the invention
[0005] In order to solve the above technical problems, the present application is proposed. The embodiments of the present application provide a communication optical fiber management method, system and electronic device based on color recognition, which extracts the characteristic distribution representation of the local characteristics of the spectral graph of the incident light and the reflected light in a high-dimensional space through a convolutional neural network to obtain a first characteristic matrix and a second characteristic matrix, and represents the change information between the spectral information of the incident light and the spectral information of the reflected light based on the transfer matrix between the two characteristic matrices, and further introduces a constraint factor based on sequence evolution to construct a loss function, so as to promote the constraint of the transformation features contained in the transfer matrix itself by optimizing the relevant feature extraction of the first characteristic matrix and the second characteristic matrix, thereby improving the accuracy of classification. In this way, the product quality can be better inspected and controlled based on color detection and recognition.
[0006] According to one aspect of the present application, a communication optical fiber management method based on color recognition is provided, which includes:
[0007] The training phase includes:
[0008] Obtaining a spectrum of incident light emitted by the communication optical fiber and a spectrum of reflected light reflected from the object under test;
[0009] Obtaining a first feature matrix from the spectrum of the incident light using a first convolutional neural network;
[0010] Obtaining a second feature matrix from the spectrum of the reflected light using a second convolutional neural network;
[0011] Calculating a transfer matrix between the first characteristic matrix and the second characteristic matrix, wherein the transfer matrix is used to represent correlation change information between spectral information of incident light and spectral information of reflected light;
[0012] Passing the transfer matrix through a classifier to obtain a classification loss function value;
[0013] Calculating a constraint factor based on sequence evolution between the first feature matrix and the second feature matrix, wherein the constraint factor based on sequence evolution includes a nuclear norm of a difference matrix between the first feature matrix and the second feature matrix and a negative of an absolute value of a logarithmic function of a quotient of a first category probability value and a second category probability value obtained by the first feature matrix and the second feature matrix respectively through the classifier;
[0014] Calculating a weighted sum between the classification loss function value and the sequence evolution-based constraint factor as a loss function value to train the first convolutional neural network and the second convolutional neural network; and
[0015] The inference phase includes:
[0016] Acquire a spectrum of incident light emitted by the communication optical fiber to be detected and a spectrum of reflected light reflected from the object to be detected;
[0017] Obtaining a first feature matrix from the spectrum of the incident light using the first convolutional neural network trained in the training phase;
[0018] Obtaining a second feature matrix from the spectrum of the reflected light using the second convolutional neural network trained in the training phase;
[0019] Calculating a transfer matrix between the first characteristic matrix and the second characteristic matrix, wherein the transfer matrix is used to represent correlation change information between spectral information of incident light and spectral information of reflected light;
[0020] The transfer matrix is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the quality of the communication optical fiber to be detected meets a preset standard.
[0021] According to another aspect of the present application, a communication optical fiber management system based on color recognition is provided, comprising:
[0022] Training modules include:
[0023] A spectrum acquisition unit, used to acquire a spectrum of incident light emitted by the communication optical fiber and a spectrum of reflected light reflected from the object to be measured;
[0024] A first feature extraction unit, configured to obtain a first feature matrix from the spectrum of the incident light obtained by the spectrum acquisition unit using a first convolutional neural network;
[0025] A second feature extraction unit, configured to obtain a second feature matrix from the spectrum of the reflected light obtained by the spectrum acquisition unit using a second convolutional neural network;
[0026] a transfer matrix calculation unit, used to calculate a transfer matrix between the first feature matrix obtained by the first feature extraction unit and the second feature matrix obtained by the second feature extraction unit, wherein the transfer matrix is used to represent correlation change information between the spectral information of the incident light and the spectral information of the reflected light;
[0027] A classification loss function value calculation unit, used for passing the transfer matrix obtained by the transfer matrix calculation unit through a classifier to obtain a classification loss function value;
[0028] a constraint factor calculation unit, used for calculating a constraint factor based on sequence evolution between the first feature matrix obtained by the first feature extraction unit and the second feature matrix obtained by the second feature extraction unit, wherein the constraint factor based on sequence evolution includes a nuclear norm of a difference matrix between the first feature matrix and the second feature matrix and a negative of an absolute value of a logarithmic function of a quotient of a first category probability value and a second category probability value obtained by the first feature matrix and the second feature matrix respectively through the classifier;
[0029] a training unit, configured to calculate a weighted sum of the classification loss function value obtained by the classification loss function value calculation unit and the sequence evolution-based constraint factor obtained by the constraint factor calculation unit as a loss function value to train the first convolutional neural network and the second convolutional neural network; and
[0030] Inference module, including:
[0031] An inference data acquisition unit, used to acquire a spectrum of incident light emitted by the communication optical fiber to be detected and a spectrum of reflected light reflected from the object to be detected;
[0032] A first feature matrix generating unit, configured to obtain a first feature matrix from the spectrum of the incident light obtained by the inference data acquiring unit using the first convolutional neural network trained in the training phase;
[0033] A second feature matrix generating unit, configured to obtain a second feature matrix from the spectrum of the reflected light obtained by the inference data acquiring unit using the second convolutional neural network trained in the training phase;
[0034] a transfer matrix generating unit, used to calculate a transfer matrix between the first characteristic matrix obtained by the first characteristic matrix generating unit and the second characteristic matrix obtained by the second characteristic matrix generating unit, wherein the transfer matrix is used to represent correlation change information between spectral information of incident light and spectral information of reflected light; and
[0035] A classification unit is used to pass the transfer matrix obtained by the transfer matrix generation unit through a classifier to obtain a classification result, wherein the classification result is used to indicate whether the quality of the communication optical fiber to be detected meets a preset standard.
[0036] According to another aspect of the present application, an electronic device is provided, comprising: a processor; and a memory, wherein computer program instructions are stored in the memory, and when the computer program instructions are executed by the processor, the processor executes the communication optical fiber management method based on color recognition as described above.
[0037] According to yet another aspect of the present application, a computer-readable medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the processor executes the communication optical fiber management method based on color recognition as described above.
[0038] According to the communication optical fiber management method, system and electronic device based on color recognition provided by the present application, the local features of the spectral graph of the incident light and the reflected light are extracted through a convolutional neural network to represent the feature distribution in a high-dimensional space to obtain a first feature matrix and a second feature matrix, and the change information between the spectral information of the incident light and the spectral information of the reflected light is represented based on the transfer matrix between the two feature matrices, and a constraint factor based on sequence evolution is further introduced to construct a loss function, so as to promote the constraint of the transformation features contained in the transfer matrix itself by optimizing the relevant feature extraction of the first feature matrix and the second feature matrix, thereby improving the accuracy of classification. In this way, the product quality can be better inspected and controlled based on color detection and recognition. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.
[0040] Figure 1 The present invention is a schematic diagram of a communication optical fiber management method based on color recognition according to an embodiment of the present application.
[0041] Figure 2A The flowchart of the training phase in the communication optical fiber management method based on color recognition according to an embodiment of the present application.
[0042] Figure 2B The flowchart is a flowchart of the inference phase in the communication optical fiber management method based on color recognition according to an embodiment of the present application.
[0043] Figure 3A Schematic diagram of the architecture of the training phase in the communication optical fiber management method based on color recognition according to an embodiment of the present application.
[0044] Figure 3B Schematic diagram of the architecture of the inference stage in the communication optical fiber management method based on color recognition according to an embodiment of the present application.
[0045] Figure 4 4 is a block diagram of a communication optical fiber management system based on color recognition according to an embodiment of the present application.
[0046] Figure 5 is a block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION
[0047] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.
[0048] Scenario Overview
[0049] As mentioned above, color detection and identification is an important means of product quality inspection and control for fiber optic sensors. Different colors of light can be expressed as different wavelengths, and color identification is essentially wavelength identification. The basic detection principle of fiber optic sensors is to use external factors to change the optical parameters of light propagating in the optical fiber. The changing light wave signal records the information of the parameter to be measured, so the parameter to be measured can be obtained by modulating and demodulating the light wave signal. The transmitting optical fiber emits a light signal to illuminate the object to be measured, and the receiving optical fiber collects the light signal reflected back by the object to be measured. The light intensity changes with the color of the surface of the object to be measured. Therefore, the change in the light intensity received by the optical fiber accurately reflects the change in color, which can achieve color detection and identification.
[0050] Traditional color recognition uses a light source to directly illuminate an object and performs color recognition based on the relationship between incident light and reflected light. However, this method is greatly affected by the external environment. In addition, the traditional luminous flux of incident light and reflected light is not sensitive to slight color changes, that is, the detection granularity of the traditional statistical model is relatively coarse. Therefore, a communication optical fiber management method based on color recognition is expected.
[0051] It should be understood that color detection and recognition is essentially a classification problem, which can be implemented based on a neural network feature extractor and classifier model.
[0052] Specifically, in the technical solution of the present application, after obtaining the characteristic matrices of the spectrum of the incident light and the spectrum of the reflected light respectively through the convolutional neural network, the change information between the spectrum information of the incident light and the spectrum information of the reflected light can be represented based on the transfer matrix between the two characteristic matrices, that is, the transfer matrix characterizes the change in the color of the optical fiber sensor. That is, the transfer matrix between the two characteristic matrices is calculated and the detection result is obtained through the classifier. However, when calculating the transfer matrix between the first characteristic matrix and the second characteristic matrix, considering that the first characteristic matrix and the second characteristic matrix may not have complete density, there may be insufficient constraints on the solution of the transfer matrix, resulting in insufficient constraints on the transformation features contained in the transfer matrix itself, affecting the classification accuracy.
[0053] In the technical solution of the present application, in addition to the classification loss function value of the transfer matrix, a constraint factor based on sequence evolution is further introduced to construct a loss function. The constraint factor based on sequence evolution includes two items. The first item is the nuclear norm of the difference matrix between the first feature matrix and the second feature matrix, that is, the sum of the eigenvalues of the difference matrix, which is used to represent the degree of feature consistency of the two feature vectors in each main dimension. The second item is the negative of the absolute value of the logarithmic function of the quotient of the first category probability value and the second category probability value obtained by the first feature matrix and the second feature matrix respectively through the classifier, that is, the constraint loss function of sequence evolution is expressed as:
[0054]
[0055] in and are the first characteristic matrix and the second characteristic matrix, and and They respectively represent the classification probability values obtained after the first feature matrix and the second feature matrix pass through the classifier.
[0056] In this way, by minimizing the degree of feature consistency between the two eigenvectors in each main dimension, the sequence evolution represented by the transfer matrix can maintain the consistency of the dependency relationship between positions, and by minimizing the negative of the absolute value of the logarithmic function of the quotient of the first category probability value and the second category probability value, the class probability evolution effect of the sequence evolution represented by the transfer matrix can be maximized. Therefore, during the training process, the structural parameters of the first convolutional neural network and the second convolutional neural network are updated by minimizing the loss function, so as to promote the constraint of the transformation features contained in the transfer matrix itself by optimizing the relevant feature extraction of the first feature matrix and the second feature matrix.
[0057] Based on this, the present application proposes a communication optical fiber management method based on color recognition, which includes: a training phase and an inference phase. Among them, the training stage includes the steps of: obtaining a spectrum of incident light emitted by a communication optical fiber and a spectrum of reflected light reflected from the object to be measured; using a first convolutional neural network to obtain a first feature matrix from the spectrum of the incident light; using a second convolutional neural network to obtain a second feature matrix from the spectrum of the reflected light; calculating a transfer matrix between the first feature matrix and the second feature matrix, wherein the transfer matrix is used to represent the correlation change information between the spectral information of the incident light and the spectral information of the reflected light; passing the transfer matrix through a classifier to obtain a classification loss function value; calculating a constraint factor based on sequence evolution between the first feature matrix and the second feature matrix, wherein the constraint factor based on sequence evolution includes the nuclear norm of the difference matrix between the first feature matrix and the second feature matrix and the negative of the absolute value of the logarithmic function of the quotient of the first category probability value and the second category probability value obtained by the first feature matrix and the second feature matrix respectively through the classifier; calculating the weighted sum between the classification loss function value and the constraint factor based on sequence evolution as the loss function value to train the first convolutional neural network and the second convolutional neural network. Among them, the inference stage includes the steps of: obtaining a spectrum of incident light emitted by the communication optical fiber to be tested and a spectrum of reflected light reflected from the object to be tested; using the first convolutional neural network trained in the training stage to obtain a first feature matrix from the spectrum of the incident light; using the second convolutional neural network trained in the training stage to obtain a second feature matrix from the spectrum of the reflected light; calculating a transfer matrix between the first feature matrix and the second feature matrix, the transfer matrix is used to represent the correlation change information between the spectral information of the incident light and the spectral information of the reflected light; and passing the transfer matrix through a classifier to obtain a classification result, the classification result is used to indicate whether the quality of the communication optical fiber to be tested meets a preset standard.
[0058] Figure 1 The figure shows a scene schematic diagram of a communication optical fiber management method based on color recognition according to an embodiment of the present application. Figure 1 As shown, in the training phase of this application scenario, first, a spectrum analyzer (e.g., Figure 1 P) is obtained by a communication optical fiber (e.g., Figure 1 The spectrum of the incident light emitted by the object under test (e.g. Figure 1 Then, the obtained spectrum of the incident light and the spectrum of the reflected light are input to a server (for example, a server) that is deployed with a communication fiber management algorithm based on color recognition. Figure 1S) as shown in , wherein the server is capable of training the first convolutional neural network and the second convolutional neural network for color-recognition-based communication fiber management algorithm based on color recognition with the spectrum of the incident light and the spectrum of the reflected light.
[0059] After training is completed, in the inference phase, first, a spectrometer (e.g. Figure 1 P) is obtained by a communication optical fiber (e.g., Figure 1 The spectrum of the incident light emitted by the object under test (e.g. Figure 1 Then, the obtained spectrum of the incident light and the spectrum of the reflected light are input to a server (for example, a server) that is deployed with a communication fiber management algorithm based on color recognition. Figure 1 S) as shown in , wherein the server is capable of processing the obtained spectrum of the incident light and the spectrum of the reflected light using a communication fiber management algorithm based on color recognition to generate a classification result indicating whether the quality of the communication fiber to be tested meets a preset standard.
[0060] After introducing the basic principles of the present application, various non-limiting embodiments of the present application will be described in detail with reference to the accompanying drawings.
[0061] Exemplary Methods
[0062] Figure 2A FIG. 1 is a flowchart of the training phase in the communication optical fiber management method based on color recognition according to an embodiment of the present application. Figure 2AAs shown, according to the communication optical fiber management method based on color recognition according to the embodiment of the present application, the method comprises: a training phase, comprising the steps of: S110, obtaining a spectrum of incident light emitted by the communication optical fiber and a spectrum of reflected light reflected from the object to be measured; S120, using a first convolutional neural network to obtain a first feature matrix from the spectrum of the incident light; S130, using a second convolutional neural network to obtain a second feature matrix from the spectrum of the reflected light; S140, calculating a transfer matrix between the first feature matrix and the second feature matrix, the transfer matrix being used to represent the correlation change information between the spectral information of the incident light and the spectral information of the reflected light; S150, converting the transfer matrix into the second feature matrix; The matrix passes through a classifier to obtain a classification loss function value; S160, calculating a constraint factor based on sequence evolution between the first feature matrix and the second feature matrix, wherein the constraint factor based on sequence evolution includes the nuclear norm of the difference matrix between the first feature matrix and the second feature matrix and the negative of the absolute value of the logarithmic function of the quotient of the first category probability value and the second category probability value obtained by the first feature matrix and the second feature matrix respectively through the classifier; S170, calculating the weighted sum between the classification loss function value and the constraint factor based on sequence evolution as the loss function value to train the first convolutional neural network and the second convolutional neural network.
[0063] Figure 2B FIG. 1 is a flowchart of the inference phase of the communication optical fiber management method based on color recognition according to an embodiment of the present application. Figure 2B As shown, according to the color recognition-based communication optical fiber management method of the embodiment of the present application, it also includes: an inference stage, including the steps of: S210, obtaining a spectrum of the incident light emitted by the communication optical fiber to be detected and a spectrum of the reflected light reflected from the object to be detected; S220, using the first convolutional neural network trained in the training stage to obtain a first feature matrix from the spectrum of the incident light; S230, using the second convolutional neural network trained in the training stage to obtain a second feature matrix from the spectrum of the reflected light; S240, calculating a transfer matrix between the first feature matrix and the second feature matrix, the transfer matrix is used to represent the correlation change information between the spectral information of the incident light and the spectral information of the reflected light; and, S250, passing the transfer matrix through a classifier to obtain a classification result, the classification result is used to indicate whether the quality of the communication optical fiber to be detected meets a preset standard.
[0064] Figure 3A The figure shows a schematic diagram of the architecture of the training phase in the communication optical fiber management method based on color recognition according to an embodiment of the present application. Figure 3A As shown, in the training phase, in the network architecture, first, a first convolutional neural network (e.g., Figure 3A The CNN shown in FIG1) obtains the spectrum of the incident light (e.g., Figure 3A ) to obtain the first feature matrix (e.g., Figure 3A MF1 as shown in FIG); then, a second convolutional neural network (e.g., Figure 3A CNN2 as shown in FIG. 2 obtains the spectrum of the reflected light (e.g., Figure 3A ) to obtain a second feature matrix (e.g., Figure 3A MF2 as shown in FIG); then, a transfer matrix between the first feature matrix and the second feature matrix is calculated (for example, Figure 3A Then, the transfer matrix is passed through a classifier (e.g., Figure 3A ) to obtain the classification loss function value (e.g., Figure 3A Then, calculating the constraint factor based on sequence evolution between the first feature matrix and the second feature matrix (for example, Figure 3A CF shown in ); finally, the weighted sum between the classification loss function value and the constraint factor based on sequence evolution is calculated as the loss function value to train the first convolutional neural network and the second convolutional neural network.
[0065] Figure 3B The figure shows a schematic diagram of the architecture of the inference stage in the communication optical fiber management method based on color recognition according to an embodiment of the present application. Figure 3B As shown, in the inference phase, in the network structure, first, the first convolutional neural network (for example, Figure 3B CN1) from the spectrum of the incident light (e.g., Figure 3B ) to obtain the first feature matrix (e.g., Figure 3B MF1 shown in FIG); then, using the second convolutional neural network trained in the training phase (for example, as Figure 3B CN2) from the spectrum of the reflected light (e.g., Figure 3B ) to obtain a second feature matrix (e.g., Figure 3B MF2 as shown in FIG); then, calculating the transfer matrix between the first feature matrix and the second feature matrix (for example, Figure 3B and, finally, passing the transfer matrix through a classifier (e.g., Figure 3B The classifier shown in FIG. 1 is used to obtain a classification result, where the classification result is used to indicate whether the quality of the communication optical fiber to be detected meets a preset standard.
[0066] More specifically, in the training phase, in step S110, step S120 and step S130, the spectrum of the incident light emitted by the communication optical fiber and the spectrum of the reflected light reflected from the object to be measured are obtained, and the first convolutional neural network is used to obtain the first feature matrix from the spectrum of the incident light, and the second convolutional neural network is used to obtain the second feature matrix from the spectrum of the reflected light. As mentioned above, color detection and recognition is an important means of product quality inspection and control of optical fiber sensors. Light of different colors can be expressed as different wavelengths, and color recognition is essentially wavelength recognition. The basic detection principle of optical fiber sensors is to use external factors to change the optical parameters of light propagating in the optical fiber. The changed light wave signal records the information of the parameter to be measured, so that the parameter to be measured can be obtained by modulating and demodulating the light wave signal. The transmitting optical fiber emits a light signal to illuminate the object to be measured, and the receiving optical fiber collects the light signal reflected by the object to be measured. The light intensity changes with the color of the surface of the object to be measured. Therefore, the change in the light intensity received by the optical fiber accurately reflects the change in color, and color detection and recognition can be achieved. This is essentially a classification problem, which can be implemented using neural network-based feature extractors and classifier models.
[0067] That is, specifically, in the technical solution of the present application, first, the spectrum of the incident light emitted by the communication optical fiber and the spectrum of the reflected light reflected from the measured object are obtained by a spectrometer. Then, the first convolutional neural network and the second convolutional neural network are used to respectively extract the characteristic distribution representation of the local characteristics of the spectrum of the incident light and the spectrum of the reflected light in the high-dimensional space, so as to obtain the first characteristic matrix and the second characteristic matrix. Correspondingly, in a specific example, each layer of the first convolutional neural network performs convolution processing based on a two-dimensional convolution kernel, pooling processing along the channel dimension and activation processing on the input data in the forward transfer of the layer to output the first characteristic matrix by the last layer of the first convolutional neural network, wherein the input of the first layer of the first convolutional neural network is the spectrum of the incident light. Each layer of the second convolutional neural network performs convolution processing based on a two-dimensional convolution kernel, pooling processing along the channel dimension and activation processing on the input data in the forward transfer of the layer to output the second characteristic matrix by the last layer of the second convolutional neural network, wherein the input of the first layer of the second convolutional neural network is the spectrum of the reflected light.
[0068] More specifically, in the training phase, in step S140 and step S150, the transfer matrix between the first feature matrix and the second feature matrix is calculated, the transfer matrix is used to represent the correlation change information between the spectral information of the incident light and the spectral information of the reflected light, and the transfer matrix is passed through a classifier to obtain a classification loss function value. That is, in the technical solution of the present application, after obtaining the first feature matrix and the second feature matrix, the transfer matrix between the two feature matrices is further calculated to represent the correlation change information between the spectral information of the incident light and the spectral information of the reflected light. That is, the transfer matrix represents the change in the color of the optical fiber sensor. Then, the transfer matrix between the two feature matrices is processed in a classifier to obtain a classification loss function value. Accordingly, in a specific example, the classifier processes the transfer matrix with the following formula to generate a classification result, wherein the formula is: ,in represents the projection of the transfer matrix into a vector, to is the weight matrix of each fully connected layer, to Representing the bias matrix of each fully connected layer; and calculating the cross entropy value between the classification result and the true value as the classification loss function value.
[0069] Specifically, in an embodiment of the present application, the process of calculating the transfer matrix between the first characteristic matrix and the second characteristic matrix includes: calculating the transfer matrix between the first characteristic matrix and the second characteristic matrix using the following formula, wherein the formula is: S=T*F ,in F represents the first characteristic matrix, T represents the transfer matrix, and S represents the second characteristic matrix.
[0070] More specifically, in the training phase, in step S160 and step S170, the constraint factor based on sequence evolution between the first feature matrix and the second feature matrix is calculated, wherein the constraint factor based on sequence evolution includes the nuclear norm of the difference matrix between the first feature matrix and the second feature matrix and the negative of the absolute value of the logarithmic function of the quotient of the first category probability value and the second category probability value obtained by the first feature matrix and the second feature matrix respectively through the classifier, and the weighted sum between the classification loss function value and the constraint factor based on sequence evolution is calculated as the loss function value to train the first convolutional neural network and the second convolutional neural network. It should be understood that in the technical solution of the present application, the transfer matrix between the two feature matrices is originally calculated and the detection result is obtained through the classifier. However, when calculating the transfer matrix between the first feature matrix and the second feature matrix, considering that the first feature matrix and the second feature matrix may not have complete density, there may be insufficient constraints on the solution of the transfer matrix, resulting in insufficient constraints on the transformation features contained in the transfer matrix itself, affecting the classification accuracy.
[0071] Therefore, in the technical solution of the present application, in addition to the classification loss function value of the transfer matrix, a constraint factor based on sequence evolution is further introduced to construct a loss function. In particular, in a specific example, the constraint factor based on sequence evolution includes two items, the first item is the nuclear norm of the difference matrix between the first feature matrix and the second feature matrix, that is, the sum of the eigenvalues of the difference matrix, which is used to represent the degree of feature consistency of the two feature vectors in each main dimension. The second item is the negative of the absolute value of the logarithmic function of the quotient of the first category probability value and the second category probability value obtained by the first feature matrix and the second feature matrix respectively through the classifier.
[0072] Then, the weighted sum between the classification loss function value and the constraint factor based on sequence evolution can be calculated as the loss function value to train the first convolutional neural network and the second convolutional neural network. It should be understood that by minimizing the degree of feature consistency of the two feature vectors in each main dimension, the sequence evolution represented by the transfer matrix can maintain the consistency of the dependency relationship between positions, and by minimizing the negative of the absolute value of the logarithmic function of the quotient of the first category probability value and the second category probability value, the class probability evolution effect of the sequence evolution represented by the transfer matrix can be maximized, so that during the training process, the structural parameters of the first convolutional neural network and the second convolutional neural network are updated by minimizing the loss function, so as to promote the constraint of the transformation features contained in the transfer matrix itself by optimizing the relevant feature extraction of the first feature matrix and the second feature matrix.
[0073] Specifically, in an embodiment of the present application, the process of calculating the constraint factor based on sequence evolution between the first feature matrix and the second feature matrix includes: calculating the constraint factor based on sequence evolution between the first feature matrix and the second feature matrix using the following formula, wherein the formula is:
[0074]
[0075] in and are the first characteristic matrix and the second characteristic matrix, and and They respectively represent classification probability values obtained after the first feature matrix and the second feature matrix pass through the classifier.
[0076] After the training is completed, the inference stage is entered. That is, after the first convolutional neural network and the second convolutional neural network are trained using the communication optical fiber management algorithm based on color recognition, the trained first convolutional neural network and the second convolutional neural network are used in the actual inference scenario, and the above steps are repeated to obtain the transfer matrix between the first feature matrix and the second feature matrix. Then, the transfer matrix can be further passed through a classifier to obtain a classification result indicating whether the quality of the communication optical fiber to be tested meets the preset standard.
[0077] Specifically, in the inference stage, first, a spectrum of the incident light emitted by the communication optical fiber to be tested and a spectrum of the reflected light reflected from the object to be tested are obtained; then, a first feature matrix is obtained from the spectrum of the incident light using the first convolutional neural network trained in the training stage; then, a second feature matrix is obtained from the spectrum of the reflected light using the second convolutional neural network trained in the training stage; then, a transfer matrix between the first feature matrix and the second feature matrix is calculated, and the transfer matrix is used to represent the correlation change information between the spectral information of the incident light and the spectral information of the reflected light; finally, the transfer matrix is passed through a classifier to obtain a classification result, and the classification result is used to represent whether the quality of the communication optical fiber to be tested meets the preset standard.
[0078] In summary, the communication optical fiber management method based on color recognition based on the embodiment of the present application is explained, which extracts the characteristic distribution representation of the local characteristics of the spectral graph of the incident light and the reflected light in the high-dimensional space through a convolutional neural network to obtain the first characteristic matrix and the second characteristic matrix, and represents the change information between the spectral information of the incident light and the spectral information of the reflected light based on the transfer matrix between the two characteristic matrices, and further introduces a constraint factor based on sequence evolution to construct a loss function, so as to promote the constraint of the transformation characteristics contained in the transfer matrix itself by optimizing the relevant feature extraction of the first characteristic matrix and the second characteristic matrix, thereby improving the accuracy of classification. In this way, the product quality can be better inspected and controlled based on color detection and recognition.
[0079] Exemplary Systems
[0080] Figure 4 FIG. 1 is a block diagram of a communication optical fiber management system based on color recognition according to an embodiment of the present application. Figure 4 As shown, a communication optical fiber management system 400 based on color recognition according to an embodiment of the present application includes: a training module 410 and an inference module 420 .
[0081] like Figure 4As shown, the training module 410 includes: a spectrum acquisition unit 410, which is used to acquire the spectrum of the incident light emitted by the communication optical fiber and the spectrum of the reflected light reflected from the object to be measured; a first feature extraction unit 420, which is used to use a first convolutional neural network to obtain a first feature matrix from the spectrum of the incident light obtained by the spectrum acquisition unit 410; in one example, in the above-mentioned color recognition-based communication optical fiber management system 400, it is used to use a second convolutional neural network to obtain a second feature matrix from the spectrum of the reflected light obtained by the spectrum acquisition unit 410; a transfer matrix calculation unit 440, which is used to calculate the transfer matrix between the first feature matrix obtained by the first feature extraction unit 420 and the second feature matrix obtained by the second feature extraction unit 430, and the transfer matrix is used to represent the correlation change information between the spectral information of the incident light and the spectral information of the reflected light; a classification loss function value calculation unit 450, which is used to convert the The transfer matrix obtained by the transfer matrix calculation unit 440 is passed through a classifier to obtain a classification loss function value; a constraint factor calculation unit 460 is used to calculate the constraint factor based on sequence evolution between the first feature matrix obtained by the first feature extraction unit 420 and the second feature matrix obtained by the second feature extraction unit 430, wherein the constraint factor based on sequence evolution includes the nuclear norm of the difference matrix between the first feature matrix and the second feature matrix and the negative of the absolute value of the logarithmic function of the quotient of the first category probability value and the second category probability value obtained by the first feature matrix and the second feature matrix respectively through the classifier; a training unit 470 is used to calculate the weighted sum between the classification loss function value obtained by the classification loss function value calculation unit 450 and the constraint factor based on sequence evolution obtained by the constraint factor calculation unit 460 as the loss function value to train the first convolutional neural network and the second convolutional neural network.
[0082] like Figure 4As shown, the inference module 420 includes: an inference data acquisition unit 421, which is used to acquire a spectrum of the incident light emitted by the communication optical fiber to be detected and a spectrum of the reflected light reflected from the object to be detected; a first feature matrix generation unit 422, which is used to obtain a first feature matrix from the spectrum of the incident light obtained by the inference data acquisition unit 421 using the first convolutional neural network trained in the training stage; a second feature matrix generation unit 423, which is used to obtain a second feature matrix from the spectrum of the reflected light obtained by the inference data acquisition unit 421 using the second convolutional neural network trained in the training stage; a transfer matrix generation unit 424, which is used to calculate a transfer matrix between the first feature matrix obtained by the first feature matrix generation unit 422 and the second feature matrix obtained by the second feature matrix generation unit 423, the transfer matrix is used to represent the correlation change information between the spectral information of the incident light and the spectral information of the reflected light; and a classification unit 425, which is used to pass the transfer matrix obtained by the transfer matrix generation unit 424 through a classifier to obtain a classification result, and the classification result is used to represent whether the quality of the communication optical fiber to be detected meets the preset standard.
[0083] In one example, in the above-mentioned color recognition-based communication fiber management system 400, the first feature extraction unit 420 is further used for: each layer of the first convolutional neural network performs convolution processing based on a two-dimensional convolution kernel, pooling processing along the channel dimension, and activation processing on the input data in the forward pass of the layer so that the last layer of the first convolutional neural network outputs the first feature matrix, wherein the input of the first layer of the first convolutional neural network is the spectrum of the incident light.
[0084] In one example, in the above-mentioned color recognition-based communication fiber management system 400, the second feature extraction unit 430 is further used for: each layer of the second convolutional neural network performs convolution processing based on a two-dimensional convolution kernel, pooling processing along the channel dimension and activation processing on the input data in the forward pass of the layer to output the second feature matrix by the last layer of the second convolutional neural network, wherein the input of the first layer of the second convolutional neural network is the spectrum of the reflected light.
[0085] In one example, in the above-mentioned color recognition-based communication optical fiber management system 400, the transfer matrix calculation unit 440 is further used to calculate the transfer matrix between the first characteristic matrix and the second characteristic matrix using the following formula, wherein the formula is: S=T*F ,in F represents the first characteristic matrix, T represents the transfer matrix, and S represents the second characteristic matrix.
[0086] In one example, in the above-mentioned color recognition-based communication optical fiber management system 400, the classification loss function value calculation unit 450 is further used for: the classifier processes the transfer matrix according to the following formula to generate a classification result, wherein the formula is: ,in represents the projection of the transfer matrix into a vector, to is the weight matrix of each fully connected layer, to Representing the bias matrix of each fully connected layer; and calculating the cross entropy value between the classification result and the true value as the classification loss function value.
[0087] In one example, in the above-mentioned communication optical fiber management system 400 based on color recognition, the constraint factor calculation unit 460 is further used to: calculate the constraint factor based on sequence evolution between the first feature matrix and the second feature matrix according to the following formula, wherein the formula is:
[0088]
[0089] in and are the first characteristic matrix and the second characteristic matrix, and and They respectively represent classification probability values obtained after the first feature matrix and the second feature matrix pass through the classifier.
[0090] Here, those skilled in the art will appreciate that the specific functions and operations of the various units and modules in the above-mentioned communication optical fiber management system 400 based on color recognition have been described in the above reference. Figures 1 to 3B The description of the communication optical fiber management method based on color recognition has been described in detail, and therefore, its repeated description will be omitted.
[0091] As described above, the communication fiber management system 400 based on color recognition according to the embodiment of the present application can be implemented in various terminal devices, such as a server of a communication fiber management algorithm based on color recognition. In one example, the communication fiber management system 400 based on color recognition according to the embodiment of the present application can be integrated into the terminal device as a software module and / or a hardware module. For example, the communication fiber management system 400 based on color recognition can be a software module in the operating device of the terminal device, or can be an application developed for the terminal device; of course, the communication fiber management system 400 based on color recognition can also be one of the many hardware modules of the terminal device.
[0092] Alternatively, in another example, the color-recognition-based communication fiber management system 400 and the terminal device may also be separate devices, and the color-recognition-based communication fiber management system 400 may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.
[0093] Exemplary Electronic Devices
[0094] Below, reference Figure 5 To describe the electronic device according to the embodiment of the present application. Figure 5 As shown, the electronic device 10 includes one or more processors 11 and a memory 12. The processor 11 may be a central processing unit (CPU) or other processing units with data processing capability and / or instruction execution capability, and may control other components in the electronic device 10 to perform desired functions.
[0095] The memory 12 may include one or more computer program products, and the computer program product may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 11 may run the program instructions to implement the functions of the communication optical fiber management method based on color recognition of the various embodiments of the present application described above and / or other desired functions. Various contents such as the first characteristic matrix, constraint factors, etc. may also be stored in the computer-readable storage medium.
[0096] In one example, the electronic device 10 may further include: an input device 13 and an output device 14 , and these components are interconnected via a bus device and / or other forms of connection mechanisms (not shown).
[0097] The input device 13 may include, for example, a keyboard, a mouse, etc.
[0098] The output device 14 can output various information to the outside, including classification results, etc. The output device 14 can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and the like.
[0099] Of course, to simplify, Figure 5Only some of the components related to the present application in the electronic device 10 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application situations, the electronic device 10 may also include any other appropriate components.
[0100] Exemplary computer program products and computer-readable storage media
[0101] In addition to the above-mentioned methods and devices, an embodiment of the present application may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to perform the steps in the functions of the color-recognition-based communication optical fiber management method according to various embodiments of the present application described in the above-mentioned "Exemplary Method" section of this specification.
[0102] The computer program product may be written in any combination of one or more programming languages to write program codes for performing the operations of the embodiments of the present application, including object-oriented programming languages, such as Java, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as an independent software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.
[0103] In addition, an embodiment of the present application may also be a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, causes the processor to execute the steps of the color-recognition-based communication fiber management method described in the above “Exemplary Method” section of this specification.
[0104] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor devices, devices or components, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.
[0105] The basic principles of the present application are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present application are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present application. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, not for limitation, and the above details do not limit the present application to being implemented by adopting the above specific details.
[0106] The block diagrams of the devices, apparatuses, equipment, and systems involved in this application are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagram. As will be appreciated by those skilled in the art, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open words, referring to "including but not limited to", and can be used interchangeably with them. The words "or" and "and" used here refer to the words "and / or" and can be used interchangeably with them, unless the context clearly indicates otherwise. The words "such as" used here refer to the phrase "such as but not limited to", and can be used interchangeably with them.
[0107] It should also be noted that in the apparatus, device and method of the present application, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present application.
[0108] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.
[0109] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.
Claims
1. A communication optical fiber management method based on color recognition, characterized in that: include: The training phase includes: Obtaining a spectrum of incident light emitted by the communication optical fiber and a spectrum of reflected light reflected from the object under test; Obtaining a first feature matrix from the spectrum of the incident light using a first convolutional neural network; Obtaining a second feature matrix from the spectrum of the reflected light using a second convolutional neural network; Calculating a transfer matrix between the first characteristic matrix and the second characteristic matrix, wherein the transfer matrix is used to represent correlation change information between spectral information of incident light and spectral information of reflected light; Passing the transfer matrix through a classifier to obtain a classification loss function value; Calculating a constraint factor based on sequence evolution between the first feature matrix and the second feature matrix, wherein the constraint factor based on sequence evolution includes a nuclear norm of a difference matrix between the first feature matrix and the second feature matrix and a negative of an absolute value of a logarithmic function of a quotient of a first category probability value and a second category probability value obtained by the first feature matrix and the second feature matrix respectively through the classifier; Calculating a weighted sum between the classification loss function value and the sequence evolution-based constraint factor as a loss function value to train the first convolutional neural network and the second convolutional neural network; and The inference phase includes: Acquire a spectrum of incident light emitted by the communication optical fiber to be detected and a spectrum of reflected light reflected from the object to be detected; Obtaining a first feature matrix from the spectrum of the incident light using the first convolutional neural network trained in the training phase; Obtaining a second feature matrix from the spectrum of the reflected light using the second convolutional neural network trained in the training phase; Calculating a transfer matrix between the first characteristic matrix and the second characteristic matrix, wherein the transfer matrix is used to represent correlation change information between spectral information of incident light and spectral information of reflected light; and The transfer matrix is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the quality of the communication optical fiber to be detected meets a preset standard.
2. The communication optical fiber management method based on color recognition according to claim 1, wherein: Obtaining a first feature matrix from the spectrum of the incident light using a first convolutional neural network, comprising: In the forward pass of the layer, each layer of the first convolutional neural network performs convolution processing based on a two-dimensional convolution kernel, pooling processing along the channel dimension, and activation processing on the input data so that the first feature matrix is output by the last layer of the first convolutional neural network, wherein the input of the first layer of the first convolutional neural network is the spectrum of the incident light.
3. The communication optical fiber management method based on color recognition according to claim 2, wherein: Using a second convolutional neural network to obtain a second feature matrix from the spectrum of the reflected light, comprising: In the forward pass of the layer, each layer of the second convolutional neural network performs convolution processing based on a two-dimensional convolution kernel, pooling processing along the channel dimension, and activation processing on the input data so that the second feature matrix is output by the last layer of the second convolutional neural network, wherein the input of the first layer of the second convolutional neural network is the spectrum of the reflected light.
4. The communication optical fiber management method based on color recognition according to claim 3, wherein: Calculating a transfer matrix between the first characteristic matrix and the second characteristic matrix includes: The transfer matrix between the first characteristic matrix and the second characteristic matrix is calculated by the following formula, wherein the formula is: S=T*F ,in F represents the first characteristic matrix, T represents the transfer matrix, and S represents the second characteristic matrix.
5. The communication optical fiber management method based on color recognition according to claim 4, wherein: Passing the transfer matrix through a classifier to obtain a classification loss function value includes: The classifier processes the transfer matrix according to the following formula to generate a classification result, wherein the formula is: ,in represents the projection of the transfer matrix into a vector, to is the weight matrix of each fully connected layer, to represents the bias matrix of each fully connected layer; and The cross entropy value between the classification result and the true value is calculated as the classification loss function value.
6. The communication optical fiber management method based on color recognition according to claim 5, wherein: Calculating a constraint factor based on sequence evolution between the first feature matrix and the second feature matrix, comprising: The constraint factor based on sequence evolution between the first feature matrix and the second feature matrix is calculated by the following formula, wherein the formula is: , Among them, the first term is the nuclear norm of the difference matrix between the first characteristic matrix and the second characteristic matrix, and are the first characteristic matrix and the second characteristic matrix, and and They respectively represent classification probability values obtained after the first feature matrix and the second feature matrix pass through the classifier.
7. A communication optical fiber management system based on color recognition, characterized in that: include: Training modules include: A spectrum acquisition unit, used to acquire a spectrum of incident light emitted by the communication optical fiber and a spectrum of reflected light reflected from the object to be measured; A first feature extraction unit, configured to obtain a first feature matrix from the spectrum of the incident light obtained by the spectrum acquisition unit using a first convolutional neural network; A second feature extraction unit, configured to obtain a second feature matrix from the spectrum of the reflected light obtained by the spectrum acquisition unit using a second convolutional neural network; a transfer matrix calculation unit, used to calculate a transfer matrix between the first feature matrix obtained by the first feature extraction unit and the second feature matrix obtained by the second feature extraction unit, wherein the transfer matrix is used to represent correlation change information between the spectral information of the incident light and the spectral information of the reflected light; A classification loss function value calculation unit, used for passing the transfer matrix obtained by the transfer matrix calculation unit through a classifier to obtain a classification loss function value; a constraint factor calculation unit, used for calculating a constraint factor based on sequence evolution between the first feature matrix obtained by the first feature extraction unit and the second feature matrix obtained by the second feature extraction unit, wherein the constraint factor based on sequence evolution includes a nuclear norm of a difference matrix between the first feature matrix and the second feature matrix and a negative of an absolute value of a logarithmic function of a quotient of a first category probability value and a second category probability value obtained by the first feature matrix and the second feature matrix respectively through the classifier; a training unit, configured to calculate a weighted sum of the classification loss function value obtained by the classification loss function value calculation unit and the sequence evolution-based constraint factor obtained by the constraint factor calculation unit as a loss function value to train the first convolutional neural network and the second convolutional neural network; and Inference modules, including: An inference data acquisition unit, used to acquire a spectrum of incident light emitted by the communication optical fiber to be detected and a spectrum of reflected light reflected from the object to be detected; A first feature matrix generating unit, configured to obtain a first feature matrix from the spectrum of the incident light obtained by the inference data acquiring unit using the first convolutional neural network trained in the training phase; A second feature matrix generating unit, configured to obtain a second feature matrix from the spectrum of the reflected light obtained by the inference data acquiring unit using the second convolutional neural network trained in the training phase; a transfer matrix generating unit, used to calculate a transfer matrix between the first characteristic matrix obtained by the first characteristic matrix generating unit and the second characteristic matrix obtained by the second characteristic matrix generating unit, wherein the transfer matrix is used to represent correlation change information between spectral information of incident light and spectral information of reflected light; and A classification unit is used to pass the transfer matrix obtained by the transfer matrix generation unit through a classifier to obtain a classification result, wherein the classification result is used to indicate whether the quality of the communication optical fiber to be detected meets a preset standard.
8. The communication optical fiber management system based on color recognition according to claim 7, wherein: The first feature extraction unit is further used to: In the forward pass of the layer, each layer of the first convolutional neural network performs convolution processing based on a two-dimensional convolution kernel, pooling processing along the channel dimension, and activation processing on the input data so that the first feature matrix is output by the last layer of the first convolutional neural network, wherein the input of the first layer of the first convolutional neural network is the spectrum of the incident light.
9. The communication optical fiber management system based on color recognition according to claim 7, wherein: The transfer matrix calculation unit is further used for: The transfer matrix between the first characteristic matrix and the second characteristic matrix is calculated by the following formula, wherein the formula is: S=T*F ,in F represents the first characteristic matrix, T represents the transfer matrix, and S represents the second characteristic matrix.
10. An electronic device, comprising: processor; as well as A memory, in which computer program instructions are stored, and when the computer program instructions are executed by the processor, the processor executes the communication optical fiber management method based on color recognition as described in any one of claims 1-6.
