Method, system, device and medium for identifying underground cable external damage

Abstract

The present invention discloses a method, system, equipment and medium for identifying external damage of underground cables, which relates to the field of cable technology. Based on different damage conditions of road surface markings, the road surface cable nameplate identification image and the road surface key image are used for correction and identification, and then a target road surface identification image is generated. The target road surface identification image and the cable space data corresponding to the underground cable are used for mapping calculation to determine the corresponding cable code in the target road surface identification image. Using the cable code as an index, the cable static data set and the cable dynamic data set are extracted to construct a cable virtual space-time sequence model. At the same time, the current construction comprehensive information is used to construct a virtual space-time sequence model of construction external damage key points. The two models are combined to perform construction external damage point identification calculations to generate multiple single-frame label mapping values ​​of construction external damage points. Based on the single-frame label mapping value of the construction external damage point and the preset standard violent construction library, the external damage risk is identified and the external damage risk warning data is determined.

Description

Technical Field

[0001] The present invention relates to the field of cable technology, and in particular to a method, system, equipment and medium for identifying underground cable damage. Background Art

[0002] With economic and social development, the density of underground cables in urban distribution networks has increased, forming a complex underground cable infrastructure system. The operational status of these cables is crucial to the reliability and quality of urban power supply. However, underground cables are inherently concealed and enclosed, creating a complex operating environment. In recent years, external damage to cables caused by construction, theft, and modification has become a frequent problem, seriously affecting their safe and reliable operation.

[0003] Currently, underground cable damage monitoring methods use image acquisition, multi-type terminal acquisition, and document interaction as the main data sources. Each data source contains a large amount of information and operates in isolation. It requires manual, step-by-step extraction of information from different data sources. The search labor cost is high, and the cost of modifying these data sources is also high. It is impossible to link and retrieve multiple types of information such as underground cable technical parameters, change records, and damage records. In addition, underground cable damage risk warnings mainly rely on regular manual inspections, supplemented by video monitoring, vibration signals, and other means. On the one hand, there is the problem of manually processing video image information and vibration monitoring signals, and it is impossible to quickly link and calculate the damage risk. On the other hand, this type of information is mostly post-event signals and cannot provide active warnings before external forces damage. Therefore, existing underground cable damage identification methods rely on regular manual inspections and damage risk identification, which is labor-intensive and inefficient. The signals obtained are post-event signals, resulting in low accuracy of the identification results. Summary of the Invention

[0004] The present invention provides a method, system, device and medium for identifying underground cable external damage, which solves the technical problem that the existing method for identifying underground cable external damage relies on manual regular inspections and external damage risk identification, which is labor-intensive, inefficient and only obtains post-event signals, resulting in low accuracy of identification results.

[0005] The present invention provides a method for identifying an underground cable breakage, comprising:

[0006] Acquire a cable nameplate identification image and a road surface key image as an initial road surface recognition image, perform correction and recognition processing on the initial road surface recognition image, and generate a target road surface recognition image;

[0007] Performing a mapping calculation using the target road surface recognition image and cable space data corresponding to the underground cable to determine the cable code corresponding to the underground cable;

[0008] The cable static data set and the cable dynamic data set corresponding to the cable coding are used to construct a model to generate a cable virtual space-time sequence model and a construction external failure key point virtual space-time sequence model;

[0009] The cable virtual space-time sequence model and the construction external break key point virtual space-time sequence model are used to identify the construction external break points, and generate a plurality of single-frame label mapping values ​​of the construction external break points corresponding to the underground cable;

[0010] External damage risk identification is performed based on the single-frame label mapping value of the construction external damage point and a preset standard violent construction library to determine external damage risk warning data corresponding to the underground cable.

[0011] Optionally, the step of acquiring the cable nameplate identification image and the road surface key image as an initial road surface recognition image, performing correction and recognition processing on the initial road surface recognition image, and generating a target road surface recognition image includes:

[0012] Obtaining cable pavement images and cable dynamic data sets corresponding to underground cables in different scenes, performing pavement identification on the initial pavement recognition images to generate initial pavement recognition data;

[0013] If the initial road surface recognition data is a road cable nameplate identification image, correcting the road cable nameplate identification image according to a preset nameplate identification correction image calculation formula to generate a first target road surface recognition image;

[0014] The calculation formula for the preset nameplate logo correction image is:

[0015] [x0 y0 1] T =A0[u d v d 1] T ;

[0016]

[0017]

[0018] Where A0 is the orientation element in the lens; k1 is the radial distortion coefficient; r (0,1) is the Euclidean distance between the distortion center and the distorted image point in the panoramic image; (x0, y0) is the physical coordinate corresponding to the image point of the cable nameplate under the influence of distortion; (u d ,v d ) is the pixel coordinate of the cable nameplate identification image point under the influence of distortion; (x (0,1) ,y (0,1)) is the physical coordinate of the cable nameplate identification image point under the correction of no distortion; (x1, y1) is the map coordinate of the cable nameplate identification image point; C cm is the exterior orientation element of the panoramic image, where m ij C cm The element in the i-th row and j-th column of the matrix; s is the scale factor;

[0019] If the initial road surface recognition data is a road surface key image, the road surface key image is corrected according to a preset template image matching calculation formula to generate a second target road surface recognition image;

[0020] The preset template image matching calculation formula is:

[0021]

[0022]

[0023] N i =b i×8 ×2 0 +b i×8+1 ×2 1 +…+b i×8+7 ×2 7 ,N 16 =N0;

[0024]

[0025] H=h0×2 0 +…+h 15 ×2 15 ;

[0026] Where D i+1 is the key image feature descriptor at position i+1; D i is the key image feature descriptor at position i; D 127 is the 127th key image feature descriptor; J0(x0,y0) is the road key image, (x0,y0) is the physical coordinate corresponding to the midpoint of the road key image; b i is the 128-bit binary code of the i-th bit; AD i is the i-th key point descriptor based on the scale-invariant feature transform matching algorithm; M is the set threshold used to calculate the binary code of the key point descriptor of the scale-invariant feature transform matching algorithm; N i The value calculated for the i-th sub-binary code group; h i is the binary code corresponding to the ith sub-binary code group; h 15 is the binary code corresponding to the 15th sub-binary code group; H is the hash address code of the key point descriptor.

[0027] Optionally, the target road surface identification image is a first target road surface recognition image; and the step of performing mapping calculation using the target road surface recognition image and the cable space data corresponding to the underground cable to determine the cable code corresponding to the underground cable includes:

[0028] Performing a mapping calculation using the first target road surface recognition image and the cable space data corresponding to the underground cable according to a preset first mapping calculation formula to determine a plurality of cable coding probabilities corresponding to the underground cable;

[0029] The preset first mapping calculation formula is:

[0030] U0=[u i ,u j ],u i,j ∈I1;

[0031] U l =ω l T U l-1 +b l ;

[0032] Y=ω N T U N-1 +b N ;

[0033]

[0034] Where U0 is the neural network input layer, where the elements of U0 are the coordinate values ​​of the nameplate correction image map (u i ,u j ); I1 is the first target road surface recognition image; U l is the hidden layer value of the neural network layer l; ω l is the weight between the l-1th layer and the lth layer of the neural network; ω N is the weight between the N-1th layer and the Nth layer of the neural network; T represents the matrix; b l is the bias value of the neural network layer l; b N is the bias value of the Nth layer of the neural network; Y is the output value of the Nth layer of the neural network; N is the total number of neural network layers; Y(i) is the classification output probability value calculated by the normalized exponential function classifier of the i-th neural network output value; e is a natural constant; e Y(i) is e to the power of Y(i);

[0035] The code corresponding to the maximum value of the cable code probabilities is used as the cable code corresponding to the underground cable.

[0036] Optionally, the target road surface identification image is a second target road surface recognition image; and the step of performing mapping calculation using the target road surface recognition image and the cable space data corresponding to the underground cable to determine the cable code corresponding to the underground cable includes:

[0037] Using the second target road surface recognition image and the cable space data corresponding to the underground cable to perform a matching calculation on the key identifiers of the surrounding environment according to a preset second mapping calculation formula, and determining multiple similarities corresponding to the underground cable;

[0038] The preset second mapping formula is:

[0039]

[0040] Where H is the hash address code of the key point descriptor of the second target road surface recognition image; is the hash address code of the key sign descriptor of the surrounding environment of the i-th cable in the database; L M is the total number of cables in the database; E is the similarity;

[0041] The code corresponding to the maximum value of the similarities is used as the cable code corresponding to the underground cable.

[0042] Optionally, the cable static data set includes a cable technical parameter group, a cable change record information group, and an external force damage alarm information group; the step of constructing a model using the cable static data set and the cable dynamic data set corresponding to the cable code to generate a cable virtual space-time sequence model and a construction external damage key point virtual space-time sequence model includes:

[0043] The cable technical parameter group is used to determine the initial position of the cable insulation sheath, and generate the cable insulation sheath position;

[0044] The cable change record information group and the external force damage alarm information group are used in combination with the cable insulation sheath position to construct a cable insulation sheath movement trajectory to generate cable insulation sheath movement trajectory data;

[0045] Modeling the cable insulation sheath movement trajectory data to generate a cable virtual space-time sequence model;

[0046] Using the cable dynamic data set to simulate and predict the motion trajectory of the construction equipment to generate construction equipment motion trajectory data;

[0047] Calculate the distance change between the movement trajectory data of the construction equipment and the corresponding movement trajectory data of the cable insulation sheath to generate multiple change distances;

[0048] Selecting key points corresponding to the change distances that are smaller than a preset threshold value to generate construction external breaking key points;

[0049] The position data corresponding to the key points of external destruction during construction are used to construct a model, and a virtual space-time sequence model of the key points of external destruction during construction is generated.

[0050] Optionally, the cable virtual space-time sequence model and the construction external failure key point virtual space-time sequence model include multiple time and space fields; the step of using the cable virtual space-time sequence model and the construction external failure key point virtual space-time sequence model to identify construction external failure points and generate multiple single-frame label mapping values ​​of construction external failure points corresponding to the underground cable includes:

[0051] Calculating the vertical Euclidean distance between each of the key construction external damage points and the cable insulation sheath corresponding to the cable virtual space-time sequence model in each of the time-space domains, and generating a vertical Euclidean distance set corresponding to the time-space domain;

[0052] The vertical Euclidean distance set is used as a single-frame label mapping value of the construction outer breaking point corresponding to the underground cable.

[0053] Optionally, the step of identifying external damage risk based on the single-frame label mapping value of the construction external damage point and a preset standard violent construction library to determine external damage risk warning data corresponding to the underground cable includes:

[0054] Classify multiple feature vectors corresponding to the single-frame label mapping values ​​of the construction external breaking points through a preset classifier to generate a set of violent construction actions;

[0055] The set of violent construction actions is compared with a preset standard violent construction library to identify external damage risks and generate an external force damage risk value;

[0056] If the external force damage risk value is greater than a preset risk threshold, the position corresponding to the external force damage risk value is used as an external damage risk warning position;

[0057] The external damage risk warning corresponding to the preset risk threshold and the external damage risk warning position are used to construct external damage risk warning data corresponding to the underground cable.

[0058] The present invention also provides an underground cable damage identification system, comprising:

[0059] a target road surface recognition image generation module, configured to obtain a cable nameplate identification image and a road surface key image as an initial road surface recognition image, perform correction and recognition processing on the initial road surface recognition image, and generate a target road surface recognition image;

[0060] a cable code determination module, configured to perform mapping calculation using the target road surface recognition image and the cable space data corresponding to the underground cable to determine the cable code corresponding to the underground cable;

[0061] A module for generating a cable virtual space-time sequence model and a construction external failure key point virtual space-time sequence model is used to construct a model using the cable static data set and the cable dynamic data set corresponding to the cable coding to generate a cable virtual space-time sequence model and a construction external failure key point virtual space-time sequence model;

[0062] A single-frame label mapping value generation module for construction external break points is used to identify construction external break points using the cable virtual space-time sequence model and the construction external break key point virtual space-time sequence model, and generate a plurality of single-frame label mapping values ​​of construction external break points corresponding to the underground cable;

[0063] The external damage risk warning data determination module is used to identify external damage risks based on the single-frame label mapping value of the construction external damage point and a preset standard violent construction library, and determine the external damage risk warning data corresponding to the underground cable.

[0064] The present invention also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of implementing any of the above-mentioned methods for identifying underground cable damage.

[0065] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed, it implements any of the above-mentioned methods for identifying underground cable damage.

[0066] It can be seen from the above technical solutions that the present invention has the following advantages:

[0067] The present invention obtains a cable nameplate identification image and a key road surface image as an initial road surface recognition image, performs correction and recognition processing on the initial road surface recognition image, and generates a target road surface recognition image. The target road surface recognition image and the cable spatial data corresponding to the underground cable are used for mapping calculation to determine the cable code corresponding to the underground cable. The cable static data set and the cable dynamic data set corresponding to the cable code are used to construct a model to generate a cable virtual space-time sequence model and a virtual space-time sequence model of key construction external damage points. The cable virtual space-time sequence model and the virtual space-time sequence model of key construction external damage points are used to identify construction external damage points, generating multiple single-frame label mapping values ​​corresponding to the underground cable. Based on the single-frame label mapping values ​​of the construction external damage points and a preset standard violent construction library, external damage risk identification is performed to determine external damage risk warning data corresponding to the underground cable. This solves the technical problem that existing underground cable external damage identification methods rely on manual regular inspections and external damage risk identification, which is labor-intensive, inefficient, and only produces ex post signals, resulting in low recognition accuracy. A virtual space-time sequence model of cable and construction external damage key points is established. By extracting the characteristics of the construction external damage key points and predicting their movement trajectories, the risk of cable external damage is identified to achieve active early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0069] Figure 1 A flowchart of a method for identifying underground cable damage provided in Example 1 of the present invention;

[0070] Figure 2 A flowchart of a method for identifying underground cable damage provided in the second embodiment of the present invention;

[0071] Figure 3 A flowchart of a method for identifying an underground cable breakage provided in the second embodiment of the present invention;

[0072] Figure 4 This is a structural block diagram of an underground cable damage identification system provided in Example 3 of the present invention. DETAILED DESCRIPTION

[0073] The embodiments of the present invention provide a method, system, device and medium for identifying underground cable external damage, which is used to solve the technical problem that the existing method for identifying underground cable external damage requires manual regular inspections and external damage risk identification, which has a large workload and low work efficiency, and obtains post-event signals, resulting in low accuracy of identification results.

[0074] In order to make the purpose, features, and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only 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 making creative work are within the scope of protection of the present invention.

[0075] See also Figure 1 , Figure 1 This is a flowchart of the steps of a method for identifying underground cable damage provided in Example 1 of the present invention.

[0076] A method for identifying an underground cable failure is provided in Example 1 of the present invention, comprising:

[0077] Step 101: Acquire a cable nameplate identification image and a road surface key image as an initial road surface recognition image, perform correction and recognition processing on the initial road surface recognition image, and generate a target road surface recognition image.

[0078] In an embodiment of the present invention, a cable nameplate image or a road surface key image is acquired based on the damage status of road surface markings, and a target road surface recognition image is generated through correction and recognition. Specifically, the cable nameplate image and the road surface key image are acquired as initial road surface recognition images, and road surface marking recognition is performed on the initial road surface recognition image to generate initial road surface recognition data. If the initial road surface recognition data is a road surface cable nameplate image, the road surface cable nameplate image is corrected according to a preset nameplate marking correction image calculation formula to generate a first target road surface recognition image. If the initial road surface recognition data is a road surface key image, the road surface key image is corrected according to a preset template image matching calculation formula to generate a second target road surface recognition image.

[0079] Step 102: Perform mapping calculation using the target road surface recognition image and the cable space data corresponding to the underground cables to determine the cable codes corresponding to the underground cables.

[0080] In an embodiment of the present invention, if the target road surface identification image is a first target road surface identification image, a mapping calculation is performed using the first target road surface identification image and the cable spatial data corresponding to the underground cable according to a preset first mapping calculation formula to determine multiple cable code probabilities corresponding to the underground cable. The code corresponding to the maximum value among the cable code probabilities is used as the cable code corresponding to the underground cable. If the target road surface identification image is a second target road surface identification image, a matching calculation is performed using the second target road surface identification image and the cable spatial data corresponding to the underground cable according to a preset second mapping calculation formula to determine multiple similarities corresponding to the underground cable. The code corresponding to the maximum value among the similarities is used as the cable code corresponding to the underground cable.

[0081] Step 103: Use the cable static data set and the cable dynamic data set corresponding to the cable coding to construct a model, and generate a cable virtual space-time sequence model and a construction external damage key point virtual space-time sequence model.

[0082] In an embodiment of the present invention, a virtual spatiotemporal sequence model of cables and key construction failure points is constructed using a static cable data set corresponding to cable coding and current construction data, namely a dynamic cable data set. The static cable data set includes a cable technical parameter set, a cable change record information set, and an external force damage alarm information set. The specific model construction process is as follows: the initial position of the cable insulation sheath is determined using the cable technical parameter set to generate the cable insulation sheath position. The cable change record information set and the external force damage alarm information set are combined with the cable insulation sheath position to construct the cable insulation sheath movement trajectory, generating cable insulation sheath movement trajectory data. The cable insulation sheath movement trajectory data is used to construct a model to generate a virtual spatiotemporal sequence model of the cable. The dynamic cable data set is used to simulate and predict the movement trajectory of construction equipment to generate construction equipment movement trajectory data. The distance change between the construction equipment movement trajectory data and the corresponding cable insulation sheath movement trajectory data is calculated to generate multiple change distances. Key points corresponding to change distances less than a preset threshold are selected to generate construction failure key points. The position data corresponding to the construction failure key points are used to construct a model to generate a virtual spatiotemporal sequence model of the construction failure key points.

[0083] Step 104: Use the cable virtual space-time sequence model and the construction external damage key point virtual space-time sequence model to identify the construction external damage points, and generate multiple construction external damage point single-frame label mapping values ​​corresponding to the underground cable.

[0084] In this embodiment of the present invention, the cable virtual space-time sequence model and the construction damage key point virtual space-time sequence model include multiple space-time domains. The vertical Euclidean distance between each construction damage key point and the cable insulation sheath corresponding to the cable virtual space-time sequence model in each space-time domain is calculated to generate a set of vertical Euclidean distances corresponding to the space-time domain. The set of vertical Euclidean distances is used as the single-frame label mapping value for the construction damage point corresponding to the underground cable.

[0085] Step 105: identify external damage risk based on the single-frame label mapping value of the construction external damage point and the preset standard violent construction library, and determine the external damage risk warning data corresponding to the underground cable.

[0086] In an embodiment of the present invention, a set of violent construction actions is generated by classifying multiple feature vectors corresponding to single-frame label mapping values ​​of preset classified construction external damage points. This set of violent construction actions is then compared with a preset standard violent construction library to identify external damage risks, generating an external damage risk value. If the external damage risk value exceeds a preset risk threshold, the location corresponding to the external damage risk value is used as an external damage risk warning location. External damage risk warning data corresponding to underground cables is constructed using the external damage risk warning and external damage risk warning locations corresponding to the preset risk threshold.

[0087] In an embodiment of the present invention, a cable nameplate identification image and a key road surface image are obtained as an initial road surface recognition image, and the initial road surface recognition image is corrected and recognized to generate a target road surface recognition image. A mapping calculation is performed using the target road surface recognition image and the cable spatial data corresponding to the underground cable to determine the cable code corresponding to the underground cable. A model is constructed using the cable static data set and the cable dynamic data set corresponding to the cable code to generate a virtual space-time sequence model of the cable and a virtual space-time sequence model of key construction external damage points. The cable virtual space-time sequence model and the virtual space-time sequence model of key construction external damage points are used to identify construction external damage points, generating multiple single-frame label mapping values ​​corresponding to the underground cable. Based on the single-frame label mapping values ​​of the construction external damage points and a preset standard violent construction library, external damage risk identification is performed to determine external damage risk warning data corresponding to the underground cable. This method solves the technical problem that existing underground cable external damage identification methods rely on manual regular inspections and external damage risk identification, which is labor-intensive, inefficient, and produces only post-event signals, resulting in low recognition accuracy. A virtual space-time sequence model of cable and construction external damage key points is established. By extracting the characteristics of the construction external damage key points and predicting their movement trajectories, the risk of cable external damage is identified to achieve active early warning.

[0088] See also Figure 2 , Figure 2 This is a flowchart of the steps of a method for identifying underground cable damage provided in Example 2 of the present invention.

[0089] Another method for identifying underground cable damage provided by Example 2 of the present invention includes:

[0090] Step 201: Acquire a cable nameplate identification image and a road surface key image as an initial road surface recognition image, perform correction and recognition processing on the initial road surface recognition image, and generate a target road surface recognition image.

[0091] Furthermore, step 201 may include the following sub-steps S11-S13:

[0092] S11. Obtain a cable nameplate identification image and a road surface key image according to different scenes as an initial road surface recognition image, perform road surface identification on the initial road surface recognition image, and generate initial road surface recognition data.

[0093] S12: If the initial road surface recognition data is a road cable nameplate identification image, correct the road cable nameplate identification image according to a preset nameplate identification correction image calculation formula to generate a first target road surface recognition image.

[0094] The calculation formula for the preset nameplate logo correction image is:

[0095] [x0 y0 1] T =A0[u d v d 1] T ;

[0096]

[0097]

[0098] Where A0 is the orientation element in the lens; k1 is the radial distortion coefficient; r (0,1) is the Euclidean distance between the distortion center and the distorted image point in the panoramic image; (x0, y0) is the physical coordinate corresponding to the image point of the cable nameplate under the influence of distortion; (u d ,v d ) is the pixel coordinate of the cable nameplate identification image point under the influence of distortion; (x (0,1) ,y (0,1) ) is the physical coordinate of the cable nameplate identification image point under the correction of no distortion; (x1, y1) is the map coordinate of the cable nameplate identification image point; C cm is the exterior orientation element of the panoramic image, where m ij C cm The element in the i-th row and j-th column of the matrix; s is the scaling factor.

[0099] S13: If the initial road surface recognition data is a road surface key image, the road surface key image is corrected according to a preset template image matching calculation formula to generate a second target road surface recognition image.

[0100] The preset template image matching calculation formula is:

[0101]

[0102]

[0103] N i =b i×8 ×2 0 +b i×8+1 ×2 1 +…+b i×8+7 ×2 7 ,N 16 =N0;

[0104]

[0105] H=h0×2 0 +…+h 15 ×2 15 ;

[0106] Where D i+1 is the key image feature descriptor at position i+1; D i is the key image feature descriptor at position i; D 127 is the 127th key image feature descriptor; J0(x0,y0) is the road key image, (x0,y0) is the physical coordinate corresponding to the midpoint of the road key image; b i is the 128-bit binary code of the i-th bit; AD i is the key point descriptor based on SIFT (Scale Invariant Feature Transform Matching Algorithm) at the i-th position; M is the set threshold used to calculate the binary code of the key point descriptor of the scale invariant feature transform matching algorithm; N i The value calculated for the ith group of binary code (the ith group of binary code is represented by b i ×8, b i ×8+1,b i ×8+2,…,b i ×8+7); h i is the binary code corresponding to the ith sub-binary code group; h 15 is the binary code corresponding to the 15th sub-binary code group; H is the hash address code of the key point descriptor.

[0107] In an embodiment of the present invention, the initial road surface recognition image includes multiple cable nameplate images and a key road surface image. The initial road surface recognition image and a set of cable dynamic data corresponding to the underground cables are obtained. Road surface recognition data is then processed on the initial road surface recognition image to determine whether the road surface markings corresponding to the underground cables are damaged. If the road surface markings are not damaged, the initial road surface recognition data is the road surface cable nameplate image. The preprocessed first target road surface recognition image is obtained by correcting, extracting, and matching the road surface cable nameplate image. Specifically, the image data corresponding to the road surface cable nameplate image is input into a preset nameplate correction image calculation formula to obtain the first target road surface recognition image.

[0108] If the road markings are damaged, the initial road recognition data is the road key image. Key features in the road key image include road signs, landmark buildings, and so on. The road key image is corrected, extracted, and matched to obtain a pre-processed second target road recognition image. Specifically, the image data corresponding to the road key image is input into a preset template image matching calculation formula to obtain the second target road recognition image.

[0109] Step 202: Perform mapping calculation using the target road surface recognition image and the cable space data corresponding to the underground cables to determine the cable codes corresponding to the underground cables.

[0110] Furthermore, the target road surface identification image is a first target road surface recognition image, and step 202 may include the following sub-steps S21-S22:

[0111] S21. Perform mapping calculation using the first target road surface recognition image and the cable space data corresponding to the underground cables according to a preset first mapping calculation formula to determine multiple cable coding probabilities corresponding to the underground cables.

[0112] The preset first mapping calculation formula is:

[0113] U0=[u i ,u j ],u i,j ∈I1;

[0114] U l =ω l T U l-1 +b l ;

[0115] Y=ω N T U N-1 +b N ;

[0116]

[0117] Where U0 is the neural network input layer, where the elements of U0 are the coordinate values ​​of the nameplate correction image map (u i ,u j ); I1 is the first target road surface recognition image; U l is the hidden layer value of the neural network layer l; ω l is the weight between the l-1th layer and the lth layer of the neural network; ω N is the weight between the N-1th layer and the Nth layer of the neural network; T represents the matrix; b l is the bias value of the neural network layer l; b N is the bias value of the Nth layer of the neural network; Y is the output value of the Nth layer of the neural network; N is the total number of neural network layers; Y(i) is the classification output probability value calculated by the normalized exponential function classifier of the i-th neural network output value; e is a natural constant; e Y(i) is e to the power of Y(i).

[0118] S22. The code corresponding to the maximum value of the cable code probability is used as the cable code corresponding to the underground cable.

[0119] In this embodiment of the present invention, a normalized exponential function classifier, or Softmax classifier, is used to compare the first target road surface recognition image and the cable spatial data corresponding to the underground cable. The corrected first target road surface recognition image and a neural network trained based on a standard image corresponding to a preset first mapping calculation formula are used to perform a comparison calculation. The cable code with the highest probability is selected as the cable code corresponding to the underground cable. Specifically, the cable code is used as the index header, and all cable marker information groups, including the cable marker shape, the coordinates of the corresponding marker's geographic location, and key identifiers of the surrounding environment of the geographic location, are mapped to obtain a cable code that uniquely identifies the cable.

[0120] Furthermore, the target road surface identification image is a second target road surface recognition image, and step 202 may include the following sub-steps S31-S32:

[0121] S31. Use the second target road surface recognition image and the cable space data corresponding to the underground cable to perform matching calculation on the key identifiers of the surrounding environment according to a preset second mapping calculation formula to determine multiple similarities corresponding to the underground cable.

[0122] The preset second mapping formula is:

[0123]

[0124] Where H is the hash address code of the key point descriptor of the second target road surface recognition image; is the hash address code of the key sign descriptor of the surrounding environment of the i-th cable in the database; L M is the total number of cables in the database; E is the similarity, and the larger the value, the greater the possibility that it belongs to the i-th cable.

[0125] S32. Use the code corresponding to the maximum value of the similarity as the cable code corresponding to the underground cable.

[0126] In this embodiment of the present invention, the second target road surface recognition image and the cable spatial data corresponding to the underground cable are matched and calculated using a preset second mapping calculation formula to obtain multiple similarities corresponding to the underground cable. Specifically, using a similarity calculation method, the second target road surface recognition image is matched with the key identifiers of the surrounding environment in the cable spatial information database, and the cable with the highest similarity is selected and the corresponding coding information, i.e., the cable code, is output.

[0127] Step 203: construct a model using the cable static data set and the cable dynamic data set corresponding to the cable coding to generate a cable virtual space-time sequence model and a construction external damage key point virtual space-time sequence model.

[0128] Furthermore, the cable static data set includes a cable technical parameter group, a cable change record information group, and an external force damage alarm information group. Step 203 may include the following sub-steps S41-S47:

[0129] S41. Use the cable technical parameter group to determine the initial position of the cable insulation sheath and generate the cable insulation sheath position.

[0130] S42. Use the cable change record information group and the external force damage alarm information group, and combine the cable insulation sheath position to construct the cable insulation sheath movement trajectory to generate cable insulation sheath movement trajectory data.

[0131] S43. Model the cable insulation sheath movement trajectory data to generate a cable virtual space-time sequence model.

[0132] S44. Use the cable dynamic data set to simulate and predict the motion trajectory of the construction equipment to generate the motion trajectory data of the construction equipment.

[0133] S45. Calculate the distance change between the construction equipment movement trajectory data and the corresponding cable insulation sheath movement trajectory data to generate multiple change distances.

[0134] S46. Select key points corresponding to change distances less than a preset threshold value to generate construction external breaking key points.

[0135] S47. Use the location data corresponding to the key points of external destruction during construction to construct a model, and generate a virtual space-time sequence model of the key points of external destruction during construction.

[0136] The preset threshold refers to the critical value set based on actual needs for selecting key points of external failure during construction.

[0137] In an embodiment of the present invention, the cable code is input into the data integration unit, and the cable code is searched and calculated in the cable static database and the cable dynamic database, and the cable technical parameter group (including cable length, cable depth, cable GIS distribution direction), the cable change record information group (including new addition, modification and movement information) and the external force damage alarm information group (vibration characteristics, light characteristics and video damage information, etc.) are extracted respectively, that is, the cable static data set is extracted. The cable calculation parameter group already contains the general parameters of the cable, as well as the cable length, cable cross-sectional area, cable spatial information, etc. Therefore, the preliminary position of the cable insulation sheath can be preliminarily determined through the cable technical parameter group to obtain the cable insulation sheath position. Combined with the cable change record information group and the external force damage alarm information group, the movement trajectory of the cable insulation sheath under the influence of construction and cable change is further considered to generate a cable virtual space-time sequence model ML (V, E, T), which represents the i-th cable insulation sheath key point information V at time T. i , the line connecting the i-th and j-th key points constitutes the information E of the cable edge ij The cable insulation movement trajectory data is generated by projecting the information onto the real image through information overlay. The cable virtual spatiotemporal sequence model ML(V,E,T) takes into account the possible movement of the cable insulation under cable changes and external force damage.

[0138] A cable dynamic data set is input through manual interaction. The data set includes construction method (SF), construction depth (SH), construction coverage (SS), construction time (ST), and ground subsidence (SD) during construction. The motion trajectories of construction equipment during the construction process within the cable dynamic data set are analyzed to generate construction equipment motion trajectory data. The distance between the construction equipment and the cable insulation is calculated based on single-frame image information per unit time. The distance change between the construction equipment and the cable insulation over multiple frames of image information over a period of time is compared. The distance change between the construction equipment motion trajectory data and the corresponding cable insulation movement trajectory data is calculated to generate multiple change distances. Based on a preset threshold, key points with the highest potential for cable damage are selected. Specifically, key points with change distances less than the preset threshold are selected to generate construction external damage key points. Construction external damage key points are determined by analyzing the motion trajectories of construction equipment during construction. Points are identified as construction external damage key points if the distance between the outer edge of the construction equipment and the cable insulation in a single-frame image is within the monitoring risk threshold and if the distance decreases continuously across multiple frames of image comparison. Summarize the edge positions of these key points at each moment, and the position information is to obtain the position data corresponding to the key points of construction external destruction, and form the virtual space-time sequence model MS (V, E, T) of the key points of construction external destruction, which represents the equipment point information V of the key point i at time T. i and the position E of the line connecting point i and point j ij .

[0139] Step 204: Use the cable virtual space-time sequence model and the construction external damage key point virtual space-time sequence model to identify the construction external damage points, and generate multiple single-frame label mapping values ​​of the construction external damage points corresponding to the underground cable.

[0140] Furthermore, the cable virtual space-time sequence model and the construction external failure key point virtual space-time sequence model include multiple time and space domains, and step 204 may include the following sub-steps S51-S52:

[0141] S51. Calculate the vertical Euclidean distance between each construction external damage key point and the cable insulation sheath corresponding to the cable virtual space-time sequence model in each time-space domain, and generate a vertical Euclidean distance set corresponding to the time-space domain.

[0142] S52. Use the vertical Euclidean distance set as a single-frame label mapping value of the construction outer breaking point corresponding to the underground cable.

[0143] In the embodiment of the present invention, the cable virtual space-time sequence model ML(V,E,T) and the construction external damage key point virtual space-time sequence model MS(V,E,T) are combined to calculate the vertical Euclidean distance set from each construction external damage key point to the cable insulation sheath in each space-time domain, and use this as the single-frame label mapping value of the construction external damage point. Specifically, d=sqrt((x1-x2) 2 +(y1-y2) 2 +(z1-z2) 2 ) The Euclidean distance calculation formula calculates the vertical distance from each external breaking key point to the cable insulation outer sheath, that is, the shortest Euclidean distance from the point to the edge.

[0144] Step 205: Classify the multiple feature vectors corresponding to the single-frame label mapping values ​​of the construction external breaking points through a preset classifier to generate a set of violent construction actions.

[0145] In this embodiment of the present invention, the preset classifier is a standard Softmax classifier. Using the standard Softmax classifier, i.e., a classifier trained based on violent construction actions within the standard safety distance red line, the feature vectors corresponding to the label mapping values ​​of single frames of the construction point outside the construction area are classified to obtain a set of violent construction actions.

[0146] Step 206: The set of violent construction actions is compared with a preset standard violent construction library to identify external damage risks, and an external force damage risk value is generated.

[0147] In this embodiment of the present invention, the pre-set standard violent construction library refers to a database, established based on actual needs, that stores external force damage risk values ​​corresponding to single-frame label mappings of different construction external damage points. The single-frame label mappings of construction external damage points corresponding to a set of violent construction actions are compared and analyzed with the standard violent construction library to obtain the corresponding external force damage risk values ​​for underground cables.

[0148] Step 207: If the external force damage risk value is greater than the preset risk threshold, the position corresponding to the external force damage risk value is used as an external damage risk warning position.

[0149] In this embodiment of the present invention, the preset risk thresholds refer to multiple critical values ​​corresponding to different risk types, each of which is assigned a corresponding external damage risk warning type. When the external damage risk value exceeds any of the preset risk thresholds, it indicates that the underground cable has been damaged. In this case, the location corresponding to the external damage risk value is used as the external damage risk warning location.

[0150] Step 208: Use the external damage risk warning and external damage risk early warning positions corresponding to the preset risk threshold to construct external damage risk early warning data corresponding to the underground cable.

[0151] In this embodiment of the present invention, the external damage risk warning corresponding to the external force damage risk value greater than the preset risk threshold is used as the external damage risk warning obtained in the current identification of the underground cable. The external damage risk warning and the external damage risk warning location are used to construct the external damage risk warning data corresponding to the underground cable.

[0152] In the embodiment of the present invention, Figure 3As shown, the underground cable damage identification method of the present invention can be implemented by an image recognition unit, a data integration unit, and a data display unit. The image recognition unit performs pavement marking recognition correction on the cable-pavement image and performs a mapping calculation to determine the cable code corresponding to the underground cable. When performing pavement marking recognition on the cable-pavement image, two scenarios may occur. In scenario 1, the pavement marking is intact. In this case, the initial pavement recognition data is the pavement cable nameplate image. To account for lens distortion, the pavement cable nameplate image is corrected according to a preset nameplate correction image calculation formula to generate a first target pavement recognition image. Then, a mapping calculation is performed using the first target pavement recognition image and the cable spatial data corresponding to the underground cable according to a preset first mapping calculation formula, selecting the cable code with the highest probability as the cable code corresponding to the underground cable. In scenario 2, the pavement marking is damaged. In this case, the initial pavement recognition data is the pavement key image. The pavement key image is corrected according to a preset template image matching calculation formula to generate a second target pavement recognition image. The second target pavement recognition image is then matched with the surrounding environment key markers in the cable spatial information database, and the cable code with the highest similarity is selected as the cable code corresponding to the underground cable. The cable code is input into the data integration unit, which extracts the cable technical parameter group from the static cable database. The cable change record information group and the cable external damage alarm information group are extracted from the dynamic cable database. The data display unit generates a cable virtual space-time sequence model ML(V, E, T) based on the cable technical parameter group, cable change record information group, and cable external damage alarm information group. Comprehensive construction information is input, including construction method SF, construction depth SH, construction coverage SS, construction time ST, and ground subsidence SD during construction. A virtual space-time sequence model MS(V, E, T) of the construction external damage key point is calculated. The Euclidean distance from the construction external damage key point to the cable insulation sheath is then calculated to determine the external damage risk value A corresponding to the underground cable. A determination is made as to whether the external damage risk value A is less than or equal to the preset risk threshold B. If not, an external damage risk warning is issued, and the external damage risk warning location Po(x, y) is displayed. Using AR visual mobile devices, inspectors can conveniently complete image acquisition and preprocessing of cable laying sections, calculating and extracting cable identification code information for various scenarios. The unique cable identification code is used as the index entry point to complete the collaborative information retrieval and unified visualization presentation of the GIS equipment ledger system, multi-channel monitoring terminals (vibration monitoring, cameras, explosion-proof blanket monitoring, etc.) and change record system. At the same time, considering the changes in the underground cable operating environment and the movement of construction tools, a virtual spatiotemporal sequence model of cables and key points of external damage during construction is established. By extracting the characteristics of key points of external damage during construction and predicting their movement trajectories, the risk of cable external damage can be identified to provide active early warning. This reduces the workload of operation and maintenance personnel in regular inspections and external damage risk identification, and improves the autonomy, intelligence, work efficiency and operational benefits of underground cable operation and maintenance management.

[0153] See also Figure 4 , Figure 4 This is a structural block diagram of an underground cable damage identification system provided in Example 3 of the present invention.

[0154] A third embodiment of the present invention provides a system for identifying underground cable damage, comprising:

[0155] The target road surface recognition image generation module 401 is used to obtain the cable nameplate identification image and the road surface key image as the initial road surface recognition image, perform correction and recognition processing on the initial road surface recognition image, and generate the target road surface recognition image.

[0156] The cable code determination module 402 is configured to perform mapping calculation using the target road surface recognition image and the cable space data corresponding to the underground cables to determine the cable code corresponding to the underground cables.

[0157] The cable virtual space-time sequence model and construction external failure key point virtual space-time sequence model generation module 403 is used to use the cable static data set and cable dynamic data set corresponding to the cable coding to construct the model and generate the cable virtual space-time sequence model and construction external failure key point virtual space-time sequence model.

[0158] The construction external break point single frame label mapping value generation module 404 is used to identify construction external break points using the cable virtual space-time sequence model and the construction external break key point virtual space-time sequence model, and generate multiple construction external break point single frame label mapping values ​​corresponding to underground cables.

[0159] The external damage risk warning data determination module 405 is used to identify external damage risks based on the single-frame label mapping value of the construction external damage point and the preset standard violent construction library, and determine the external damage risk warning data corresponding to the underground cable.

[0160] Optionally, the target road surface recognition image generation module 401 includes:

[0161] The initial road surface recognition data generation module is used to obtain the cable nameplate identification image and the road surface key image according to different scenes as the initial road surface recognition image, perform road surface identification on the initial road surface recognition image, and generate initial road surface recognition data.

[0162] The first target road surface recognition image generation module is used to correct the road cable nameplate identification image according to a preset nameplate identification correction image calculation formula to generate a first target road surface recognition image if the initial road surface recognition data is a road cable nameplate identification image.

[0163] The calculation formula for the preset nameplate logo correction image is:

[0164] [x0 y0 1] T=A0[u d v d 1] T ;

[0165]

[0166]

[0167] Where A0 is the orientation element in the lens; k1 is the radial distortion coefficient; r (0,1) is the Euclidean distance between the distortion center and the distorted image point in the panoramic image; (x0, y0) is the physical coordinate corresponding to the image point of the cable nameplate under the influence of distortion; (u d ,v d ) is the pixel coordinate of the cable nameplate identification image point under the influence of distortion; (x (0,1) ,y (0,1) ) is the physical coordinate of the cable nameplate identification image point under the correction of no distortion; (x1, y1) is the map coordinate of the cable nameplate identification image point; C cm is the exterior orientation element of the panoramic image, where m ij C cm The element in the i-th row and j-th column of the matrix; s is the scaling factor.

[0168] The second target road surface recognition image generation module is used to correct the road surface key image according to a preset template image matching calculation formula to generate a second target road surface recognition image if the initial road surface recognition data is a road surface key image.

[0169] The preset template image matching calculation formula is:

[0170]

[0171]

[0172] N i =b i×8 ×2 0 +b i×8+1 ×2 1 +…+b i×8+7 ×2 7 ,N 16 =N0;

[0173]

[0174] H=h0×2 0 +…+h 15 ×2 15 ;

[0175] Where D i+1is the key image feature descriptor at position i+1; D i is the key image feature descriptor at position i; D 127 is the 127th key image feature descriptor; J0(x0,y0) is the road key image, (x0,y0) is the physical coordinate corresponding to the midpoint of the road key image; b i is the 128-bit binary code of the i-th bit; AD i is the i-th key point descriptor based on the scale-invariant feature transform matching algorithm; M is the set threshold used to calculate the binary code of the key point descriptor of the scale-invariant feature transform matching algorithm; N i The value calculated for the i-th sub-binary code group; h i is the binary code corresponding to the ith sub-binary code group; h 15 is the binary code corresponding to the 15th sub-binary code group; H is the hash address code of the key point descriptor.

[0176] Optionally, the target road surface identification image is a first target road surface recognition image, and the cable code determination module 402 includes:

[0177] The cable coding probability determination module is used to perform mapping calculation using the first target road surface recognition image and the cable space data corresponding to the underground cables according to a preset first mapping calculation formula to determine multiple cable coding probabilities corresponding to the underground cables.

[0178] The preset first mapping calculation formula is:

[0179] U0=[u i ,u j ],u i,j ∈I1;

[0180] U l =ω l T U l-1 +b l ;

[0181] Y=ω N T U N-1 +b N ;

[0182]

[0183] Where U0 is the neural network input layer, where the elements of U0 are the coordinate values ​​of the nameplate correction image map (u i ,u j ); I1 is the first target road surface recognition image; U l is the hidden layer value of the neural network layer l; ω l is the weight between the l-1th layer and the lth layer of the neural network; ωN is the weight between the N-1th layer and the Nth layer of the neural network; T represents the matrix; b l is the bias value of the neural network layer l; b N is the bias value of the Nth layer of the neural network; Y is the output value of the Nth layer of the neural network; N is the total number of neural network layers; Y(i) is the classification output probability value calculated by the normalized exponential function classifier of the i-th neural network output value; e is a natural constant; e Y(i) is e to the power of Y(i).

[0184] The first submodule for determining the cable code is configured to use the code corresponding to the maximum value of the cable code probabilities as the cable code corresponding to the underground cable.

[0185] Optionally, the target road surface identification image is a second target road surface recognition image, and the cable code determination module 402 includes:

[0186] The similarity determination module is used to use the second target road surface recognition image and the cable space data corresponding to the underground cable to perform matching calculation on the key identifiers of the surrounding environment according to the preset second mapping calculation formula to determine multiple similarities corresponding to the underground cable.

[0187] The preset second mapping formula is:

[0188]

[0189] Where H is the hash address code of the key point descriptor of the second target road surface recognition image; is the hash address code of the key sign descriptor of the surrounding environment of the i-th cable in the database; L M is the total number of cables in the database; E is the similarity.

[0190] The second submodule for determining the cable code is configured to use the code corresponding to the maximum value of the similarity as the cable code corresponding to the underground cable.

[0191] Optionally, the cable static data set includes a cable technical parameter group, a cable change record information group, and an external force damage alarm information group. The cable virtual time-space sequence model and the construction external damage key point virtual time-space sequence model generation module 403 include:

[0192] The cable insulation sheath position generation module is used to determine the initial position of the cable insulation sheath using the cable technical parameter group and generate the cable insulation sheath position.

[0193] The cable insulation sheath movement trajectory data generation module is used to use the cable change record information group and the external force damage alarm information group, combined with the cable insulation sheath position, to construct the cable insulation sheath movement trajectory and generate the cable insulation sheath movement trajectory data.

[0194] The cable virtual space-time sequence model generation module is used to construct a model based on the cable insulation sheath movement trajectory data to generate a cable virtual space-time sequence model.

[0195] The construction machinery motion trajectory data generation module is used to simulate and predict the motion trajectory of construction machinery using the cable dynamic data set to generate the construction machinery motion trajectory data.

[0196] The change distance generation module is used to calculate the distance change between the construction equipment motion trajectory data and the corresponding cable insulation sheath movement trajectory data to generate multiple change distances.

[0197] The construction external destruction key point generation module is used to select key points corresponding to the change distance less than a preset threshold and generate construction external destruction key points.

[0198] The virtual space-time sequence model generation module of the key points of external construction demolition is used to construct a model using the position data corresponding to the key points of external construction demolition to generate a virtual space-time sequence model of the key points of external construction demolition.

[0199] Optionally, the cable virtual space-time sequence model and the construction external damage key point virtual space-time sequence model include multiple time and space domains. The construction external damage point single frame label mapping value generation module 404 includes:

[0200] The vertical Euclidean distance set generation module is used to calculate the vertical Euclidean distance between each construction external damage key point and the cable insulation sheath corresponding to the cable virtual space-time sequence model in each time and space field, and generate the vertical Euclidean distance set corresponding to the time and space field.

[0201] The submodule for generating single-frame label mapping values ​​of construction external break points is used to use the vertical Euclidean distance set as the single-frame label mapping value of the construction external break point corresponding to the underground cable.

[0202] Optionally, the external damage risk warning data determination module 405 includes:

[0203] The violent construction action set generation module is used to classify multiple feature vectors corresponding to the single-frame label mapping values ​​of the construction external breaking points through a preset classifier to generate a violent construction action set.

[0204] The external force damage risk value generation module is used to identify external damage risks by comparing the violent construction action set with the preset standard violent construction library and generate an external force damage risk value.

[0205] The external damage risk warning position determination module is used to use the position corresponding to the external damage risk value as the external damage risk warning position if the external damage risk value is greater than a preset risk threshold.

[0206] The external damage risk warning data determination submodule is used to construct external damage risk warning data corresponding to underground cables by using external damage risk warnings and external damage risk warning positions corresponding to preset risk thresholds.

[0207] An embodiment of the present invention further provides an electronic device, comprising: a memory and a processor, wherein a computer program is stored in the memory; when the computer program is executed by the processor, the processor executes the underground cable damage identification method as described in any of the above embodiments.

[0208] The memory can be an electronic memory such as a flash memory, an EEPROM (Electrically Erasable Programmable Read-Only Memory), an EPROM, a hard disk or a ROM. The memory has storage space for program codes for executing any method step in the above method. For example, the storage space for program codes can include individual program codes for implementing the various steps in the above method respectively. These program codes can be read from or written to one or more computer program products. These computer program products include program code carriers such as a hard disk, a compact disk (CD), a memory card or a floppy disk. The program code can be compressed, for example, in an appropriate form. When these codes are run by a computing and processing device, the computing and processing device executes the individual steps in the underground cable external damage identification method described above.

[0209] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for identifying underground cable damage as described in any of the above embodiments is implemented.

[0210] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0211] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0212] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0213] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0214] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments 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 code.

[0215] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. 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 various embodiments of the present invention.

Claims

1. A method for identifying underground cable damage, characterized in that: include: Acquire a cable nameplate identification image and a road surface key image as an initial road surface recognition image, perform correction and recognition processing on the initial road surface recognition image, and generate a target road surface recognition image; Performing a mapping calculation using the target road surface recognition image and cable space data corresponding to the underground cable to determine the cable code corresponding to the underground cable; The cable static data set and the cable dynamic data set corresponding to the cable coding are used to construct a model to generate a cable virtual space-time sequence model and a construction external failure key point virtual space-time sequence model; The cable virtual space-time sequence model and the construction external break key point virtual space-time sequence model are used to identify the construction external break points, and generate a plurality of single-frame label mapping values ​​of the construction external break points corresponding to the underground cable; Identify the external damage risk based on the single-frame label mapping value of the construction external damage point and the preset standard violent construction library, and determine the external damage risk warning data corresponding to the underground cable; The cable static data set includes a cable technical parameter group, a cable change record information group, and an external force damage alarm information group; the step of using the cable static data set corresponding to the cable code and the cable dynamic data set to construct a model to generate a cable virtual space-time sequence model and a construction external damage key point virtual space-time sequence model includes: The cable technical parameter group is used to determine the initial position of the cable insulation sheath, and generate the cable insulation sheath position; The cable change record information group and the external force damage alarm information group are used in combination with the cable insulation sheath position to construct a cable insulation sheath movement trajectory to generate cable insulation sheath movement trajectory data; Modeling the cable insulation sheath movement trajectory data to generate a cable virtual space-time sequence model; Using the cable dynamic data set to simulate and predict the motion trajectory of the construction equipment to generate construction equipment motion trajectory data; Calculate the distance change between the movement trajectory data of the construction equipment and the corresponding movement trajectory data of the cable insulation sheath to generate multiple change distances; Selecting key points corresponding to the change distances that are smaller than a preset threshold value to generate construction external breaking key points; The position data corresponding to the key points of external destruction during construction are used to construct a model, and a virtual space-time sequence model of the key points of external destruction during construction is generated.

2. The underground cable damage identification method according to claim 1, characterized in that: The step of acquiring the cable nameplate identification image and the road surface key image as an initial road surface recognition image, performing correction and recognition processing on the initial road surface recognition image, and generating a target road surface recognition image includes: Acquire a cable nameplate identification image and a road surface key image according to different scenes as an initial road surface recognition image, perform road surface identification on the initial road surface recognition image, and generate initial road surface recognition data; If the initial road surface recognition data is a road cable nameplate identification image, correcting the road cable nameplate identification image according to a preset nameplate identification correction image calculation formula to generate a first target road surface recognition image; The calculation formula for the preset nameplate logo correction image is: ； ； ； Where, It is the directional element existing in the lens; is the radial distortion coefficient; is the Euclidean distance between the distortion center and the distorted image point in the panoramic image; The physical coordinates of the cable nameplate identification image points under the influence of distortion; is the pixel coordinate of the cable nameplate identification image point under the influence of distortion; To correct the physical coordinates of the cable nameplate identification image points without distortion; The map coordinates corresponding to the cable nameplate identification image points; is the exterior orientation element of the panoramic image, where for The element in the i-th row and j-th column of the matrix; s is the scale factor; If the initial road surface recognition data is a road surface key image, the road surface key image is corrected according to a preset template image matching calculation formula to generate a second target road surface recognition image; The preset template image matching calculation formula is: ； Where, is the key image feature descriptor of the i+1th position; is the key image feature descriptor of the i-th position; It is the 127th key image feature descriptor; is the key image of the road surface, is the physical coordinate corresponding to the midpoint of the key image of the road surface; is the 128-bit i-th binary code; is the i-th key point descriptor based on the scale-invariant feature transform matching algorithm; M is the set threshold used to calculate the binary code of the key point descriptor of the scale-invariant feature transform matching algorithm; The value calculated for the i-th sub-binary code group; is the binary code corresponding to the i-th sub-binary code group; is the binary code corresponding to the 15th sub-binary code group; H is the hash address code of the key point descriptor.

3. The underground cable damage identification method according to claim 2, characterized in that: The target road surface recognition image is a first target road surface recognition image; The step of performing mapping calculation using the target road surface recognition image and the cable space data corresponding to the underground cable to determine the cable code corresponding to the underground cable includes: Performing a mapping calculation using the first target road surface recognition image and the cable space data corresponding to the underground cable according to a preset first mapping calculation formula to determine a plurality of cable coding probabilities corresponding to the underground cable; The preset first mapping calculation formula is: ； Where, is the neural network input layer, where The elements are the coordinate values ​​of the nameplate correction image map ; A first target road surface recognition image; is the hidden layer value of the neural network layer l; is the weight between the l-1th layer and the lth layer of the neural network; is the weight between the N-1th layer and the Nth layer of the neural network; T represents the matrix; is the bias value of the lth layer of the neural network; is the bias value of the Nth layer of the neural network; Y is the output value of the Nth layer of the neural network; N is the total number of neural network layers; Y(i) is the classification output probability value calculated by the normalized exponential function classifier of the i-th neural network output value; e is a natural constant; is e to the power of Y(i); The code corresponding to the maximum value of the cable code probabilities is used as the cable code corresponding to the underground cable.

4. The underground cable damage identification method according to claim 2, characterized in that: The target road surface recognition image is a second target road surface recognition image; The step of performing mapping calculation using the target road surface recognition image and the cable space data corresponding to the underground cable to determine the cable code corresponding to the underground cable includes: Using the second target road surface recognition image and the cable space data corresponding to the underground cable to perform a matching calculation on the key identifiers of the surrounding environment according to a preset second mapping calculation formula, and determining multiple similarities corresponding to the underground cable; The preset second mapping calculation formula is: ； Where, A hash address code for a key point descriptor of a second target road surface recognition image; is the hash address code of the key descriptor of the surrounding environment of the i-th cable in the database; is the total number of cables in the database; E is the similarity; The code corresponding to the maximum value of the similarities is used as the cable code corresponding to the underground cable.

5. The underground cable damage identification method according to claim 1, characterized in that: The cable virtual space-time sequence model and the construction external failure key point virtual space-time sequence model include multiple time and space fields; the step of using the cable virtual space-time sequence model and the construction external failure key point virtual space-time sequence model to identify construction external failure points and generate multiple single-frame label mapping values ​​of construction external failure points corresponding to the underground cable includes: Calculating the vertical Euclidean distance between each of the key construction external damage points and the cable insulation sheath corresponding to the cable virtual space-time sequence model in each of the time-space domains, and generating a vertical Euclidean distance set corresponding to the time-space domain; The vertical Euclidean distance set is used as a single-frame label mapping value of the construction external break point corresponding to the underground cable.

6. The underground cable damage identification method according to claim 1, characterized in that: The step of identifying external damage risk based on the single-frame label mapping value of the construction external damage point and a preset standard violent construction library to determine the external damage risk warning data corresponding to the underground cable includes: Classifying multiple feature vectors corresponding to the single-frame label mapping values ​​of the construction external breaking points through a preset classifier to generate a set of violent construction actions; The set of violent construction actions is compared with a preset standard violent construction library to identify external damage risks and generate an external force damage risk value; If the external force damage risk value is greater than a preset risk threshold, the position corresponding to the external force damage risk value is used as an external damage risk warning position; The external damage risk warning corresponding to the preset risk threshold and the external damage risk warning position are used to construct external damage risk warning data corresponding to the underground cable.

7. An underground cable damage identification system, characterized in that: The underground cable damage identification system is used to implement the underground cable damage identification method according to any one of claims 1 to 6, comprising: a target road surface recognition image generation module, configured to obtain a cable nameplate identification image and a road surface key image as an initial road surface recognition image, perform correction and recognition processing on the initial road surface recognition image, and generate a target road surface recognition image; a cable code determination module, configured to perform mapping calculation using the target road surface recognition image and cable space data corresponding to the underground cable to determine the cable code corresponding to the underground cable; A module for generating a cable virtual space-time sequence model and a construction external failure key point virtual space-time sequence model is used to construct a model using the cable static data set and the cable dynamic data set corresponding to the cable coding to generate a cable virtual space-time sequence model and a construction external failure key point virtual space-time sequence model; A single-frame label mapping value generation module for construction external break points is used to identify construction external break points using the cable virtual space-time sequence model and the construction external break key point virtual space-time sequence model, and generate a plurality of single-frame label mapping values ​​of construction external break points corresponding to the underground cable; An external damage risk warning data determination module is used to identify external damage risks based on the single-frame label mapping value of the construction external damage point and a preset standard violent construction library, and determine the external damage risk warning data corresponding to the underground cable; The cable static data set includes a cable technical parameter group, a cable change record information group, and an external force damage alarm information group; the cable virtual space-time sequence model and the construction external damage key point virtual space-time sequence model generation module include: a cable insulation sheath position generating module, configured to determine an initial position of the cable insulation sheath using the cable technical parameter group and generate a cable insulation sheath position; a cable insulation sheath movement trajectory data generating module, configured to construct a cable insulation sheath movement trajectory using the cable change record information group and the external force damage alarm information group in combination with the cable insulation sheath position, and generate cable insulation sheath movement trajectory data; A cable virtual space-time sequence model generation module is used to construct a model based on the cable insulation sheath movement trajectory data to generate a cable virtual space-time sequence model; A construction equipment motion trajectory data generation module is used to simulate and predict the motion trajectory of the construction equipment using the cable dynamic data set to generate the construction equipment motion trajectory data; a change distance generating module, configured to calculate the distance change between the motion trajectory data of the construction equipment and the corresponding movement trajectory data of the cable insulation sheath, and generate a plurality of change distances; A construction external failure key point generation module is used to select key points corresponding to the change distance less than a preset threshold value to generate construction external failure key points; The virtual space-time sequence model generation module of the construction external demolition key points is used to use the position data corresponding to the construction external demolition key points to construct a model and generate a virtual space-time sequence model of the construction external demolition key points.

8. An electronic device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor executes the steps of the underground cable damage identification method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the underground cable damage identification method according to any one of claims 1 to 6 is implemented.

Images