Icing Detection Method, Device, Equipment and Storage Medium for Micro-Terrain of Transmission Lines

Abstract

The present invention relates to the technical field of power grids, and discloses a method, device, equipment and storage medium for detecting icing on micro-topography of transmission lines. The method includes: acquiring an image of a transmission line, and preprocessing the transmission line image through a geographic information system to obtain a micro-topography image of the transmission line; inputting the micro-topography image of the transmission line into a pre-trained icing detection model for micro-topography of transmission lines, and obtaining an icing detection area and a corresponding icing category output by the icing detection model for micro-topography of transmission lines; and visually displaying the icing detection area and the corresponding icing category. The solution of this embodiment can strengthen the icing feature extraction ability and the position feature capture ability by combining a multi-branch icing feature enhancement module and an icing position feature attention enhancement module for icing recognition, and can realize rapid and accurate recognition of icing on the micro-topography along the transmission line in a low-temperature environment.

Description

Technical Field

[0001] The present invention relates to the technical field of power grids, and in particular, to a method, device, equipment and storage medium for detecting icing on micro-topography of transmission lines. Background Art

[0002] The micro-topography of a transmission line refers to the topographic features along the power transmission line. During the layout of the transmission line, due to the irregularity of the terrain and the change of the terrain, the line may pass through various different terrain areas, and these topographic features are called micro-topography. Micro-topography icing refers to the icing phenomenon that occurs in local areas due to specific topographic features or meteorological conditions in these micro-topographies. This icing may not be evenly distributed like in flat areas, but concentrated in some terrain depressions or shadows, increasing the non-uniformity and risk of icing.

[0003] When icing exists on the micro-topography of a transmission line, it may increase the lateral wind load on the pole tower, reducing the stability of the pole tower and thus increasing the risk of pole tower collapse. In addition, it will also make the transmission channel environment around the transmission line dangerous, threatening the surrounding residents and traffic. Therefore, in order to ensure the safety of the transmission line, it is very necessary to detect the icing on the micro-topography of the transmission line.

[0004] Currently, the existing methods for detecting icing on the micro-topography of transmission lines mainly adopt manual methods or sensor detection methods. Among them, for the manual method, not only a large amount of human resources are required, but there is also a certain degree of danger, and the real-time detection effect cannot be achieved, resulting in a time delay. For the sensor detection method, both the efficiency and accuracy of automatic recognition are relatively low, and it cannot meet the actual needs. Summary of the Invention

[0005] The present invention provides a method, device, equipment and storage medium for detecting icing on the micro-topography of transmission lines, which can achieve rapid and accurate identification of icing on the micro-topography along the transmission line in a low-temperature environment, avoid waste of human resources, and ensure the personal safety of maintenance personnel.

[0006] According to an aspect of the present invention, there is provided a method for detecting icing on the micro-topography of a transmission line, including:

[0007] Obtain an image of the transmission line, and preprocess the image of the transmission line through a geographic information system to obtain an image of the micro-topography of the transmission line;

[0008] Input the image of the micro-topography of the transmission line into a pre-trained icing detection model for the micro-topography of the transmission line, and obtain the icing detection area and the corresponding icing category output by the icing detection model for the micro-topography of the transmission line;

[0009] Among them, the transmission line micro-topography icing detection model includes a multi-branch icing feature enhancement module and an icing position feature attention enhancement module;

[0010] Visualize the icing detection area and the corresponding icing category.

[0011] According to another aspect of the present invention, there is provided an icing detection device for transmission line micro-topography, including:

[0012] An image acquisition module, configured to acquire a transmission line image, and preprocess the transmission line image through a geographic information system to obtain a transmission line micro-topography image;

[0013] An icing detection module, configured to input the transmission line micro-topography image into a pre-trained transmission line micro-topography icing detection model, and obtain the icing detection area and the corresponding icing category output by the transmission line micro-topography icing detection model;

[0014] Among them, the transmission line micro-topography icing detection model includes a multi-branch icing feature enhancement module and an icing position feature attention enhancement module;

[0015] A category display module, configured to visually display the icing detection area and the corresponding icing category.

[0016] According to another aspect of the present invention, there is provided an electronic device, the electronic device includes:

[0017] At least one processor; and

[0018] A memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the icing detection method for transmission line micro-topography according to any embodiment of the present invention.

[0020] According to another aspect of the present invention, there is provided a computer-readable storage medium, the computer-readable storage medium stores a computer program, and the computer program is used to implement the icing detection method for transmission line micro-topography according to any embodiment of the present invention when executed by a processor.

[0021] According to another aspect of the present invention, there is provided a computer program product, including a computer program, and the computer program implements the icing detection method for transmission line micro-topography according to any embodiment of the present invention when executed by a processor.

[0022] The technical solution of the embodiment of the present invention includes: obtaining an image of a transmission line, preprocessing the image of the transmission line through a geographic information system to obtain a micro-topography image of the transmission line; inputting the micro-topography image of the transmission line into a pre-trained icing detection model for the micro-topography of the transmission line, and obtaining an icing detection area and a corresponding icing category output by the icing detection model for the micro-topography of the transmission line; wherein, the icing detection model for the micro-topography of the transmission line includes a multi-branch icing feature enhancement module and an icing position feature attention enhancement module; visually displaying the icing detection area and the corresponding icing category; by combining the multi-branch icing feature enhancement module and the icing position feature attention enhancement module for icing identification, the ability to extract icing features and capture position features can be enhanced, and rapid and accurate identification of icing on the micro-topography along the transmission line in a low-temperature environment can be achieved.

[0023] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0025] Figure 1 is a flowchart of a method for detecting icing on the micro-topography of a transmission line according to Embodiment 1 of the present invention;

[0026] Figure 2 is a schematic diagram of the network structure of an icing detection model for the micro-topography of a transmission line according to Embodiment 1 of the present invention;

[0027] Figure 3 is a schematic diagram of the structure of a backbone module according to Embodiment 1 of the present invention;

[0028] Figure 4 is a schematic diagram of the structure of an MIE module according to Embodiment 1 of the present invention;

[0029] Figure 5 is a schematic diagram of the structure of a PCM module according to Embodiment 1 of the present invention;

[0030] Figure 6 is a schematic diagram of the structure of an icing detection device for the micro-topography of a transmission line according to Embodiment 2 of the present invention;

[0031] Figure 7It is a schematic structural diagram of an electronic device for implementing the icing detection method of the transmission line microtopography in the embodiments of the present invention. Detailed implementation manners

[0032] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0033] It should be noted that the terms "first", "second", "target", etc. in the description and claims of the present invention and the above drawings are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units does not have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0034] Embodiment 1

[0035] Figure 1 The flowchart of an icing detection method for the transmission line microtopography provided in Embodiment 1 of the present invention is applicable to the situation of real-time monitoring and identification of the icing condition of the transmission line microtopography. This method can be executed by an icing detection device for the transmission line microtopography, and the icing detection device for the transmission line microtopography can be implemented in the form of hardware and / or software. Typically, the icing detection device for the transmission line microtopography can be configured in an electronic device, such as a computer device or a server, etc. As Figure 1 shown, this method includes:

[0036] S110. Obtain a transmission line image, and preprocess the transmission line image through a geographic information system to obtain a transmission line microtopography image.

[0037] In this embodiment, the transmission line image can be obtained by aerial photography with a drone or by a camera fixed on a transmission tower to photograph the defective transmission line.

[0038] Among them, a Geographic Information System (GIS) is a system for processing and managing geospatial data. Image data such as satellite images, aerial photographs, and terrain models can all be processed through the GIS system. In this embodiment, the captured transmission line images can be preprocessed through the GIS system. For example, micro-topography area segmentation, micro-topography category judgment, etc., and the preprocessed transmission line images can be used as transmission line micro-topography images. Typically, micro-topography categories can include valleys, hills, rivers, trees, etc.

[0039] Optionally, preprocessing the transmission line images through the geographic information system to obtain transmission line micro-topography images may include:

[0040] Obtaining a plurality of selected regions corresponding to the transmission line images through the geographic information system, and obtaining supervised classification corresponding to each of the selected regions through the maximum likelihood classification method;

[0041] Determining transmission line micro-topography images among the plurality of selected regions according to the supervised classification corresponding to each of the selected regions.

[0042] Specifically, the GIS tool software can be used to import the transmission line images, and by dragging the vector range polygon, the icing areas in the images can be selected to obtain a plurality of selected regions corresponding to the transmission line images. Then, supervised classification can be performed on the selected regions based on the maximum likelihood classification method. In the maximum likelihood classification, it is assumed that the pixel values of each supervised classification follow a multivariate normal distribution (multivariate Gaussian distribution), and its probability density function can be expressed as:

[0043]

[0044] Among them, x is an n-dimensional vector representing the eigenvalue of a pixel, usually the spectral feature of the pixel, μ is an n-dimensional vector representing the mean vector of the supervised classification, and Σ is an n×n covariance matrix representing the covariance matrix of the supervised classification.

[0045] The decision rule expression in the maximum likelihood classification is Among them, represents the supervised classification of the pixel, i represents the index of the classification, and P(x|μ i ,∑ i ) represents the probability that the pixel x belongs to the supervised classification i. Thus, the probability density of the selected region can be calculated first, and then based on the decision rule expression, according to this probability density, the supervised classification corresponding to the selected region can be determined. Finally, the selected regions belonging to the icing category can be screened out from all the selected regions to obtain the transmission line micro-topography images.

[0046] S120. Input the micro-topography image of the transmission line into a pre-trained ice detection model for the micro-topography of the transmission line, and obtain the ice detection region and the corresponding ice category output by the ice detection model for the micro-topography of the transmission line.

[0047] Among them, the ice detection model for the micro-topography of the transmission line may include a multi-branch ice feature enhancement module (Multi-Branch Ice-Feature Enhancement Module, MIE) and a point-wise convolutional block attention module (Point-Wise Conv Convolutional Block Attention Module, PCM).

[0048] In this embodiment, by adding a multi-branch ice feature enhancement module and a point-wise convolutional block attention module to the deep learning model, the ice recognition ability of the model for the micro-topography can be improved. Typically, the multi-branch ice feature enhancement module can adopt a multi-branch structure, and respectively introduce a deformable attention mechanism and a spatial attention mechanism to enhance the attention to the local key regions in the feature map; the point-wise convolutional block attention module can enhance the attention to the ice feature position information in the micro-topography along the transmission line under low-temperature conditions by introducing a channel attention mechanism. In a specific example, the network structure of the ice detection model for the micro-topography of the transmission line can be as Figure 2 shown. MIE, PCM, a convolutional layer (Conv), a BN (Batch Normalization) layer, and a ReLU (Rectified Linear Unit) layer form a backbone module (PCM-MIE Block, PM Block), and the structure of the PM Block can be as Figure 3 shown.

[0049] Specifically, the ice detection model for the micro-topography of the transmission line can be used to perform ice recognition on the micro-topography image of the transmission line to obtain a predicted bounding box and an ice category, and the overlapping bounding boxes can be removed through non-maximum suppression to obtain the final ice detection region and the corresponding ice category.

[0050] Optionally, before inputting the micro-topography image of the transmission line into the pre-trained ice detection model for the micro-topography of the transmission line, it may further include: establishing an initial detection model based on preset initial model parameters and hyperparameters; collecting the micro-topography ice image of the transmission line, and preprocessing the micro-topography ice image of the transmission line through a GIS system to obtain a recognition data set, and the recognition data set can be divided into a training set, a validation set, and a test set; training the initial detection model based on the training set, the validation set, and the test set to obtain a trained ice detection model for the micro-topography of the transmission line.

[0051] In a specific example, in a power transmission line, a camera or inspection drone deployed on a tower is used to collect images of ice-covered areas along the micro-topography of the transmission line. After collecting 2,000 images of ice-covered micro-topography of the transmission line {I1,I2,…,I 2000}, the ArcGIS tool is used to perform supervised classification data preprocessing on the original transmission line micro-topography ice-covered images. First, 50 transmission line micro-topography ice-covered images {I1, I2, …, I 50} as the training set, select the "Draw Polygon" vector range polygon drawing module in turn, drag the vector range polygon in the module, select the ice-covered area in the image, and after drawing the ice-covered area polygons of several transmission line micro-terrain ice-covered images, the transmission line micro-terrain ice-covered image training set {I1',I2',…,I 50 '}. Then, the ice-covered polygons in the training set are merged. After the merging is completed, the original multiple polygons are integrated into the information of the same ice-covered category, and the name and color of the information are configured. After the configuration is completed, a .gsg format file can be obtained, which is the input file for supervised classification. Then, the file and the transmission line micro-topography ice-covered image are imported into the image classification module, and the maximum likelihood classification is selected to perform data preprocessing on the transmission line micro-topography ice-covered image classification.

[0052] Furthermore, the LabelImg image annotation tool can be used to annotate the categories and coordinate frame information of the ice-covered areas of all transmission line micro-topography ice-covered images. In the LabelImg tool, first select the VOC annotation format so that the generated annotation file is in .xml format. Then, for each collected transmission line micro-topography ice-covered image data, drag the anchor box to select it, and select the corresponding category label after selecting it (when annotating a new category for the first time, enter the category information in the category pre-selection box, and the category of the data is a single ice-covered category). Each image contains one or more ice-covered areas. If there are multiple ice-covered areas, the corresponding image ice-covered targets need to be selected with multiple anchor boxes. After the ice-covered targets of all transmission line micro-topography ice-covered images are annotated, 2000 annotation files in xml format can be obtained. The name of each annotation file is consistent with the name of the corresponding transmission line micro-topography ice-covered image, and each annotation file can contain annotation anchor box information and category information of multiple targets.

[0053] After the annotation of the icing images of the micro-topography of the transmission line is completed, the dataset can be divided. To ensure that the detection model has sufficient training data, the training set, validation set, and test set can be randomly divided in a ratio of 8:1:1. After division, 1600 pieces of training data, 200 pieces of validation data, and 200 pieces of test data can be obtained.

[0054] Then, the parameters and hyperparameters of the icing detection model for the micro-topography of the transmission line can be initialized, such as the number of training times, momentum factor, batch size, optimizer, learning rate type, and initial learning rate, etc. After completing the above preparations, the above training set, validation set, and test set can be used for model training. Typically, the initialized optimizer can be the optimizer optimizer, the initialized number of training rounds can be 200, the learning rate type can be the cosine learning rate, and the initial learning rate can be 0.001.

[0055] In each round of training, the model grabs the preprocessed icing data of the micro-topography of the transmission line in the training set according to the preset batch size for training. Suppose one piece of image data of the icing of the micro-topography of the transmission line input into the model for training is Y1. First, Y1 is input into the embedding layer (the parameters of the embedding layer will be learned and optimized during training, and the embedding layer can be regarded as matrix operation) for feature magnification to obtain the output Y2 (taking the dimension size of Y1 as 640×640×3 as an example, inputting Y1 into the embedding layer to obtain the output Y2 of 320×320×64 for subsequent feature extraction, and this value varies with the dimensions of different input icing images of the micro-topography of the transmission line). Then, Y2 is input into the PM Block. In the PM Block, Y2 is successively input into the MIE module, PCM module, convolutional layer, BN layer, and ReLU activation function to obtain the output Y3. Subsequently, Y3 is input into the combination of n embedding layers and PM Block for feature extraction (the value of n is determined by the network performance requirements. A higher n value has higher accuracy, and relatively a smaller n value has faster detection speed) to obtain the output Y7. Y7 is input into the embedding layer, and this embedding layer plays a role in adjusting the feature map size and channels to obtain the output Y8. Subsequently, Y8 is input into the global pooling layer, and this global pooling layer reduces Y8 to a vector Y9 of a fixed size to reduce the number of model parameters and computational amount, and at the same time improve the generalization ability and anti-overfitting ability of the model. Then, Y9 is successively input into the convolutional layer and fully connected layer to map the features to the final output Y of the model 11 。Training updates the loss function and various parameters in the model through backpropagation. The loss function uses the cross-entropy loss function. After each round of training is completed, the icing data of the micro-topography of the transmission line in the validation set is also grabbed according to the preset batch size for validation to check the effectiveness of each round of training. When the loss value of the model training tends to converge, the training of the icing detection model for the micro-topography of the transmission line ends.

[0056] S130. Visualize the icing detection area and the corresponding icing category.

[0057] Specifically, after obtaining the icing detection area and the corresponding icing category, the icing detection result can be visually displayed to users on the corresponding page, so that line maintenance personnel can timely check the icing condition of the transmission line.

[0058] The technical solution of the embodiment of the present invention obtains a transmission line image, preprocesses the transmission line image through a geographic information system to obtain a micro-topography image of the transmission line; inputs the micro-topography image of the transmission line into a pre-trained icing detection model for the micro-topography of the transmission line, and obtains the icing detection area and the corresponding icing category output by the icing detection model for the micro-topography of the transmission line; wherein, the icing detection model for the micro-topography of the transmission line includes a multi-branch icing feature enhancement module and an icing position feature attention enhancement module; visualizes the icing detection area and the corresponding icing category; by combining the multi-branch icing feature enhancement module and the icing position feature attention enhancement module for icing recognition, the icing feature extraction ability and the position feature capture ability can be enhanced, and rapid and accurate recognition of icing on the micro-topography along the transmission line in a low-temperature environment can be achieved.

[0059] In an optional implementation manner of this embodiment, inputting the micro-topography image of the transmission line into a pre-trained icing detection model for the micro-topography of the transmission line, and obtaining the icing detection area and the corresponding icing category output by the icing detection model for the micro-topography of the transmission line may include:

[0060] Through the multi-branch icing feature enhancement module, perform initial feature extraction on the micro-topography image of the transmission line to obtain an initial feature map;

[0061] Input the initial feature map into the icing position feature attention enhancement module to obtain an intermediate feature map;

[0062] Input the intermediate feature map into the first basic model unit to obtain an output feature map, and based on the output feature map, obtain the icing detection area and the corresponding icing category.

[0063] Specifically, the micro-topography image of the transmission line can be successively processed by the multi-branch icing feature enhancement module, the icing position feature attention enhancement module and the first basic model unit in the backbone module to obtain an output feature map, and the output feature map can be processed according to the network structure of the icing detection model for the micro-topography of the transmission line to obtain the final icing detection area and the corresponding icing category. Among them, the last output channels of the icing detection model for the micro-topography of the transmission line respectively include a regression part and a classification part, which are used to output the predicted bounding box and the target category.

[0064] Optionally, through a multi-branch icing feature enhancement module, initial feature extraction is performed on the micro-topography image of the transmission line to obtain an initial feature map, which may include:

[0065] Obtain the original feature map corresponding to the micro-topography image of the transmission line, and input the original feature map into the first basic model unit to obtain an input feature map;

[0066] In the first branch, a deformable attention method is used to obtain a first-branch feature map according to the input feature map;

[0067] In the second branch, the input feature map is sequentially input into the first basic model unit and the second basic model unit to obtain a second-branch feature map;

[0068] In the third branch, a spatial attention method is used to obtain a third-branch feature map according to the input feature map;

[0069] The first-branch feature map, the second-branch feature map, and the third-branch feature map are concatenated to obtain a temporary feature map, and the temporary feature map is input into the first basic model unit to obtain an initial feature map.

[0070] Among them, the first basic model unit includes a convolutional layer (Conv), a batch normalization layer (BN), and a rectified linear unit layer (ReLU), and the second basic model unit includes a point-wise convolutional layer (Point-Wise Conv, PW Conv), a batch normalization layer (BN), and a rectified linear unit layer (ReLU). Among them, the point-wise convolutional layer can be a convolutional layer with a kernel size of 1×1.

[0071] In this embodiment, by using the MIE module, feature extraction can be performed on the icing image of the micro-topography of the transmission line through a multi-branch structure, paying attention to the local key regions in the feature map while suppressing the background noise in the image, rather than processing the entire image data, which can significantly reduce the computational amount brought by the attention operation and enhance the feature extraction ability for the icing image of the micro-topography of the transmission line.

[0072] In a specific example, the structure of the MIE module can be as Figure 4 shown, where the uppermost branch is the first branch, the middle branch is the second branch, and the lowermost branch is the third branch. The first branch uses a deformable attention (Deformable Attention, DAttention) method, the third branch uses a spatial attention (spatialattention) method, and the second branch is composed of a first basic model unit and a second basic model unit connected in sequence.

[0073] Among them, the calculation process of the deformable attention method is Among them, T2 represents the feature map input to the first branch, and W0 represents the output projection matrix. represents the attention output result of the m-th head, and its calculation process is m = 1, 2, …, M, q (m) , k (m)T and v (m) respectively represent the query vector, key-value vector, and value vector under the m-th head. d represents the dimension of each head, and the calculation method is the number of input channels divided by the number of heads. σ represents the softmax function. Assuming the input is {x1, x2, …, x i}}, its calculation formula is

[0074] In the third branch, the spatial attention method is adopted to first perform maximum pooling and average pooling on the input feature map to extract the maximum value and average value at each spatial position respectively; then, these two results can be concatenated to obtain a new feature map, and its number of channels is twice that of the original; subsequently, the obtained new feature map can be input into a convolutional layer, and the convolutional result is processed through the sigmoid function to normalize its value in the range of [0, 1] to obtain the final feature map of the third branch.

[0075] Finally, the output feature maps of the three branches can be merged and concatenated (Concat) to obtain a temporary feature map, and the temporary feature map can be processed through a combination of a Conv layer, a BN layer, and a ReLU layer to obtain the initial feature map.

[0076] Optionally, inputting the initial feature map into the icing position feature attention enhancement module to obtain an intermediate feature map may include:

[0077] Inputting the initial feature map into the second basic model unit to obtain the current feature map, and splitting the current feature map along the channel dimension to obtain the first split feature map and the second split feature map;

[0078] Adopting the convolutional block attention method to obtain the icing position attention enhancement feature map according to the first split feature map, and obtaining the intermediate feature map according to the icing position attention enhancement feature map and the second split feature map.

[0079] In this embodiment, the PCM module can enhance the attention to the icing feature position information in the micro-topography along the transmission line under low-temperature conditions, and at the same time can control the number of model parameters, and can improve the efficiency of icing detection.

[0080] In a specific example, the structure of the PCM module can be as Figure 5As shown in the figure. First, the initial feature map can be input into the second basic model unit composed of a PWConv layer, a BN layer, and a Relu layer to obtain the current feature map, and the current feature map can be input into the feature map splitting module (Split) to split it into two parts along the channel dimension, namely the first split feature map and the second split feature map. Subsequently, through CBAM (Convolutional Block Attention Module), the convolutional block attention method can be used to process the first split feature map, and the processing result can be further input into the first basic model unit to obtain the icing position attention enhanced feature map. Finally, the icing position attention enhanced feature map and the second split feature map can be merged and concatenated (Concat) to obtain the intermediate feature map.

[0081] Among them, the calculation process of the CBAM attention mechanism is as follows. First, based on the formula the input feature map F3 is input into a channel attention, and M c represents the channel attention mechanism; subsequently, the output result can be input into a spatial attention mechanism again, as shown in the formula M s represents the spatial attention mechanism to obtain the final output.

[0082] In this embodiment, the PCM module is adopted. The detection efficiency of the model can be ensured through the PWConv layer, and secondly, the ability of the model to capture position information can be improved through the CBAM module.

[0083] The technical solution of the embodiment of the present invention designs a feature extraction module according to the icing target in the micro-topography scene of the transmission line, enhances the ability of the model to extract icing features while controlling the number of parameters, and secondly, through the GIS system, the dataset collected along the transmission line can be preprocessed, and a more efficient and accurate model can be trained to achieve fast and accurate icing recognition in the actual transmission line.

[0084] Embodiment 2

[0085] Figure 6 This is a schematic structural diagram of an icing detection device for the micro-topography of a transmission line provided by the second embodiment of the present invention. As Figure 6 shown, the device includes: an image acquisition module 210, an icing detection module 220, and a category display module 230; among them,

[0086] The image acquisition module 210 is used to acquire a transmission line image and preprocess the transmission line image through a geographic information system to obtain a micro-topography image of the transmission line;

[0087] An icing detection module 220, configured to input the micro-topography image of the transmission line into a pre-trained icing detection model for the micro-topography of the transmission line, and obtain an icing detection area and a corresponding icing category output by the icing detection model for the micro-topography of the transmission line;

[0088] Wherein, the icing detection model for the micro-topography of the transmission line includes a multi-branch icing feature enhancement module and an icing position feature attention enhancement module;

[0089] A category display module 230, configured to visually display the icing detection area and the corresponding icing category.

[0090] The technical solution of the embodiment of the present invention obtains a transmission line image, preprocesses the transmission line image through a geographic information system to obtain a micro-topography image of the transmission line; inputs the micro-topography image of the transmission line into a pre-trained icing detection model for the micro-topography of the transmission line, and obtains an icing detection area and a corresponding icing category output by the icing detection model for the micro-topography of the transmission line; wherein, the icing detection model for the micro-topography of the transmission line includes a multi-branch icing feature enhancement module and an icing position feature attention enhancement module; visually display the icing detection area and the corresponding icing category; by combining the multi-branch icing feature enhancement module and the icing position feature attention enhancement module for icing identification, the icing feature extraction ability and the position feature capture ability can be enhanced, and rapid and accurate identification of icing on the micro-topography along the transmission line in a low-temperature environment can be achieved.

[0091] Optionally, the image acquisition module 210 is specifically configured to obtain a plurality of boxed areas corresponding to the transmission line image through a geographic information system, and obtain a supervised classification corresponding to each of the boxed areas through a maximum likelihood classification method;

[0092] Determine a micro-topography image of the transmission line among the plurality of boxed areas according to the supervised classification corresponding to each of the boxed areas.

[0093] Optionally, the icing detection module 220 is specifically configured to perform initial feature extraction on the micro-topography image of the transmission line through a multi-branch icing feature enhancement module to obtain an initial feature map;

[0094] Input the initial feature map into the icing position feature attention enhancement module to obtain an intermediate feature map;

[0095] Input the intermediate feature map into a first basic model unit to obtain an output feature map, and obtain an icing detection area and a corresponding icing category according to the output feature map.

[0096] Optionally, the icing detection module 220 is specifically configured to obtain the original feature map corresponding to the micro-topography image of the transmission line, and input the original feature map into the first basic model unit to obtain the input feature map;

[0097] In the first branch, the deformable attention method is used to obtain the first branch feature map according to the input feature map;

[0098] In the second branch, the input feature map is sequentially input into the first basic model unit and the second basic model unit to obtain the second branch feature map;

[0099] In the third branch, the spatial attention method is used to obtain the third branch feature map according to the input feature map;

[0100] The first branch feature map, the second branch feature map, and the third branch feature map are spliced to obtain a temporary feature map, and the temporary feature map is input into the first basic model unit to obtain the initial feature map.

[0101] Optionally, the icing detection module 220 is specifically configured to input the initial feature map into the second basic model unit to obtain the current feature map, and divide the current feature map along the channel dimension to obtain the first divided feature map and the second divided feature map;

[0102] The convolutional block attention method is used to obtain the icing position attention enhanced feature map according to the first divided feature map, and the intermediate feature map is obtained according to the icing position attention enhanced feature map and the second divided feature map.

[0103] Optionally, the first basic model unit includes a convolutional layer, a batch normalization layer, and a rectified linear unit layer, and the second basic model unit includes a pointwise convolutional layer, a batch normalization layer, and a rectified linear unit layer.

[0104] The icing detection device for the micro-topography of the transmission line provided by the embodiment of the present invention can execute the icing detection method for the micro-topography of the transmission line provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the method.

[0105] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision, and disclosure of the user's personal information and other processes all comply with the provisions of relevant laws and regulations and do not violate public order and good customs.

[0106] Embodiment III

[0107] Figure 7FIG. 0 shows a schematic structural diagram of an electronic device 30 that can be used to implement an embodiment of the present invention. The electronic device 30 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device 30 can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0108] As Figure 7 shown, the electronic device 30 includes at least one processor 31, and a memory communicatively connected to the at least one processor 31, such as a read-only memory (ROM) 32, a random access memory (RAM) 33, etc. The memory stores a computer program executable by the at least one processor. The processor 31 can perform various appropriate actions and processes according to the computer program stored in the read-only memory 32 or the computer program loaded from the storage unit 38 into the random access memory 33. In the RAM 33, various programs and data required for the operation of the electronic device 30 can also be stored. The processor 31, the ROM 32, and the RAM 33 are connected to each other through a bus 34. An input / output (I / O) interface 35 is also connected to the bus 34.

[0109] Multiple components in the electronic device 30 are connected to the I / O interface 35, including: an input unit 36, such as a keyboard, a mouse, etc.; an output unit 37, such as various types of displays, speakers, etc.; a storage unit 38, such as a magnetic disk, an optical disk, etc.; and a communication unit 39, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 39 allows the electronic device 30 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0110] The processor 31 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 31 include, but are not limited to, a central processing unit, a graphics processing unit, various dedicated artificial intelligence computing chips, various processors running machine learning model algorithms, a digital signal processor, and any appropriate processor, controller, microcontroller, etc. The processor 31 executes the various methods and processes described above, such as the icing detection method for the micro-topography of transmission lines.

[0111] In some embodiments, the icing detection method for the micro-topography of a transmission line can be implemented as a computer program, which is tangibly embodied in a computer-readable storage medium, such as storage unit 38. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 30 via the ROM 32 and / or the communication unit 39. When the computer program is loaded into the RAM 33 and executed by the processor 31, one or more steps of the above-described icing detection method for the micro-topography of a transmission line can be performed. Alternatively, in other embodiments, the processor 31 can be configured to perform the icing detection method for the micro-topography of a transmission line by any other suitable means (e.g., by means of firmware).

[0112] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays, application-specific integrated circuits, application-specific standard products, systems-on-a-chip, programmable logic devices loaded with a program, computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a dedicated or general-purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0113] The computer programs for implementing the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a dedicated computer, or other programmable data processing device, such that when the computer programs are executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer programs can be executed entirely on the machine, partially on the machine, as a stand-alone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0114] In the context of the present invention, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, the computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical fiber, a portable compact disk read-only memory, an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0115] To provide for interaction with a user, the systems and techniques described herein can be implemented on an electronic device 30 having: a display device (e.g., a cathode ray tube or a liquid crystal display) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device 30. Other kinds of devices can also be used to provide for interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic, speech, or tactile input).

[0116] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or in a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local area networks, wide area networks, blockchain networks, and the Internet.

[0117] The computing system can include a client and a server. The client and the server are generally remote from each other and typically interact through a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server.

[0118] This embodiment may further include a computer program product, which includes a computer program that, when executed by a processor, implements the icing detection method for micro-topography of a transmission line provided in any embodiment of the present invention.

[0119] It should be understood that various forms of the processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in the present invention can be executed in parallel, sequentially, or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and no limitations are imposed herein.

[0120] The above specific embodiments do not constitute a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for detecting icing on micro-topography of a transmission line, characterized in that, Including: Obtain an image of a transmission line, and preprocess the transmission line image through a geographic information system to obtain a micro-topography image of the transmission line; Input the micro-topography image of the transmission line into a pre-trained icing detection model for the micro-topography of the transmission line, and obtain the icing detection area and the corresponding icing category output by the icing detection model for the micro-topography of the transmission line; Among them, the icing detection model for the micro-topography of the transmission line includes a multi-branch icing feature enhancement module and an icing position feature attention enhancement module; Visualize the icing detection area and the corresponding icing category; Inputting the micro-topography image of the transmission line into a pre-trained icing detection model for the micro-topography of the transmission line, and obtaining the icing detection area and the corresponding icing category output by the icing detection model for the micro-topography of the transmission line includes: Through the multi-branch icing feature enhancement module, perform initial feature extraction on the micro-topography image of the transmission line to obtain an initial feature map; Input the initial feature map into the icing position feature attention enhancement module to obtain an intermediate feature map; Input the intermediate feature map into the first basic model unit, obtain an output feature map, and based on the output feature map, obtain the icing detection area and the corresponding icing category.

2. The method according to claim 1, wherein Preprocessing the transmission line image through a geographic information system to obtain a micro-topography image of the transmission line includes: Through the geographic information system, obtain a plurality of boxed areas corresponding to the transmission line image, and through the maximum likelihood classification method, obtain the supervised classification corresponding to each boxed area; Determine the micro-topography image of the transmission line among the plurality of boxed areas according to the supervised classification corresponding to each boxed area.

3. The method according to claim 1, characterized in that, Performing initial feature extraction on the micro-topography image of the transmission line through the multi-branch icing feature enhancement module to obtain an initial feature map includes: Obtain the original feature map corresponding to the micro-topography image of the transmission line, and input the original feature map into the first basic model unit to obtain an input feature map; In the first branch, use the deformable attention method to obtain a first branch feature map according to the input feature map; In the second branch, input the input feature map into the first basic model unit and the second basic model unit in sequence to obtain a second branch feature map; In the third branch, use the spatial attention method to obtain a third branch feature map according to the input feature map; Stitch the first branch feature map, the second branch feature map, and the third branch feature map to obtain a temporary feature map, and input the temporary feature map into the first basic model unit to obtain an initial feature map.

4. The method according to claim 1, wherein Inputting the initial feature map into the icing position feature attention enhancement module to obtain an intermediate feature map includes: Input the initial feature map into the second basic model unit to obtain a current feature map, and divide the current feature map along the channel dimension to obtain a first divided feature map and a second divided feature map; Use the convolutional block attention method to obtain an icing position attention enhancement feature map according to the first divided feature map, and obtain an intermediate feature map according to the icing position attention enhancement feature map and the second divided feature map.

5. The method according to any one of claims 3-4, characterized in that, The first basic model unit includes a convolutional layer, a batch normalization layer, and a rectified linear unit layer, and the second basic model unit includes a pointwise convolutional layer, a batch normalization layer, and a rectified linear unit layer.

6. An icing detection device for microtopography of a transmission line, characterized in that, Comprising: An image acquisition module, configured to acquire a transmission line image, and perform preprocessing on the transmission line image through a geographic information system to obtain a transmission line micro-topography image; An icing detection module, configured to input the transmission line micro-topography image into a pre-trained transmission line micro-topography icing detection model, and obtain an icing detection area and a corresponding icing category output by the transmission line micro-topography icing detection model; Wherein, the transmission line micro-topography icing detection model includes a multi-branch icing feature enhancement module and an icing position feature attention enhancement module; A category display module, configured to visually display the icing detection area and the corresponding icing category; The icing detection module is specifically configured to perform initial feature extraction on the transmission line micro-topography image through the multi-branch icing feature enhancement module to obtain an initial feature map; Input the initial feature map into the icing position feature attention enhancement module to obtain an intermediate feature map; Input the intermediate feature map into the first basic model unit to obtain an output feature map, and obtain an icing detection area and a corresponding icing category according to the output feature map.

7. An electronic device, characterized in that, The electronic device includes: At least one processor, and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the icing detection method for the transmission line micro-topography according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and the computer program is used to implement the icing detection method for the transmission line micro-topography according to any one of claims 1-5 when executed by a processor.

9. A computer program product, characterized in that, Including a computer program, and the computer program implements the icing detection method for the transmission line micro-topography according to any one of claims 1-5 when executed by a processor.

Images