Fiber core splice closure information input method, device, equipment and medium
Through the feature extraction and prediction model of the OTDR test curve, the automatic core continuation box information entry solves the problem of low efficiency in the existing technology and realizes efficient and accurate information entry and management.
Patent Information
- Application Number
- CN202411264434.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-10
- Publication Date
- 2024-12-13
AI Technical Summary
In the prior art, the information input efficiency of the fiber core connecting box is low, requires a lot of manual time, and is prone to calculation and entry errors.
By obtaining the OTDR test curve, extracting feature data, using prediction models to predict and classify core continuation box information, it is automatically entered into the resource management system, and manual intervention is reduced.
It improves the accuracy and efficiency of core connection box information entry, realizes automated management, reduces noise interference, and improves the efficiency and accuracy of fiber network maintenance and fault diagnosis.
Smart Images
Figure CN119131819A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method, device, equipment and medium for entering information of a fiber core splice box. Background Art
[0002] At present, optical cable inspection mainly relies on the Optical Time-Domain Reflectometer (OTDR) technology. OTDR measures by sending and receiving pulsed optical signals to generate OTDR curves, which are graphs that reflect the changes in optical signal strength over distance in an optical fiber link.
[0003] The existing technology mainly relies on manually querying the OTDR curve, calculating various fiber core splice box information based on the queried OTDR curve, and manually entering various information into the system, which takes a lot of time and effort, and may also cause various errors, such as calculation errors or input errors, etc. Therefore, the efficiency of the existing technology is low. Summary of the invention
[0004] The invention provides a method, device, equipment and medium for entering information of a core splice box, which are used to solve the problem of low efficiency of the information entry method of the core splice box in the prior art.
[0005] In a first aspect, the present invention provides a method for entering information of a core splice box, comprising: Obtain multiple target optical time domain reflectometer OTDR test curves; Based on the multiple target OTDR test curves, the fiber core splice box information is predicted to obtain a fiber core splice box information set corresponding to each target OTDR test curve; any of the fiber core splice box information sets includes multiple types of fiber core splice box information; Based on multiple fiber core splice box information sets, the fiber core splice box information of the same type is classified to obtain target information sets corresponding to multiple types of fiber core splice box information respectively; Enter the target information sets corresponding to each type into the fiber core splice box resource management system one by one.
[0006] In one embodiment, when the core splice box information is predicted based on the multiple target OTDR test curves to obtain the core splice box information set corresponding to each target OTDR test curve, the following steps are performed for each target OTDR test curve: Feature extraction is performed on the target OTDR test curve to obtain multiple types of OTDR curve feature data; the multiple types of OTDR curve feature data at least include first OTDR curve feature data, second OTDR curve feature data and third OTDR curve feature data; the first OTDR curve feature data includes the time of the rising point appearing in the descending curve and the time of the initial end of the curve; the second OTDR curve feature data includes the curve abscissa and the height of the trajectory before and after the event; the third OTDR curve feature data includes the height of the reflection time front end and the reflection peak; Inputting various types of OTDR curve feature data into a core splice box information prediction model, obtaining various types of core splice box information corresponding to the target OTDR test curve output by the core splice box information prediction model; the various types of core splice box information at least include core length, attenuation value and fault point location; the core splice box information prediction model is obtained by model training based on various types of OTDR curve feature samples and core splice box information labels corresponding to various types of OTDR curve feature samples.
[0007] In one embodiment, the fiber core splice box information prediction model is trained in the following manner: Get OTDR test curve samples; Extracting features from the OTDR test curve samples to obtain multiple types of OTDR curve feature samples; Label each type of OTDR curve characteristic sample to obtain a fiber core splicing box information label corresponding to each type of OTDR curve characteristic sample; Model training is performed based on each type of OTDR curve feature sample and the fiber core splice box information label corresponding to each type of OTDR curve feature sample to obtain the fiber core splice box information prediction model.
[0008] In one embodiment, the target information sets corresponding to the core splice box information of the same type are obtained by classifying the multiple core splice box information sets based on the multiple core splice box information sets, including: Splicing is performed based on multiple fiber core splicing box information sets to obtain a splicing information set; Based on the splicing information set, generate an initial table image; Performing image preprocessing on the initial table image to obtain a target table image; Performing table detection on the target table image to obtain a fiber core splicing box information table image corresponding to each target OTDR test curve; Marking the area where each type of fiber core splice box information is located in each fiber core splice box information table picture, to obtain a plurality of table areas marked in each fiber core splice box information table picture; Based on the multiple table areas marked in each fiber core splice closure information table picture, the fiber core splice closure information of the same type is classified to obtain target information sets corresponding to multiple types of fiber core splice closure information.
[0009] In one embodiment, the target information sets corresponding to the core splice box information of the same type are obtained by classifying the multiple table areas marked in each core splice box information table picture, including: Based on each table area marked in each fiber core splicing box information table image, content extraction is performed to obtain text content corresponding to each table area marked in each fiber core splicing box information table image; Based on the text content corresponding to each table area marked in each fiber core splice box information table picture, the fiber core splice box information of the same type is integrated to obtain target information sets corresponding to multiple types of fiber core splice box information.
[0010] In one embodiment, when content extraction is performed based on each table area marked in each fiber core splice box information table image to obtain text content corresponding to each table area marked in each fiber core splice box information table image, the following steps are performed for each table area: Get the cells in the table area; Performing text recognition on the cells to obtain the text to be checked corresponding to the table area; Perform text error correction on the text to be checked to obtain text content corresponding to the table area.
[0011] In one embodiment, when acquiring multiple target OTDR test curves, the following steps are performed for each target OTDR test curve: Get the initial OTDR test curve; Performing wavelet decomposition on the initial OTDR test curve to obtain wavelet decomposition coefficients at multiple levels; Performing threshold quantization processing on the wavelet decomposition coefficients of the multiple levels to obtain multiple wavelet decomposition coefficients after threshold quantization processing; Wavelet reconstruction is performed based on the multiple wavelet decomposition coefficients after the threshold quantization process to obtain the target OTDR test curve.
[0012] In a second aspect, the present invention further provides a device for entering information of a core splice box, comprising: An acquisition module, used for acquiring multiple target optical time domain reflectometer OTDR test curves; A core splice box information prediction module is used to predict the core splice box information based on the multiple target OTDR test curves to obtain a core splice box information set corresponding to each target OTDR test curve; any of the core splice box information sets includes multiple types of core splice box information; A classification module, for classifying the same type of fiber core splice box information based on multiple fiber core splice box information sets to obtain target information sets corresponding to multiple types of fiber core splice box information respectively; The input module is used to input the target information set corresponding to each type into the fiber core splicing box resource management system one by one.
[0013] In a third aspect, the present invention provides a device, comprising an electronic device, wherein the electronic device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of any of the above-mentioned methods for entering information of a core splicing box are implemented.
[0014] In a fourth aspect, the present invention further provides a medium, comprising a non-transitory computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of any of the above-mentioned methods for entering fiber core splice box information are implemented.
[0015] The fiber core splice box information entry method, device, equipment and medium provided by the present invention predict the fiber core splice box information based on multiple target OTDR test curves, and can obtain the corresponding fiber core splice box information set, and classify and manage them according to types, so as to further realize the entry of the fiber core splice box information into the fiber core splice box resource management system one by one according to different types, without manual intervention in the calculation and entry of the core splice box information, thereby improving the accuracy and efficiency of the fiber core splice box information entry, thereby achieving the goals of automated management and improved work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0017] Figure 1 It is a flow chart of the method for entering information of a fiber core splicing box provided by the present invention.
[0018] Figure 2 It is a flow chart of OTDR signal denoising process provided by the present invention.
[0019] Figure 3 It is a basic structural diagram of the RNN provided by the present invention.
[0020] Figure 4 It is a schematic diagram of the steps of the RNN provided by the present invention.
[0021] Figure 5 It is a schematic diagram of the process of digitizing the information of the fiber core splicing box provided by the present invention.
[0022] Figure 6 It is a schematic diagram of the process of image preprocessing provided by the present invention.
[0023] Figure 7 It is a schematic diagram of the process of marking a region of interest provided by the present invention.
[0024] Figure 8 It is a structural schematic diagram of the fiber core splicing box information input device provided by the present invention.
[0025] Fig. 9 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION
[0026] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0027] The terms "first", "second", etc. in the present invention are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention can be implemented in an order other than those illustrated or described herein.
[0028] Combine the following Figure 1-Figure 9 The present invention describes the fiber core splice box information entry method, device, equipment and medium provided by the present invention.
[0029] It should be noted that the fiber core splice box information entry method provided in the embodiment of the present invention is implemented based on the fiber core splice box information entry device. The embodiment of the present invention takes the fiber core splice box information entry device as an example to describe the fiber core splice box information entry method.
[0030] Combination Figure 1 , Figure 1 It is a flow chart of the method for entering information of a fiber core splicing box provided by the present invention.
[0031] like Figure 1 As shown, the method includes the following: Step 101, obtaining multiple target optical time domain reflectometer OTDR test curves; Step 102: Predict the fiber core splice box information based on the multiple target OTDR test curves to obtain a fiber core splice box information set corresponding to each target OTDR test curve; Step 103: Based on the multiple fiber core splice closure information sets, the fiber core splice closure information of the same type is classified to obtain target information sets corresponding to the multiple types of fiber core splice closure information respectively; Step 104: Enter the target information sets corresponding to each type into the fiber core splice box resource management system one by one.
[0032] Specifically, testers use OTDR technology to test optical cables, measure the loss and reflection on the optical fiber, and draw corresponding multiple initial OTDR test curves, which can be used to further analyze fiber core splice box information such as optical fiber length, loss, and fault point location.
[0033] Furthermore, the fiber core splicing box information input device obtains a plurality of initial OTDR test curves.
[0034] It should be noted that the test curve can be obtained through OTDR technology, but there is a part of noise data. When performing data analysis, useful data will be covered by noise, resulting in calculation errors. Therefore, for the obtained multiple initial OTDR test curves, signal denoising can be performed by wavelet transform.
[0035] Furthermore, the fiber core splicing box information input device performs signal denoising processing on the multiple initial OTDR test curves respectively to obtain multiple target OTDR test curves.
[0036] Furthermore, the fiber core splice box information entry device performs feature extraction based on multiple target OTDR test curves to obtain multiple types of OTDR curve feature data for each target OTDR test curve, wherein each type of OTDR curve feature can be used to calculate and analyze a type of fiber core splice box information.
[0037] Furthermore, the core splice box information entry device predicts the core splice box information based on multiple types of OTDR curve characteristic data of each target OTDR test curve, and obtains multiple types of core splice box information corresponding to each target OTDR test curve, that is, a core splice box information set corresponding to each target OTDR test curve.
[0038] Furthermore, the core splice closure information input device classifies the core splice closure information of the same type based on the multiple core splice closure information sets to obtain target information sets corresponding to the multiple types of core splice closure information.
[0039] Furthermore, the fiber core splice closure information input device performs structured data conversion on target information sets respectively corresponding to the various types of fiber core splice closure information to obtain target information sets respectively corresponding to the various types of fiber core splice closure information after structured data conversion.
[0040] Furthermore, the core splice closure information entry device enters the target information sets corresponding to the various types of core splice closure information converted from the structured data into the core splice closure resource management system one by one.
[0041] The fiber core splice box information entry method provided by the present invention predicts the fiber core splice box information based on multiple target OTDR test curves, can obtain the corresponding fiber core splice box information set, and classify and manage it according to type, and further realize the fiber core splice box information being entered into the fiber core splice box resource management system one by one according to different types, without manual intervention in the calculation and entry of the fiber core splice box information, thereby improving the accuracy and efficiency of the fiber core splice box information entry, thereby achieving the goals of automated management and improved work efficiency.
[0042] Further, based on step 101, when multiple target OTDR test curves are obtained, the following steps are performed for each target OTDR test curve: Get the initial OTDR test curve; Performing wavelet decomposition on the initial OTDR test curve to obtain wavelet decomposition coefficients at multiple levels; Performing threshold quantization processing on the wavelet decomposition coefficients of the multiple levels to obtain multiple wavelet decomposition coefficients after threshold quantization processing; Wavelet reconstruction is performed based on the multiple wavelet decomposition coefficients after the threshold quantization process to obtain the target OTDR test curve.
[0043] Combination Figure 2 , Figure 2 It is a flow chart of OTDR signal denoising process provided by the present invention.
[0044] Specifically, the fiber core splicing box information input device obtains an initial OTDR test curve.
[0045] Furthermore, the fiber core splicing box information input device performs wavelet decomposition on the initial OTDR test curve to obtain wavelet decomposition coefficients at multiple levels.
[0046] In one embodiment, a wavelet decomposition level P is set, and wavelet decomposition is performed on the data in each level. Therefore, the calculation formula of the wavelet decomposition coefficient is as follows:
[0047] Among them, k represents the base noise; t represents the time; , Respectively represent the i-th layer wavelet coefficient and the M-th layer wavelet coefficient at position P; Characterizes the wavelet decomposition coefficients at position P.
[0048] Furthermore, the core splice box information input device performs threshold quantization processing on the wavelet decomposition coefficients of multiple levels to obtain multiple wavelet decomposition coefficients after the threshold quantization processing.
[0049] In one embodiment, the calculation formula for wavelet decomposition coefficient threshold quantization is as follows:
[0050] in, Characterize the wavelet decomposition coefficients at position P; Represents the Mth layer wavelet coefficient at time t; Characterize the wavelet decomposition coefficients at time t; Characterizes the wavelet decomposition coefficients after threshold quantization processing.
[0051] Furthermore, the fiber core splice box information input device performs wavelet reconstruction based on a plurality of wavelet decomposition coefficients after threshold quantization processing to obtain a target OTDR test curve.
[0052] In one embodiment, the low-frequency coefficient d is determined in position P, and the wavelet decomposition coefficients after threshold quantization are combined. Perform wavelet reconstruction, so the calculation method of wavelet reconstruction is as follows:
[0053] in, Characterizes the wavelet decomposition level at the low frequency coefficient d; Characterize the wavelet decomposition level at the current time; Characterize the target OTDR test curve.
[0054] The embodiment of the present invention effectively processes and optimizes the original curve through wavelet decomposition and reconstruction, which can significantly reduce noise, improve signal quality, and extract accurate target signals so that subsequent input into the model can accurately predict the fiber core splice box information.
[0055] Further, based on step 102, when the core splice box information is predicted based on the multiple target OTDR test curves to obtain the core splice box information set corresponding to each target OTDR test curve, the following steps are performed for each target OTDR test curve: Feature extraction is performed on the target OTDR test curve to obtain multiple types of OTDR curve feature data; the multiple types of OTDR curve feature data at least include first OTDR curve feature data, second OTDR curve feature data and third OTDR curve feature data; the first OTDR curve feature data includes the time of the rising point appearing in the descending curve and the time of the initial end of the curve; the second OTDR curve feature data includes the curve abscissa and the height of the trajectory before and after the event; the third OTDR curve feature data includes the height of the reflection time front end and the reflection peak; Inputting various types of OTDR curve feature data into a core splice box information prediction model, obtaining various types of core splice box information corresponding to the target OTDR test curve output by the core splice box information prediction model; the various types of core splice box information at least include core length, attenuation value and fault point location; the core splice box information prediction model is obtained by model training based on various types of OTDR curve feature samples and core splice box information labels corresponding to various types of OTDR curve feature samples.
[0056] Specifically, the fiber core splice box information input device extracts features from the target OTDR test curve to obtain multiple types of OTDR curve feature data, including at least first OTDR curve feature data, second OTDR curve feature data and third OTDR curve feature data.
[0057] It should be noted that the first OTDR curve characteristic data includes the time of the rising point of the descending curve in the target OTDR test curve and the time of the initial end of the curve, which can be used to predict the core length; the second OTDR curve characteristic data includes the horizontal coordinate of the curve in the target OTDR test curve and the height of the trajectory before and after the event, which can be used to predict the attenuation value; the third OTDR curve characteristic data includes the height of the reflection time front end and the reflection peak in the target OTDR test curve, which can be used to predict the location of the fault point.
[0058] Furthermore, the core splice box information entry device inputs various types of OTDR curve characteristic data into the core splice box information prediction model to obtain various types of core splice box information corresponding to the target OTDR test curve output by the core splice box information prediction model, which at least includes the core length, attenuation value and fault point location.
[0059] The embodiment of the present invention extracts multiple types of features from the target OTDR test curve and inputs the multiple types of feature data into a trained fiber core splice box information prediction model, thereby being able to accurately obtain fiber core splice box information such as fiber core length, attenuation value, and fault point location, thereby improving the automation and accuracy of OTDR curve analysis, thereby significantly improving the efficiency and accuracy of optical fiber network maintenance and fault diagnosis.
[0060] Furthermore, the fiber core splice box information prediction model is trained in the following way: Get OTDR test curve samples; Extracting features from the OTDR test curve samples to obtain multiple types of OTDR curve feature samples; Label each type of OTDR curve characteristic sample to obtain a fiber core splicing box information label corresponding to each type of OTDR curve characteristic sample; Model training is performed based on each type of OTDR curve feature sample and the fiber core splice box information label corresponding to each type of OTDR curve feature sample to obtain the fiber core splice box information prediction model.
[0061] Specifically, the fiber core splice box information entry device obtains an initial OTDR test curve sample. Further, the fiber core splice box information entry device performs signal denoising on the initial OTDR test curve sample to obtain an OTDR test curve sample after signal denoising.
[0062] It should be noted that the OTDR test curve samples after signal denoising can be stored in the optical fiber information database at the same time, and the relevant information in the optical fiber information database can be processed by data equalization to be used as training data for subsequent model training process.
[0063] Furthermore, the fiber core splicing box information entry device extracts features from the OTDR test curve samples to obtain various types of OTDR curve feature samples, such as the time of the rising point of the descending curve in the OTDR test curve sample and the time of the initial end of the curve, the height of the horizontal coordinate of the curve in the OTDR test curve sample and the trajectory before and after the event, and the height of the reflection time front end and the reflection peak in the OTDR test curve sample.
[0064] Furthermore, the core splice box information entry device marks each type of OTDR curve characteristic sample to obtain a core splice box information label corresponding to each type of OTDR curve characteristic sample, such as a core length label, an attenuation value label and a fault point location label.
[0065] Furthermore, the core splice box information input device performs model training based on each type of OTDR curve feature sample and the core splice box information label corresponding to each type of OTDR curve feature sample to obtain a core splice box information prediction model.
[0066] In one embodiment, the model can use a recurrent neural network (RNN) combined with Figure 3 , Figure 3 It is a basic structural diagram of the RNN provided by the present invention.
[0067] like Figure 3 As shown in the figure, the basic structure of RNN includes X, h, u, v, w and y. X represents the input sequence, which contains input information at multiple times. The hidden state h is the memory unit of RNN, which is used to capture long-term dependencies in the sequence. The input weight matrix u, the hidden state weight matrix v, and the output weight matrix w are used to adjust the input information, calculate the hidden state, and map to the output layer, respectively. y represents the output of RNN.
[0068] Combination Figure 4 As shown, Figure 4 It is a schematic diagram of the steps of the RNN provided by the present invention.
[0069] Step S01, initialization sequence: First, initialize the time series t, input sequence X, hidden sequence h, and output sequence y respectively: .
[0070] Step S02: Update hidden status: For the hidden state update in the recurrent neural network, it can be expressed as:
[0071] in, Characterize activation functions; Represents the input at time t; Represents the hidden state at time t-1; u represents the weight matrix, which determines the current input How to affect the current hidden state ; v represents the weight matrix, which determines the previous hidden state How to affect the current hidden state ; Represents the hidden state at time t.
[0072] Step S03, output result: By mapping the hidden state to the output layer, we get the result , the expression is:
[0073] in, Represents the intermediate representation before the hidden state is mapped to the output layer; w represents the intermediate representation before the hidden state is mapped to the output layer; The weight matrix that maps to the final output.
[0074] In order to minimize the error, the cross entropy loss function is usually used. The overall loss function L can be obtained by accumulating the loss at each moment in the time series, formally expressed as:
[0075] During the training phase, the loss function is minimized through optimization algorithms such as gradient descent. Specifically, u, v, and w are derived separately, and then the parameters are updated based on the gradient information.
[0076] Use layer normalization to optimize RNN, and set the hidden sequence to ,in Characterizing the hidden state at time t, the basic principle of layer normalization is to introduce a standardized operation after the linear transformation of the hidden state. The optimization process is described below.
[0077] Step S01: Standardization: For each element of the hidden state ( i Represents the index of an element), its standardized calculation method is:
[0078] in, Represents the mean of the hidden state at time t; Characterize variance; Represents a positive number and prevents division by 0.
[0079] Step S02: Linear transformation and translation: Introducing learnable parameters for linear transformation and translation, the final layer normalized output is:
[0080] in, and They represent the learnable scaling factor and translation factor respectively, which are adjusted through gradient updates during training to improve the adaptability of the model.
[0081] The embodiment of the present invention obtains OTDR test curve samples and performs feature extraction and annotation to generate various types of OTDR curve feature samples and their corresponding fiber core splice box information labels, and performs model training based on these data. The fiber core splice box information prediction model finally obtained can efficiently and accurately predict the fiber core length, attenuation value and fault point location and other fiber core splice box information in the optical fiber system, thereby realizing the automation and precision of optical fiber maintenance and fault diagnosis.
[0082] Further, based on step 103, the target information sets corresponding to the multiple types of core splice closure information are obtained by classifying the multiple types of core splice closure information based on the multiple core splice closure information sets, including: Splicing is performed based on multiple fiber core splicing box information sets to obtain a splicing information set; Based on the splicing information set, generate an initial table image; Performing image preprocessing on the initial table image to obtain a target table image; Performing table detection on the target table image to obtain a fiber core splicing box information table image corresponding to each target OTDR test curve; Marking the area where each type of fiber core splice box information is located in each fiber core splice box information table picture, to obtain a plurality of table areas marked in each fiber core splice box information table picture; Based on the multiple table areas marked in each fiber core splice closure information table picture, the fiber core splice closure information of the same type is classified to obtain target information sets corresponding to multiple types of fiber core splice closure information.
[0083] Combination Figure 5 , Figure 5 It is a schematic diagram of the process of digitizing the information of the fiber core splicing box provided by the present invention.
[0084] Specifically, the core splice box information input device generates multiple core splice box information tables based on multiple core splice box information sets. Therefore, each target OTDR test curve corresponds to a core splice box information table, and each core splice box information table records multiple types of core splice box information.
[0085] Furthermore, the fiber core splice box information input device splices a plurality of fiber core splice box information tables to obtain a spliced information set.
[0086] It should be noted that the splicing method can be horizontal splicing or vertical splicing.
[0087] Furthermore, the fiber core splice box information entry device generates an initial table image including the splicing information set based on the splicing information set.
[0088] Furthermore, the fiber core splice box information input device performs image preprocessing on the initial table image to obtain a target table image.
[0089] It should be noted that image technology can be used to pre-process images, including image tilt, blur, underexposure, table tilt, etc. Figure 6 , Figure 6 Schematic diagram of the image preprocessing process provided by the present invention. Figure 6 As shown, the preprocessing process of the image is as follows: Step S01: input the picture to be processed; Step S02: uniformly format the pictures to a uniform size; Step S03: grayscale the image; Step S04: Binarization processing, converting into a form more suitable for human or machine analysis and processing; Step S05: contour detection, obtaining the contour edge of the wiring table in the image; Step S06: obtaining the angle, obtaining the enclosing matrix of the table outline, and obtaining the tilt angle; L1: orientation check, machine learning training picture orientation check training model, determine whether the angle is normal, if normal, then go to step S08, if not normal, go to step S07; Step S07: rotation correction, performing rotation correction according to the angle; Step S08: extracting the image area within the outline to form an independent logo image; Step S09: image enhancement, suppression of useless information, improvement of image usability, image sharpening, smoothing, and denoising.
[0090] Use spatial domain processing and its transformation to remove or reduce noise:
[0091] in, is the position / coordinate of the pixel in the image; is the position / coordinate in the convolution kernel, and the coordinate of the center point is (0,0); is the convolution kernel in The weight parameter on ; is with The corresponding image pixel value; In the picture The filtering / convolution result of the pixel.
[0092] By using frequency domain processing, low-frequency filtering is used to remove noise, and high-frequency filtering is used to enhance the signal to make the picture clear.
[0093] Furthermore, the fiber core splice box information input device performs table detection on the target table image to obtain the fiber core splice box information table image corresponding to each target OTDR test curve.
[0094] In one embodiment, a target detection model YOLO (You Only Look Once) may be used to implement target detection, target positioning, feature point detection, and table features using a YOLO neural network algorithm to obtain all core splice box information table images.
[0095] Furthermore, the fiber core splice closure information input device marks the area where each type of fiber core splice closure information is located in each fiber core splice closure information table picture, and obtains a plurality of table areas marked in each fiber core splice closure information table picture.
[0096] It should be noted that in machine vision and image processing, the area to be processed is outlined in the processed image in the form of a box, circle, ellipse, irregular polygon, etc., which is called a region of interest (ROI). To classify, extract and identify the fiber core splice box information, the fiber core length, attenuation value and fault point in each fiber core splice box information table image are all regions of interest, so the regions of interest need to be marked according to different types. Figure 7 , Figure 7 1 is a flow chart of marking a region of interest provided by the present invention, and the following steps are described.
[0097] Step S01: obtaining a table template, and then obtaining corresponding interest point templates according to different templates; Step S02: Matching the interest point template and the wiring picture; Step S03: Obtain the coordinate information of the four corners of the cell; Step S04: adjusting the scaling ratio and image of the ROI area template; L1: If the ratio is matched, proceed to the next step S06; L2: If not matched, then go to the next step S05; Step S05: scaling the image to fit the proportion of the points of interest and the proportion of the wiring image; Step S06: Calculate the matching ROI area according to the template and then mark it; Step S07: Output and mark all ROI regions, and mark the ROI regions.
[0098] The embodiment of the present invention generates corresponding fiber core splice box information tables based on multiple fiber core splice box information sets, further generates an initial table image by splicing, and accurately extracts the fiber core splice box information table image corresponding to each target OTDR test curve after image preprocessing and table detection. By marking and classifying different types of fiber core splice box information areas, accurate classification and aggregation of multiple fiber core splice box information are achieved, and finally target information sets corresponding to multiple types of fiber core splice box information are obtained, which greatly improves the systematicness and accuracy of data management, and then the classified information is entered into the fiber core splice box resource management system one by one, without manual intervention in the calculation and entry of the core splice box information, thereby improving the accuracy and efficiency of the core splice box information entry.
[0099] Furthermore, the target information sets corresponding to the core splice box information of the same type are obtained by classifying the multiple table areas marked in each core splice box information table picture, including: Based on each table area marked in each fiber core splicing box information table image, content extraction is performed to obtain text content corresponding to each table area marked in each fiber core splicing box information table image; Based on the text content corresponding to each table area marked in each fiber core splice box information table picture, the fiber core splice box information of the same type is integrated to obtain target information sets corresponding to multiple types of fiber core splice box information.
[0100] Specifically, the fiber core splice box information input device extracts content based on each table area marked in each fiber core splice box information table image, and obtains text content corresponding to each table area marked in each fiber core splice box table image.
[0101] Furthermore, the core splice closure information input device integrates the same type of core splice closure information based on the text content corresponding to each table area marked in each core splice closure information table image to obtain target information sets corresponding to multiple types of core splice closure information.
[0102] The embodiment of the present invention realizes efficient classification and aggregation of multiple fiber splice box information by extracting content from multiple table areas marked in each fiber splice box information table image and integrating text content corresponding to the same type of fiber splice box information, and finally obtains target information sets corresponding to the multiple types of core splice box information, which greatly improves the systematicness and accuracy of data management, and then enters the classified information one by one into the fiber splice box resource management system, without manual intervention in the calculation and entry of the core splice box information, thereby improving the accuracy and efficiency of the core splice box information entry.
[0103] Further, when content extraction is performed based on each table area marked in each fiber core splice box information table image to obtain text content corresponding to each table area marked in each fiber core splice box information table image, the following steps are performed for each table area: Get the cells in the table area; Performing text recognition on the cells to obtain the text to be checked corresponding to the table area; Perform text error correction on the text to be checked to obtain text content corresponding to the table area.
[0104] Specifically, the fiber core splice box information input device obtains a cell in the table area. It should be noted that there may be multiple cells in the table area, and a single cell is used as an example here, and multiple cells are also applicable.
[0105] Furthermore, the fiber core splice box information input device performs text recognition on the cells to obtain the text to be checked corresponding to the table area. In one embodiment, the machine learning technology of Convolutional Neural Networks (CNN) can be used to perform text recognition on the cells.
[0106] Furthermore, the fiber core splice box information input device performs text error correction on the text to be checked to obtain the text content corresponding to the table area. In one embodiment, the communication service dictionary can be used to correct the text to be checked.
[0107] The embodiment of the present invention performs detailed processing on each table area, first obtains the cells and performs text recognition, and then corrects the recognized text, thereby ensuring the accuracy and completeness of the extracted text content, effectively improving the accuracy of data extraction and processing, and providing high-quality data support for subsequent classification and aggregation of fiber core splice box information.
[0108] The fiber core splice box information entry device provided by the present invention is described below. The fiber core splice box information entry device described below and the fiber core splice box information entry method described above can be referred to each other.
[0109] Reference Figure 8 , Figure 8 It is a structural schematic diagram of the fiber core splicing box information input device provided by the present invention.
[0110] The core splice box information input device comprises: The acquisition module 810 is used to acquire multiple target optical time domain reflectometer OTDR test curves.
[0111] The core splice box information prediction module 820 is used to predict the core splice box information based on the multiple target OTDR test curves to obtain a core splice box information set corresponding to each target OTDR test curve; any of the core splice box information sets includes multiple types of core splice box information.
[0112] The classification module 830 is used to classify the same type of core splice closure information based on the multiple core splice closure information sets to obtain target information sets corresponding to the multiple types of core splice closure information.
[0113] The input module 840 is used to input the target information set corresponding to each type into the fiber core splice box resource management system one by one.
[0114] Furthermore, the acquisition module 810 is also used for: Get the initial OTDR test curve; Performing wavelet decomposition on the initial OTDR test curve to obtain wavelet decomposition coefficients at multiple levels; Performing threshold quantization processing on the wavelet decomposition coefficients of the multiple levels to obtain multiple wavelet decomposition coefficients after threshold quantization processing; Wavelet reconstruction is performed based on the multiple wavelet decomposition coefficients after the threshold quantization process to obtain the target OTDR test curve.
[0115] Furthermore, the fiber core splice box information prediction module 820 is also used for: Feature extraction is performed on the target OTDR test curve to obtain multiple types of OTDR curve feature data; the multiple types of OTDR curve feature data at least include first OTDR curve feature data, second OTDR curve feature data and third OTDR curve feature data; the first OTDR curve feature data includes the time of the rising point appearing in the descending curve and the time of the initial end of the curve; the second OTDR curve feature data includes the curve abscissa and the height of the trajectory before and after the event; the third OTDR curve feature data includes the height of the reflection time front end and the reflection peak; Inputting various types of OTDR curve feature data into a core splice box information prediction model, obtaining various types of core splice box information corresponding to the target OTDR test curve output by the core splice box information prediction model; the various types of core splice box information at least include core length, attenuation value and fault point location; the core splice box information prediction model is obtained by model training based on various types of OTDR curve feature samples and core splice box information labels corresponding to various types of OTDR curve feature samples.
[0116] Furthermore, the fiber core splice box information entry device is also used for: Get OTDR test curve samples; Extracting features from the OTDR test curve samples to obtain multiple types of OTDR curve feature samples; Label each type of OTDR curve characteristic sample to obtain a fiber core splicing box information label corresponding to each type of OTDR curve characteristic sample; Model training is performed based on each type of OTDR curve feature sample and the fiber core splice box information label corresponding to each type of OTDR curve feature sample to obtain the fiber core splice box information prediction model.
[0117] Furthermore, the classification module 830 is also used for: Splicing is performed based on multiple fiber core splicing box information sets to obtain a splicing information set; Based on the splicing information set, generate an initial table image; Performing image preprocessing on the initial table image to obtain a target table image; Performing table detection on the target table image to obtain a fiber core splicing box information table image corresponding to each target OTDR test curve; Marking the area where each type of fiber core splice box information is located in each fiber core splice box information table picture, to obtain a plurality of table areas marked in each fiber core splice box information table picture; Based on the multiple table areas marked in each fiber core splice closure information table picture, the fiber core splice closure information of the same type is classified to obtain target information sets corresponding to multiple types of fiber core splice closure information.
[0118] Furthermore, the classification module 830 is also used for: Based on each table area marked in each fiber core splicing box information table image, content extraction is performed to obtain text content corresponding to each table area marked in each fiber core splicing box information table image; Based on the text content corresponding to each table area marked in each fiber core splice box information table picture, the fiber core splice box information of the same type is integrated to obtain target information sets corresponding to multiple types of fiber core splice box information.
[0119] Furthermore, the classification module 830 is also used for: Get the cells in the table area; Performing text recognition on the cells to obtain the text to be checked corresponding to the table area; Perform text error correction on the text to be checked to obtain text content corresponding to the table area.
[0120] Fig. 9is a schematic diagram of the structure of the electronic device provided by the present invention, such as Fig. 9 As shown, the electronic device may include: a processor 910, a communication interface 920, a memory 930 and a communication bus 940, wherein the processor 910, the communication interface 920 and the memory 930 communicate with each other through the communication bus 940. The processor 910 may call the logic instructions in the memory 930 to execute the method for entering the information of the core splice box, which includes: obtaining multiple target optical time domain reflectometer OTDR test curves; predicting the information of the core splice box based on the multiple target OTDR test curves to obtain the core splice box information set corresponding to each target OTDR test curve; any of the core splice box information sets includes multiple types of core splice box information; based on the multiple core splice box information sets, classifying the core splice box information of the same type to obtain the target information sets corresponding to the multiple types of core splice box information; and entering the target information sets corresponding to each type into the core splice box resource management system one by one.
[0121] In addition, the logic instructions in the above-mentioned memory 930 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.
[0122] On the other hand, the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the fiber core splice box information entry method provided in the above-mentioned embodiments, the method including: obtaining multiple target optical time domain reflectometer OTDR test curves; predicting the fiber core splice box information based on the multiple target OTDR test curves to obtain a fiber core splice box information set corresponding to each target OTDR test curve; any of the fiber core splice box information sets includes multiple types of fiber core splice box information; based on the multiple fiber core splice box information sets, classifying according to the same type of core splice box information to obtain target information sets corresponding to the multiple types of core splice box information; and entering the target information sets corresponding to each type into the core splice box resource management system one by one.
[0123] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the fiber core splice box information entry method provided in the above-mentioned embodiments, the method comprising: obtaining multiple target optical time domain reflectometer OTDR test curves; predicting the fiber core splice box information based on the multiple target OTDR test curves to obtain a fiber core splice box information set corresponding to each target OTDR test curve; any of the fiber core splice box information sets including multiple types of fiber core splice box information; based on the multiple fiber core splice box information sets, classifying the fiber core splice box information according to the same type to obtain target information sets corresponding to the multiple types of core splice box information; and entering the target information sets corresponding to each type into the core splice box resource management system one by one.
[0124] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. Those of ordinary skill in the art may understand and implement the present invention without creative effort.
[0125] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for entering information of a fiber core splice box, It is characterized in that include: Obtain multiple target optical time domain reflectometer OTDR test curves; Perform fiber core splice box information prediction based on the multiple target OTDR test curves to obtain a fiber core splice box information set corresponding to each target OTDR test curve; Any of the fiber core splice box information sets includes multiple types of fiber core splice box information; Based on multiple fiber core splice box information sets, the fiber core splice box information of the same type is classified to obtain target information sets corresponding to multiple types of fiber core splice box information respectively; Enter the target information sets corresponding to each type into the fiber core splice box resource management system one by one.
2. The method for entering information of a core splice box according to claim 1, It is characterized in that When the core splice box information is predicted based on the multiple target OTDR test curves to obtain the core splice box information set corresponding to each target OTDR test curve, the following steps are performed for each target OTDR test curve: Feature extraction is performed on the target OTDR test curve to obtain multiple types of OTDR curve feature data; the multiple types of OTDR curve feature data at least include first OTDR curve feature data, second OTDR curve feature data and third OTDR curve feature data; the first OTDR curve feature data includes the time of the rising point appearing in the descending curve and the time of the initial end of the curve; the second OTDR curve feature data includes the curve abscissa and the height of the trajectory before and after the event; the third OTDR curve feature data includes the height of the reflection time front end and the reflection peak; Inputting various types of OTDR curve feature data into a core splice box information prediction model, obtaining various types of core splice box information corresponding to the target OTDR test curve output by the core splice box information prediction model; the various types of core splice box information at least include core length, attenuation value and fault point location; the core splice box information prediction model is obtained by model training based on various types of OTDR curve feature samples and core splice box information labels corresponding to various types of OTDR curve feature samples.
3. The method for entering information of a core splice box according to claim 2, It is characterized in that The core splice box information prediction model is trained in the following way: Get OTDR test curve samples; Extracting features from the OTDR test curve samples to obtain multiple types of OTDR curve feature samples; Label each type of OTDR curve characteristic sample to obtain a fiber core splicing box information label corresponding to each type of OTDR curve characteristic sample; Model training is performed based on each type of OTDR curve feature sample and the fiber core splice box information label corresponding to each type of OTDR curve feature sample to obtain the fiber core splice box information prediction model.
4. The method for entering information of a core splice box according to claim 1, It is characterized in that The target information sets corresponding to the core splice box information of the same type are obtained by classifying the multiple core splice box information sets based on the multiple core splice box information sets, including: Splicing is performed based on multiple fiber core splicing box information sets to obtain a splicing information set; Based on the splicing information set, generate an initial table image; Performing image preprocessing on the initial table image to obtain a target table image; Performing table detection on the target table image to obtain a fiber core splicing box information table image corresponding to each target OTDR test curve; Marking the area where each type of fiber core splice box information is located in each fiber core splice box information table picture, to obtain a plurality of table areas marked in each fiber core splice box information table picture; Based on the multiple table areas marked in each fiber core splice closure information table picture, the fiber core splice closure information of the same type is classified to obtain target information sets corresponding to multiple types of fiber core splice closure information.
5. The method for entering information of a core splice box according to claim 4, It is characterized in that The target information sets corresponding to the core splice box information of the same type are obtained by classifying the multiple table areas marked in each core splice box information table picture, including: Based on each table area marked in each fiber core splicing box information table image, content extraction is performed to obtain text content corresponding to each table area marked in each fiber core splicing box information table image; Based on the text content corresponding to each table area marked in each fiber core splice box information table picture, the fiber core splice box information of the same type is integrated to obtain target information sets corresponding to multiple types of fiber core splice box information.
6. The method for entering information of a core splice box according to claim 5, It is characterized in that When content extraction is performed based on each table area marked in each fiber core splicing box information table image to obtain text content corresponding to each table area marked in each fiber core splicing box information table image, the following steps are performed for each table area: Get the cells in the table area; Performing text recognition on the cells to obtain the text to be checked corresponding to the table area; Perform text error correction on the text to be checked to obtain text content corresponding to the table area.
7. The method for entering information of a core splice box according to any one of claims 1 to 6, It is characterized in that When acquiring multiple target OTDR test curves, perform the following steps for each target OTDR test curve: Get the initial OTDR test curve; Performing wavelet decomposition on the initial OTDR test curve to obtain wavelet decomposition coefficients at multiple levels; Performing threshold quantization processing on the wavelet decomposition coefficients of the multiple levels to obtain multiple wavelet decomposition coefficients after threshold quantization processing; Wavelet reconstruction is performed based on the multiple wavelet decomposition coefficients after the threshold quantization process to obtain the target OTDR test curve.
8. A fiber core splice box information entry device, It is characterized in that include: An acquisition module, used for acquiring multiple target optical time domain reflectometer OTDR test curves; A core splice box information prediction module, used to predict the core splice box information based on the multiple target OTDR test curves, and obtain a core splice box information set corresponding to each target OTDR test curve; Any of the fiber core splice box information sets includes multiple types of fiber core splice box information; A classification module, for classifying the same type of fiber core splice box information based on multiple fiber core splice box information sets to obtain target information sets corresponding to multiple types of fiber core splice box information respectively; The input module is used to input the target information set corresponding to each type into the fiber core splicing box resource management system one by one.
9. A device, comprising an electronic device, the electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the computer program, the steps of the method for entering information of a core splice box according to any one of claims 1 to 7 are implemented.
10. A medium comprising a non-transitory computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the steps of the method for entering information of a core splice box according to any one of claims 1 to 7 are implemented.