Icing thickness detection method and device, electronic equipment and storage medium

Abstract

The application relates to the field of power transmission line detection, in particular to an icing thickness detection method and device, an electronic device and a storage medium, the icing thickness detection method comprising the following steps: acquiring an icing image of a power transmission line; inputting the icing image into a pre-trained icing semantic extraction model to obtain icing semantic information of the power transmission line; determining an icing contour of the power transmission line based on the icing semantic information; and determining the icing thickness of the power transmission line based on the icing contour. According to the embodiment of the application, the icing thickness can be detected without power-off maintenance, the hardware devices required for icing thickness detection can be simplified, and the maintenance cost and difficulty of the hardware devices are reduced.

Description

Technical Field

[0001] This application relates to the field of power transmission line testing, specifically to a method, apparatus, electronic device, and storage medium for detecting icing thickness. Background Technology

[0002] Line icing refers to a situation where all power lines within a certain area are encased in ice. Line icing can cause problems such as conductor galloping, tower tilting and collapse, wire breaks, and insulator flashover, causing great inconvenience to production and daily life, as well as resulting in huge economic losses.

[0003] Monitoring the icing status of transmission lines is beneficial for addressing icing-affected lines before icing disasters occur, such as activating de-icing equipment to effectively prevent icing from spreading. When detecting ice thickness, tension sensors can be installed on the conductors and ground wires of the transmission line to measure insulator tension, insulator tilt angle, conductor stress, and conductor galloping, among other values. Then, an algorithm is used to determine the current ice thickness of the transmission line based on the relationship between these measurements and the ice thickness.

[0004] However, the sensors using the above measurement methods are complex to install, and may even require a power outage during installation, affecting the normal use of power transmission lines. Moreover, the maintenance costs and difficulties are also high. Summary of the Invention

[0005] In view of the above, embodiments of this application provide an ice thickness detection method, electronic device, apparatus, and storage medium, which aim to reduce the installation complexity, maintenance cost, and maintenance difficulty of hardware devices while accurately detecting ice thickness.

[0006] This application provides a method for detecting ice thickness, including:

[0007] Acquire images of icing on power transmission lines;

[0008] The icing image is input into a pre-trained icing semantic extraction model to obtain the icing semantic information of the transmission line;

[0009] Based on the icing semantic information, the icing profile of the transmission line is determined;

[0010] Based on the icing profile, the icing thickness of the transmission line is determined.

[0011] This application embodiment extracts the semantic information of icing on the transmission line based on the icing image to obtain the icing contour. The icing thickness can be obtained from the icing contour. There is no need to install sensors for measuring values ​​such as the tensile strength of the transmission line. Not only can the icing thickness be detected without power outage maintenance, but it can also simplify the hardware equipment required for icing thickness detection, thereby reducing the installation complexity, maintenance cost and maintenance difficulty of the hardware equipment.

[0012] In some embodiments, the icing semantic extraction model includes a meteorological analysis model and a semantic segmentation model;

[0013] The step of inputting the icing image into a pre-trained icing semantic extraction model to obtain the icing semantic information includes:

[0014] Acquire micrometeorological data that matches the time of the icing image capture;

[0015] The micrometeorological data is input into the meteorological analysis model to obtain the weather information at the time the icing image was taken;

[0016] Detect whether the weather information at the time the icing image was taken meets the meteorological conditions for icing;

[0017] If the weather information at the time the icing image was taken meets the icing meteorological conditions, the icing image is input into the semantic segmentation model to obtain the icing semantic information.

[0018] In some embodiments, the training steps of the icing semantic extraction model include:

[0019] Acquire icing image samples and micrometeorological data samples;

[0020] Based on the icy image samples and the current semantic segmentation model, the first loss function value corresponding to the current semantic segmentation model is obtained;

[0021] Based on the micrometeorological data sample and the current meteorological analysis model, the second loss function value corresponding to the current meteorological analysis model is obtained;

[0022] Based on the first loss function value and the second loss function value, the total loss function value of the icing semantic extraction model is obtained;

[0023] If the total loss function value indicates that the icing semantic extraction model has not converged, the model parameters in the current semantic segmentation model and the current meteorological analysis model are adjusted to obtain the adjusted icing semantic extraction model.

[0024] Continue training the adjusted icing semantic extraction model until it converges.

[0025] In some embodiments, the icing semantic extraction model includes: a first feature pyramid and a second feature pyramid;

[0026] The first feature pyramid is used to downsample the icing image several times to obtain a first matrix;

[0027] The second feature pyramid is used to upsample the first matrix several times to obtain the icing semantic information.

[0028] In some embodiments, the icing profile includes a first profile line and a second profile line along the transmission line, and determining the icing thickness of the transmission line based on the icing profile includes:

[0029] Find the first position on the first contour line and the second position on the second contour line; the distance between the first position and the second position is the shortest distance between the first contour line and the second contour line;

[0030] Based on the first location and the second location, the icing diameter of the transmission line is determined;

[0031] The ice thickness is determined based on the ice diameter and the diameter of the transmission line.

[0032] In some embodiments, determining the icing diameter of the transmission line based on the first location and the second location includes:

[0033] Obtain a first set of neighboring points, the first set of neighboring points including the first position and a plurality of positions adjacent to the first position on the first contour line;

[0034] Calculate the shortest distance between each position in the first set of nearest points and the second contour line to obtain multiple distance values;

[0035] The icing diameter is determined based on the multiple distance values.

[0036] In some embodiments, determining the icing profile of the transmission line based on the icing semantic information includes:

[0037] The icing semantic information is binarized to obtain a binarized image;

[0038] The icing contour of the transmission line is extracted from the binarized image.

[0039] This application also provides an ice thickness detection device, including:

[0040] The acquisition module is used to acquire images of icing on transmission lines;

[0041] The extraction module is used to input the icing image into a pre-trained icing semantic extraction model to obtain the icing semantic information of the transmission line;

[0042] The calculation module is used to determine the icing profile of the transmission line based on the icing semantic information; and to determine the icing thickness of the transmission line based on the icing profile.

[0043] This application also provides an electronic device, which includes a processor and a memory. The memory is used to store instructions, and the processor is used to call the instructions in the memory to cause the electronic device to execute the above-described ice thickness detection method.

[0044] This application also provides a computer-readable storage medium that stores computer instructions that, when executed on an electronic device, cause the electronic device to perform the above-described ice thickness detection method. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of a scenario for an ice thickness detection system provided in an embodiment of this application.

[0046] Figure 2 This is a flowchart illustrating the steps of an ice thickness detection method provided in an embodiment of this application.

[0047] Figure 3 This is a schematic diagram of the network structure of a semantic segmentation model provided in an embodiment of this application.

[0048] Figure 4 This is a schematic diagram of the network structure of a meteorological analysis model provided in an embodiment of this application.

[0049] Figure 5 A flowchart illustrating the steps of a training method for an icing semantic extraction model provided in an embodiment of this application.

[0050] Figure 6 This is a schematic diagram of the outline of an icing profile provided in one embodiment of this application.

[0051] Figure 7 A flowchart of the sub-steps of step 204 provided in an embodiment of this application.

[0052] Figure 8 This is a schematic diagram of the structure of an ice thickness detection device provided in an embodiment of this application.

[0053] Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0054] To better understand the above-mentioned objectives, features, and advantages of this application, the application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0055] The following description sets forth many specific details to provide a full understanding of this application. The described embodiments are only some, not all, of the embodiments of this application.

[0056] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the specification of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application.

[0057] It should be further noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0058] In this application, "at least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and drawings of this application are used to distinguish similar objects, not to describe a specific order or sequence.

[0059] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0060] This application provides a method, apparatus, electronic device 101, and computer-readable storage medium for detecting ice thickness, which will be described in detail below.

[0061] Please see Figure 1As shown, Figure 1 This is a schematic diagram of a scenario for an ice thickness detection system provided in an embodiment of this application.

[0062] The ice thickness detection system may include an electronic device 101 and a camera device 102, wherein the electronic device 101 integrates the ice thickness detection device. The electronic device 101 is communicatively connected to the camera device 102.

[0063] The camera device 102 is used to capture images of the icing on the power transmission line and to transmit these images to the electronic device 101. The camera device 102 can be a camera or other device with video recording and photography capabilities.

[0064] The electronic device 101 can be used to acquire icing images of transmission lines; input the icing images into a pre-trained icing semantic extraction model to obtain icing semantic information of the transmission lines; determine the icing contour of the transmission lines based on the icing semantic information; and determine the icing thickness of the transmission lines based on the icing contour.

[0065] Electronic device 101 is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions. Its hardware includes, but is not limited to, processors, microprogrammed controllers (MCUs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc. Electronic device 101 can be a personal computer, server, etc.

[0066] Those skilled in the art will understand that Figure 1 The application environment shown is merely one application scenario of the solution in this application and does not constitute a limitation on the application scenario of the solution in this application. Other application environments may include those that are more specific to this application. Figure 1 The more or fewer devices / equipment shown, for example Figure 1 Only one electronic device 101 and one camera device 102 are shown in the diagram. It is understood that the ice thickness detection system may also include one or more other devices, such as the camera device 102 or other devices with other functions, which are not limited here.

[0067] In addition, such as Figure 1As shown, the icing thickness detection system may further include: a micro-meteorological monitoring device 103, which is communicatively connected to an electronic device 101. The micro-meteorological monitoring device 103 can be used to collect micro-meteorological data of the environment where the transmission line is located, and transmit the collected micro-meteorological data to the electronic device 101, so that the electronic device 101 can also combine the micro-meteorological data to obtain the icing thickness.

[0068] Micrometeorology refers to changes in small-scale meteorological elements in the near-surface atmosphere and upper soil layer caused by certain structural features (such as micro-topography). Micrometeorological data may include, but are not limited to, temperature, humidity, wind speed, wind direction, atmospheric pressure, rainfall, and sunlight.

[0069] The micro-weather monitoring device 103 may include a micro-weather sensor.

[0070] The aforementioned camera device 102, micro-meteorological monitoring device 103, etc., can be powered by solar energy to achieve long-term ice thickness detection without human intervention.

[0071] Please refer to Figure 2 As shown, Figure 2 This is a flowchart illustrating the steps of an ice thickness detection method according to an embodiment of this application. This ice thickness detection method can be applied to electronic device 101.

[0072] It is understandable that the order of steps in this flowchart can be changed, and some steps can be omitted, depending on different needs.

[0073] See Figure 2 As shown, the ice thickness detection method may include the following steps.

[0074] Step 201: Obtain an image of the icing on the transmission line.

[0075] For example, electronic device 101 can acquire images of icing on power transmission lines from camera device 102.

[0076] In some embodiments, after obtaining the icing image, preprocessing operations such as image enhancement and spatial filtering can be performed on the icing image, and then step 202 is executed so that the subsequent icing semantic extraction model can process the icing image.

[0077] Step 202: Input the icing image into the pre-trained icing semantic extraction model to obtain the icing semantic information of the transmission line.

[0078] The semantic information of icing refers to the areas in the icing image where there is icing on the transmission lines.

[0079] In some embodiments, the icing semantic extraction model includes a semantic segmentation model.

[0080] For example, the semantic segmentation model can be U-Net, FCN, SegNet, etc., but is not limited to these.

[0081] For example, please see Figure 3 As shown, Figure 3 This is a schematic diagram of the network structure of a semantic segmentation model provided in an embodiment of this application.

[0082] The semantic segmentation model may include a first feature pyramid 301 and a second feature pyramid 302. The first feature pyramid 301 is used to downsample the icing image several times to obtain a first matrix; the second feature pyramid 302 is used to upsample the first matrix several times to obtain icing semantic extraction information.

[0083] The first feature pyramid 301 includes several groups of convolutional layers and max pooling layers, each group of convolutional layers and max pooling layers is used for downsampling once. The second feature pyramid 302 also includes several groups of convolutional layers and max pooling layers, each group of convolutional layers and max pooling layers is used for upsampling once.

[0084] The scaling factor for upsampling in the second feature pyramid 302 is the antisymmetric coefficient of the scaling factor for downsampling in the first feature pyramid. That is, the scaling factor for the first upsampling in the second feature pyramid 302 is the reciprocal of the scaling factor for the last downsampling in the first feature pyramid 301. For example, if the scaling factor for the last downsampling in the first feature pyramid 301 is 1:2, then the scaling factor for the first upsampling in the second feature pyramid 302 is 2:1; the scaling factor for the second upsampling in the second feature pyramid 302 is the reciprocal of the scaling factor for the second downsampling in the first feature pyramid 301, and so on.

[0085] For example, refer to Figure 3 As shown, let the length and width of the icing image be denoted as a×b. Assuming that the scaling ratios of the first feature pyramid 301 for downsampling the length and width of the icing image are {1:4, 1:2, 1:2, 1:2}, then... Figure 3In the equation, n1 equals 4, n2 equals 8, n3 equals 16, and n4 equals 32. That is, the first feature pyramid 301 downsamples the icing image several times to obtain a matrix with sizes of {(a / 4)×(b / 4), (a / 8)×(b / 8), (a / 16)×(b / 16), (a / 32)×(b / 32)}, where the size of the first matrix is ​​(a / 32)×(b / 32). The second feature pyramid 302 upsamples the first matrix several times, and the scaling ratio of the second feature pyramid 302 upsampling several times is {2:1, 2:1, 2:1}. The size of the matrix obtained by the second feature pyramid 302 upsampling several times is {(a / 16)×(b / 16), (a / 8)×(b / 8), (a / 4)×(b / 4).

[0086] In other embodiments, the icing semantic extraction model includes a meteorological analysis model and a semantic segmentation model.

[0087] Step 202 may include: acquiring micro-meteorological data that matches the shooting time of the icing image, such as micro-meteorological data measured when the icing image was taken, or micro-meteorological data measured within a preset time period when the icing image was taken; then, inputting the micro-meteorological data into a meteorological analysis model to obtain weather information; and obtaining icing semantic information based on the weather information, the icing image, and the semantic segmentation model.

[0088] Weather information can be used to describe the temperature and humidity of the climate environment where power transmission lines are located. For example, weather information can include sunny days, rainy days, snowy days, etc.

[0089] Electronic devices can verify whether a power transmission line is under icing conditions (i.e., icing meteorological conditions) based on this weather information, thereby verifying the accuracy of the extracted icing semantic information. The icing meteorological conditions can be set according to the actual environment, such as the altitude, air pressure, and region where the power transmission line is located, as well as other practical needs; this application embodiment does not limit this setting.

[0090] If the weather information meets the icing meteorological conditions, such as the weather information indicating that the temperature of the transmission line is below zero and the air humidity is high, and it is a snowy day, it means that the transmission line is likely to be iced. This shows that the correct icing semantic information can be extracted based on the icing image and semantic segmentation model. Then, the icing thickness of the transmission line can be determined based on the icing semantic information, i.e., step 203 is executed.

[0091] If the weather information does not meet the meteorological conditions for icing, such as the weather information indicating that the current temperature is in the thirties and it is a sunny day with plenty of sunshine, it means that the transmission line is unlikely to have icing. This suggests that the icing semantic information extracted based on the icing image and semantic segmentation model may be incorrect, for example, due to incorrect extraction caused by reflection.

[0092] In some embodiments, obtaining icing semantic information based on weather information, icing images, and a semantic segmentation model may include: detecting whether the weather information at the time the icing image was taken meets the icing meteorological conditions. If the weather information at the time the icing image was taken meets the icing meteorological conditions, the icing image is input into the semantic segmentation model to obtain the icing semantic information.

[0093] If the weather information at the time the icing image was taken does not meet the icing meteorological conditions, the icing thickness need not be calculated.

[0094] This application embodiment utilizes micro-meteorological data to determine whether icing may occur on transmission lines, thereby determining whether accurate icing semantic information can be extracted from icing images, improving the reliability of icing semantic information, and thus improving the accuracy of icing thickness calculation.

[0095] The meteorological analysis model described above is used to obtain weather information based on micrometeorological data. The meteorological analysis model can be a CNN model.

[0096] For example, refer to Figure 4 As shown, Figure 4 This is a schematic diagram of the structure of a meteorological analysis model provided in an embodiment of this application.

[0097] Figure 4 The seven meteorological elements shown include micrometeorological data such as temperature, humidity, wind speed, wind direction, atmospheric pressure, rainfall, and sunshine. Electronic device 101 can input these seven meteorological elements. Figure 3 The weather analysis model shown yields weather information.

[0098] In practical applications, the type of micro-meteorological data input into the meteorological analysis model can be set according to the actual application requirements, and this application embodiment does not limit this.

[0099] In some embodiments, reference Figure 5 As shown, the training steps for the icing semantic extraction model include:

[0100] Step 501: Obtain icing image samples and micrometeorological data samples.

[0101] The icing image samples are marked with icing contours, and the icing contours form closed loops. The pixel labels inside the icing contour and the pixel labels of the icing contour are positive samples, while the pixel labels of the icing contour outside the icing contour are negative samples.

[0102] Micro-meteorological data samples include micro-meteorological data groups, such as the seven meteorological elements and their corresponding labels, and weather information, such as sunny, rainy, or snowy days.

[0103] Step 502: Based on the icy image samples and the current semantic segmentation model, obtain the first loss function value corresponding to the current semantic segmentation model.

[0104] For example, when the pixel label of a pixel in an icy image sample is a positive sample, the corresponding loss function is denoted as f. 11 f 11 as follows:

[0105] f 11 =-(1-p) γ logp;

[0106] Where p is the predicted probability that the pixel belongs to a positive or negative sample; γ is a hyperparameter in the semantic segmentation model.

[0107] When the pixel label is a negative sample, the corresponding loss function is denoted as f. 12 f 12 as follows:

[0108] f 12 =-p γ logp;

[0109] Where p is the predicted probability that the pixel belongs to a positive or negative sample; γ is a hyperparameter in the semantic segmentation model.

[0110] The loss function of the semantic segmentation model is denoted as F1, and F1 is shown below:

[0111] F1=∑αw i f 11 +(1-α)f 12 ;

[0112] Among them, w i w is the weight parameter. i =(NN) i ) / N, where N is the total number of pixels in the icing image, N i Let α be the number of pixels in the icy image sample labeled with category i, where category i refers to either a positive or negative sample. When the pixel is a positive sample, α = 1; when the pixel is a negative sample, α = 0.

[0113] Step 503: Based on the micro-meteorological data sample and the current meteorological analysis model, obtain the second loss function value corresponding to the current meteorological analysis model.

[0114] For example, meteorological analysis models can be referenced. Figure 4 As shown, the meteorological analysis model can be a CNN network composed of multiplicative layers, weighted layers, normalization layers, rectified linear units (ReLU), fully connected layers, and dropout layers.

[0115] The second loss function of this meteorological analysis model can be denoted as F2, as shown below:

[0116] F2 = 1 - exp(-(xy)) 2 / x 2 );

[0117] Where x is the label in the micro-meteorological data sample, y is the output data of the meteorological analysis model, i.e. the predicted weather information, and exp is the exponential function.

[0118] Step 504: Based on the first loss function value and the second loss function value, obtain the total loss function value of the icing semantic extraction model.

[0119] The total loss function can be denoted as F, where F = θF1 + βF2; θ is the weight of the first loss function F1, and β is the weight of the second loss function F2. The values ​​of θ and β can be set according to actual application requirements, and this application embodiment does not limit them.

[0120] Step 505: If the icing semantic extraction model has not converged based on the total loss function value, adjust the model parameters in the current semantic segmentation model and the current meteorological analysis model to obtain the adjusted icing semantic extraction model.

[0121] Step 506: Continue training the adjusted icing semantic extraction model until it converges.

[0122] That is, steps 501 to 505 can be repeated until the icing semantic extraction model converges.

[0123] The embodiments of this application combine meteorological analysis models and semantic segmentation models for training, which can comprehensively consider the features of micro-meteorological data and icing images, thereby improving the accuracy of icing semantic information extraction.

[0124] Once the icing semantic information is obtained, step 203 can be executed.

[0125] Step 203: Determine the icing profile of the transmission line based on icing semantic information.

[0126] In some embodiments, step 203 may include: binarizing the icing semantic information to obtain a binarized image; and extracting the icing contour of the transmission line from the binarized image.

[0127] The ice-covered outline is also texture information in the semantic information of ice-covering.

[0128] For example, the icing contours of transmission lines can be obtained using image binarization algorithms. Image binarization algorithms can include, but are not limited to, the averaging method, the bimodal method, and the Otsu algorithm.

[0129] Step 204: Determine the icing thickness of the transmission line based on the icing profile.

[0130] refer to Figure 6 As shown, the icing profile includes a first profile line 601 and a second profile line 602 along the transmission line 600.

[0131] In some embodiments, reference Figure 7 As shown, step 204 may include:

[0132] Step 701: Determine the icing diameter of the transmission line based on the first contour line and the second contour line.

[0133] In some embodiments, step 701 may include: randomly selecting a position on the first contour line 601, calculating the distance between the position and the second contour line 602, and using the distance as the icing diameter of the transmission line 600.

[0134] In some other embodiments, step 701 may include: taking the shortest distance between the first profile line 601 and the second profile line 602 as the icing diameter of the transmission line 600.

[0135] In some other embodiments, step 701 may include: taking the longest distance between the first contour line 601 and the second contour line 602 as the icing diameter of the transmission line 600.

[0136] In other embodiments, step 701 may include:

[0137] Step 7011: Locate the first position on the first contour line and the second position on the second contour line.

[0138] The distance between the first position and the second position is the shortest distance between the first contour line 601 and the second contour line 602.

[0139] For example, the first and second positions can be found using search algorithms such as depth-first search or breadth-first search.

[0140] Then, the icing diameter of the transmission line 600 can be determined based on the first and second positions. For example, steps 7012 to 7014 below can be performed.

[0141] Step 7012: Obtain the first set of neighboring points.

[0142] The first set of neighboring points includes a first location and a plurality of locations adjacent to the first location on the first contour line 601.

[0143] In some embodiments, several contour points near the first position can be uniformly selected on the first contour line 601 as multiple positions adjacent to the first position. For example, positions spaced 5, 10, or 15 pixels from the first position can be selected as multiple positions adjacent to the first position.

[0144] Step 7013: Calculate the shortest distance between each position in the first set of neighboring points and the second contour line to obtain multiple distance values.

[0145] For example, refer again Figure 6 As shown, assuming the first position is denoted as a0, and the first neighboring point set X = {a0, a1, ..., an}, the position corresponding to the shortest distance from each position in set X to the second contour line 602 can be found on the second contour line 602. These positions are denoted as set Y, set Y = {b0, b1, ..., bn}, where multiple distance values ​​are the distance between a0 and b0, the distance between a1 and b1, ..., and the distance between an and bn, respectively.

[0146] Step 7014: Determine the icing diameter based on multiple distance values.

[0147] In some embodiments, the mean of multiple distance values ​​can be calculated and used as the ice diameter; or, the highest and lowest values ​​can be removed from the multiple distance values, and the average value can be calculated and used as the ice diameter.

[0148] In other embodiments, the mode or median distance value can be selected from multiple distance values ​​as the ice diameter.

[0149] The above method for calculating the icing diameter based on multiple distance values ​​is only an example. In actual applications, it can be set according to requirements, and the embodiments of this application do not limit it.

[0150] Step 702: Determine the icing thickness based on the icing diameter and the transmission line diameter.

[0151] For example, let Z be the diameter of the icing, d be the diameter of the transmission line 600, and (Zd) / 2 be the icing thickness.

[0152] This application embodiment extracts the semantic information of icing on the transmission line 600 based on the icing image to obtain the icing profile. The icing thickness can be obtained from the icing profile. There is no need to install sensors for measuring values ​​such as the tension value of the transmission line. Not only can the icing thickness be detected without power outage maintenance, but it can also simplify the hardware equipment required for icing thickness detection and reduce maintenance complexity.

[0153] In addition, the real-time example of this application also incorporates micro-meteorological data, taking into account the impact of weather information on ice thickness, which can avoid misidentification of ice outlines caused by factors such as sunlight reflection, and improve the accuracy of ice thickness calculation.

[0154] Based on the same idea as the ice thickness detection method in the above embodiments, this application also provides an ice thickness detection device, which can be used to perform the above ice thickness detection method. For ease of explanation, the structural schematic diagram of the ice thickness detection device embodiment only shows the parts related to the embodiments of this application. Those skilled in the art will understand that the illustrated structure does not constitute a limitation on the device, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0155] like Figure 8 As shown, the icing thickness detection device includes an acquisition module 801, an extraction module 802, and a calculation module 803. In some embodiments, the above modules can be programmable software instructions stored in memory and executable by a processor. It is understood that in other embodiments, the above modules can also be program instructions or firmware embedded in the processor.

[0156] The acquisition module 801 is used to acquire images of icing on the transmission lines;

[0157] Extraction module 802 is used to input the icing image into a pre-trained icing semantic extraction model to obtain the icing semantic information of the transmission line;

[0158] The calculation module 803 is used to determine the icing profile of the transmission line based on the icing semantic information; and to determine the icing thickness of the transmission line based on the icing profile.

[0159] Figure 9 This is a schematic diagram of an embodiment of the electronic device 101 of this application.

[0160] The electronic device 101 includes a memory 20, a processor 30, and a computer program 40 stored in the memory 20 and executable on the processor 30. When the processor 30 executes the computer program 40, it implements the steps described in the above-described ice thickness detection method embodiment, for example... Figure 2 Steps 201 to 204 are shown.

[0161] For example, computer program 40 can also be divided into one or more modules / units, which are stored in memory 20 and executed by processor 30. The one or more modules / units can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of computer program 40 in electronic device 101. For example, it can be divided into... Figure 8 The diagram shows the acquisition module 801, the extraction module 802, and the calculation module 803.

[0162] Those skilled in the art will understand that the schematic diagram is merely an example of electronic device 101 and does not constitute a limitation on electronic device 101. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device 101 may also include input / output devices, network access devices, buses, etc.

[0163] Processor 30 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor, a single-chip microcomputer, or any conventional processor.

[0164] The memory 20 can be used to store computer programs 40 and / or modules / units. The processor 30 implements various functions of the electronic device 101 by running or executing the computer programs and / or modules / units stored in the memory 20 and by calling the data stored in the memory 20. The memory 20 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device 101 (such as audio data), etc. In addition, the memory 20 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other non-volatile solid-state storage device.

[0165] If the modules / units integrated in electronic device 101 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0166] In the several embodiments provided in this application, it should be understood that the disclosed electronic device 101 and method can be implemented in other ways. For example, the above-described embodiments of the electronic device 101 are merely illustrative; for example, the division of the units is merely a logical functional division, and there may be other division methods in actual implementation.

[0167] Furthermore, the functional units in the various embodiments of this application can be integrated into the same processing unit, or each unit can exist physically separately, or two or more units can be integrated into the same unit. The integrated units described above can be implemented in hardware or in the form of hardware plus software functional modules.

[0168] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from the spirit or essential characteristics of this application. Therefore, the embodiments should be considered exemplary and not restrictive in all respects. Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units or electronic device 101 recited in the claims of electronic device 101 may also be implemented by the same unit or electronic device 101 through software or hardware. The terms "first," "second," etc., are used to indicate names and do not indicate any particular order.

[0169] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit it. Although this application has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of this application without departing from the spirit and scope of the technical solutions of this application.

Claims

1. A method for detecting ice thickness, characterized in that, include: Acquire images of icing on power transmission lines; The icing images are input into a pre-trained icing semantic extraction model to obtain icing semantic information of the transmission line. The icing semantic extraction model includes a meteorological analysis model and a semantic segmentation model, where the meteorological analysis model is a CNN model. The training steps of the icing semantic extraction model include: acquiring icing image samples and micro-meteorological data samples; obtaining a first loss function value corresponding to the current semantic segmentation model based on the icing image samples and the current semantic segmentation model; obtaining a second loss function value corresponding to the current meteorological analysis model based on the micro-meteorological data samples and the current meteorological analysis model; obtaining a total loss function value of the icing semantic extraction model based on the first loss function value and the second loss function value; if it is determined that the icing semantic extraction model has not converged based on the total loss function value, adjusting the model parameters in the current semantic segmentation model and the current meteorological analysis model to obtain an adjusted icing semantic extraction model; continuing to train the adjusted icing semantic extraction model until it converges. Based on the icing semantic information, the icing profile of the transmission line is determined; Based on the icing profile, the icing thickness of the transmission line is determined.

2. The method for detecting ice thickness as described in claim 1, characterized in that, The step of inputting the icing image into a pre-trained icing semantic extraction model to obtain the icing semantic information includes: Acquire micrometeorological data that matches the time of the icing image capture; The micrometeorological data is input into the meteorological analysis model to obtain the weather information at the time the icing image was taken; Detect whether the weather information at the time the icing image was taken meets the meteorological conditions for icing; If the weather information at the time the icing image was taken meets the icing meteorological conditions, the icing image is input into the semantic segmentation model to obtain the icing semantic information.

3. The method for detecting ice thickness as described in claim 1, characterized in that, The icing semantic extraction model includes: a first feature pyramid and a second feature pyramid; The first feature pyramid is used to downsample the icing image several times to obtain a first matrix; The second feature pyramid is used to upsample the first matrix several times to obtain the icing semantic information.

4. The method for detecting ice thickness as described in any one of claims 1 to 3, characterized in that, The icing profile includes a first profile line and a second profile line along the transmission line. Determining the icing thickness of the transmission line based on the icing profile includes: Find the first position on the first contour line and the second position on the second contour line; The distance between the first position and the second position is the shortest distance between the first contour line and the second contour line; Based on the first location and the second location, the icing diameter of the transmission line is determined; The ice thickness is determined based on the ice diameter and the diameter of the transmission line.

5. The method for detecting ice thickness as described in claim 4, characterized in that, Determining the icing diameter of the transmission line based on the first and second locations includes: Obtain a first set of neighboring points, the first set of neighboring points including the first position and a plurality of positions adjacent to the first position on the first contour line; Calculate the shortest distance between each position in the first set of nearest points and the second contour line to obtain multiple distance values; The icing diameter is determined based on the multiple distance values.

6. The method for detecting ice thickness as described in claim 1, characterized in that, Determining the icing profile of the transmission line based on the icing semantic information includes: The icing semantic information is binarized to obtain a binarized image; The icing contour of the transmission line is extracted from the binarized image.

7. A device for detecting ice thickness, characterized in that, include: The acquisition module is used to acquire images of icing on transmission lines; An extraction module is used to input the icing image into a pre-trained icing semantic extraction model to obtain the icing semantic information of the transmission line. The icing semantic extraction model includes a meteorological analysis model and a semantic segmentation model, wherein the meteorological analysis model is a CNN model. The training steps of the icing semantic extraction model include: acquiring icing image samples and micro-meteorological data samples; obtaining a first loss function value corresponding to the current semantic segmentation model based on the icing image samples and the current semantic segmentation model; obtaining a second loss function value corresponding to the current meteorological analysis model based on the micro-meteorological data samples and the current meteorological analysis model; obtaining a total loss function value of the icing semantic extraction model based on the first loss function value and the second loss function value; if it is determined that the icing semantic extraction model has not converged based on the total loss function value, adjusting the model parameters in the current semantic segmentation model and the current meteorological analysis model to obtain an adjusted icing semantic extraction model; continuing to train the adjusted icing semantic extraction model until the icing semantic extraction model converges. The calculation module is used to determine the icing profile of the transmission line based on the icing semantic information; and to determine the icing thickness of the transmission line based on the icing profile.

8. An electronic device, the electronic device comprising a processor and a memory, characterized in that, The memory is used to store instructions, and the processor is used to call the instructions in the memory to cause the electronic device to perform the ice thickness detection method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed on an electronic device, cause the electronic device to perform the ice thickness detection method as described in any one of claims 1 to 6.

Images