Power transmission line insulator snow cover identification method and device and electronic equipment

Abstract

The application discloses a power transmission line insulator snow cover identification method and device and electronic equipment. The method comprises the following steps: acquiring multiple insulator images; using an image enhancement method to respectively perform image enhancement processing on the multiple insulator images to obtain multiple first insulator images; using a pre-constructed denoising convolutional neural network to respectively perform denoising processing on the multiple first insulator images to obtain multiple second insulator images; inputting the multiple second insulator images into a pre-constructed initial insulator snow cover identification model for training to obtain a target insulator snow cover identification model; and using the target insulator snow cover identification model to identify a to-be-identified insulator image to obtain a snow cover identification result corresponding to the to-be-identified insulator image. The application solves the technical problem of low identification accuracy of the power transmission line insulator snow cover identification method in the related art and the inability to identify potential fault risks of the power transmission line in a timely manner.

Description

Methods, devices and electronic equipment for identifying snow cover on transmission line insulators Technical Field

[0001] This invention relates to the field of power transmission line technology, and more specifically, to a method, apparatus, and electronic device for identifying snow cover on power transmission line insulators. Background Technology

[0002] Under high-altitude micro-meteorological conditions, snow accumulation on insulators poses a significant threat to power grid safety. Especially during cold winters, snow accumulation on insulators is a frequent occurrence, and in severe cases, it can lead to transmission line breakage and tower collapse, thereby threatening the safe operation of the power grid and even causing huge economic losses and casualties. Therefore, timely and accurate identification of snow accumulation on insulators is crucial. How to monitor the snow accumulation status of transmission line insulators under high-altitude micro-meteorological conditions in real time, so as to take timely measures to prevent accidents, is currently a hot research topic.

[0003] Currently, the main technologies for identifying snow cover on insulators include: automatic identification of snow cover on transmission lines using image processing technology; analysis and identification of snow cover using mechanical methods; and online monitoring methods for snow cover on transmission lines based on three-dimensional reconstruction. Among these, automatic identification of snow cover on transmission lines using image processing technology is mostly only applicable to insulator snow cover in ideal or simple environments, and it is difficult to achieve accurate and effective identification of insulator snow cover in complex environments around the clock.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This invention provides a method, apparatus, and electronic device for identifying snow cover on transmission line insulators, to at least solve the technical problems of low identification accuracy and inability to timely identify potential fault risks of transmission lines in related technologies.

[0006] According to one aspect of the present invention, a method for identifying snow-covered insulators of transmission lines is provided, comprising: acquiring multiple insulator images, wherein the multiple insulator images include snow-covered insulator images and snow-free insulator images acquired under various acquisition environments; performing image enhancement processing on the multiple insulator images respectively using an image enhancement method to obtain multiple first insulator images; performing denoising processing on the multiple first insulator images respectively using a pre-constructed denoising convolutional neural network to obtain multiple second insulator images; inputting the multiple second insulator images into a pre-constructed initial insulator snow-covered identification model for training to obtain a target insulator snow-covered identification model; and using the target insulator snow-covered identification model to identify the insulator image to be identified to obtain the snow-covered identification result corresponding to the insulator image to be identified.

[0007] According to another aspect of the present invention, a snow-covered insulator identification device for transmission lines is also provided, comprising: an acquisition module for acquiring multiple insulator images, wherein the multiple insulator images include snow-covered insulator images and snow-free insulator images acquired under various acquisition environments; a first image processing module for performing image enhancement processing on the multiple insulator images using an image enhancement method to obtain multiple first insulator images; a second image processing module for performing denoising processing on the multiple first insulator images using a pre-constructed denoising convolutional neural network to obtain multiple second insulator images; a training module for inputting the multiple second insulator images into a pre-constructed initial insulator snow-covered identification model for training to obtain a target insulator snow-covered identification model; and an image recognition module for using the target insulator snow-covered identification model to identify the insulator image to be identified, thereby obtaining the snow-covered identification result corresponding to the insulator image to be identified.

[0008] According to another aspect of the present invention, a non-volatile storage medium is also provided, which stores a plurality of instructions adapted for a processor to load and execute any one of the above-described methods for identifying snow cover on transmission line insulators.

[0009] According to another aspect of the present invention, an electronic device is also provided, including one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any one of the above-described methods for identifying snow cover on transmission line insulators.

[0010] In this embodiment of the invention, multiple insulator images are acquired, including snow-covered and snow-free images of insulators acquired under various acquisition environments. An image enhancement method is used to enhance each of the multiple insulator images, resulting in multiple first insulator images. A pre-constructed denoising convolutional neural network is then used to denoise each of the multiple first insulator images, resulting in multiple second insulator images. These second insulator images are then input into a pre-constructed initial insulator snow-covered recognition model for training, resulting in a target insulator snow-covered recognition model. The target insulator snow-covered recognition model is then used to recognize the insulator image to be identified, obtaining the corresponding snow-covered recognition result. This achieves the goal of accurately identifying and detecting insulator snow-covered conditions in complex environments, thereby improving the accuracy of insulator snow-covered recognition for transmission lines and enabling timely identification of potential fault risks in transmission lines. This solves the technical problem of low recognition accuracy and inability to timely identify potential fault risks in transmission line insulator snow-covered recognition methods in related technologies. Attached Figure Description

[0011] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0012] Figure 1 is a flowchart of a method for identifying snow cover on transmission line insulators according to an embodiment of the present invention;

[0013] Figure 2 is a schematic diagram of an optional denoising convolutional neural network according to an embodiment of the present invention;

[0014] Figure 3 is a schematic diagram of a snow-covered identification device for transmission line insulators according to an embodiment of the present invention.

[0015] Figure 4 is a schematic diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0016] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0017] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0018] According to an embodiment of the present invention, a method for identifying snow cover on transmission line insulators is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0019] Figure 1 is a flowchart of a method for identifying snow cover on transmission line insulators according to an embodiment of the present invention. As shown in Figure 1, the method includes the following steps:

[0020] Step S102: Obtain multiple images of insulators.

[0021] Optionally, the aforementioned multiple insulator images include snow-covered and snow-free insulator images acquired under various acquisition environments. These multiple insulator images may be, but are not limited to, acquired under high-altitude micro-meteorological conditions.

[0022] Step S104: Using image enhancement methods, image enhancement processing is performed on the above multiple insulator images to obtain multiple first insulator images.

[0023] Optionally, an image enhancement method can be set to perform image enhancement processing on multiple insulator images to obtain multiple compensated insulator images as multiple first insulator images.

[0024] In an optional embodiment, the above-mentioned image enhancement method performs image enhancement processing on the multiple insulator images to obtain multiple first insulator images, including: performing grayscale processing on the multiple first insulator images to obtain multiple third insulator images; dividing the multiple third insulator images into pixels based on the pixel intensity values ​​corresponding to each of the multiple third insulator images to obtain a first number of pixel sets corresponding to each of the multiple third insulator images, wherein the first number of pixel sets correspond to different pixel intensity ranges; and performing enhancement processing on the first number of pixel sets using different image enhancement processing methods to obtain the multiple first insulator images. Through the above method, multiple insulator images are processed in grayscale to obtain multiple third insulator images, and the pixels in the multiple third insulator images are divided into a first number of pixel sets (such as a first pixel set, a second pixel set, and a third pixel set) according to the pixel intensity values. Different image enhancement processing methods are used to enhance different pixel sets, and the compensated pixel sets are used for image reconstruction to obtain compensated multiple third insulator images as multiple first insulator images.

[0025] In an optional embodiment, where the first set of pixels includes a first set of pixels, a second set of pixels, and a third set of pixels, wherein the first set of pixels corresponds to a first pixel intensity range, the second set of pixels corresponds to a second pixel intensity range, and the third set of pixels corresponds to a third pixel intensity range, and the upper limit of the first pixel intensity range is less than or equal to the lower limit of the second pixel intensity range, and the upper limit of the second pixel intensity range is less than or equal to the lower limit of the third pixel intensity range, different image enhancement processing methods are used to enhance the first set of pixels respectively to obtain the multiple first-size pixels. An insulator image includes: increasing the pixel intensity value of each pixel in the third pixel set by a first preset pixel intensity value to obtain a fourth pixel set; increasing the pixel intensity compensation value of each pixel in the fourth pixel set by a pixel intensity value less than a second preset pixel intensity value to obtain a fifth pixel set; decreasing the pixel intensity compensation value of each pixel in the first pixel set to obtain a sixth pixel set; and decreasing the pixel intensity compensation value of each pixel in the second pixel set to obtain a seventh pixel set; and reconstructing an image based on the fifth pixel set, the sixth pixel set, and the seventh pixel set to obtain the plurality of first insulator images.

[0026] In an optional embodiment, the method further includes: obtaining the total number of pixels in the fourth pixel set and the number of pixels in the fourth pixel set that are greater than a third preset pixel intensity value; and obtaining the pixel intensity compensation value based on the total number of pixels and the first preset pixel intensity value in the following manner:

[0027]

[0028] in, The above pixel intensity compensation value is represented by N, the above number of pixels is represented by α, and the above first preset pixel intensity value is represented by N. A3 This represents the total number of pixels mentioned above.

[0029] Optionally, the pixels in multiple third insulator images are divided into three sets: a first pixel set A1, a second pixel set A2, and a third pixel set A3. The pixel intensity ranges corresponding to the first pixel set A1, the second pixel set A2, and the third pixel set A3 are as follows:

[0030]

[0031] In the formula, x i Let represent the pixel intensity value of the i-th pixel. In the third pixel set A3, a first preset intensity value α is added to each pixel, and the number N of pixels whose pixel intensity value exceeds the third preset pixel intensity value 255 is determined. Preferably, α = 25.

[0032] In the third pixel set A3, pixels with intensity values ​​less than the second preset pixel intensity value (255-α) are identified. An intensity compensation value φ is added to each pixel, resulting in a compensated set A3, which is then used as the fifth pixel set.

[0033]

[0034] In the formula, N A3 This represents the total number of pixels in the third pixel set A3.

[0035] In the first pixel set A1 and the second pixel set A2, the intensity compensation value φ is reduced for each pixel, resulting in a compensated set A1 as the sixth pixel set and a compensated set A2 as the seventh pixel set. Image reconstruction is then performed using the compensated pixel sets (i.e., the fifth pixel set, the sixth pixel set, and the seventh pixel set) corresponding to multiple third insulator images, resulting in multiple compensated third insulator images that serve as multiple first insulator images.

[0036] By employing the above methods, areas with higher pixel intensity values ​​are selected for image enhancement processing to further increase their intensity values, while reducing the intensity values ​​of background areas. This aims to highlight the target, reduce image noise, and improve the accuracy of insulator snow-covered identification in complex environments. Simultaneously, an adaptive intensity compensation value is set, which adjusts the value based on the number of pixels corresponding to high pixel intensity values ​​to prevent pixel distortion and loss of target information.

[0037] Step S106: Using a pre-constructed denoising convolutional neural network, the above-mentioned multiple first insulator images are denoised to obtain multiple second insulator images.

[0038] Optionally, the above-mentioned pre-constructed denoising convolutional neural network is used to denoise the above-mentioned multiple first insulator images to obtain multiple second insulator images, including: constructing the above-mentioned denoising convolutional neural network by combining convolution and deconvolution, upsampling and downsampling, and deep features and shallow features; and using the above-mentioned denoising convolutional neural network to perform the above-mentioned denoising processing on the above-mentioned multiple first insulator images to obtain the above-mentioned multiple second insulator images.

[0039] Optionally, as shown in Figure 2, the aforementioned denoising convolutional neural network includes a first convolutional module, a second convolutional module, a first downsampling module, a second downsampling module, a first residual module, a second residual module, a first upsampling module, a second upsampling module, a first deconvolutional module, a second deconvolutional module, a first addition module, and a second addition module. The input of the first convolutional module is used to receive the multiple first insulator images. The output of the first convolutional module is connected to the input of the second convolutional module. The output of the second convolutional module is connected to the input of the first downsampling module. The output of the first downsampling module is connected to the input of the second downsampling module. The output of the second downsampling module is connected to the input of the first residual module. The output of the first residual module is connected to... The input and output of the second residual module are connected to the input of the first upsampling module, the output of the first upsampling module is connected to the input of the second upsampling module, the output of the second upsampling module is connected to the input of the first deconvolution module, the output of the first deconvolution module is connected to the first input of the first adder module, the second input of the first adder module is connected to the output of the first convolution module, the output of the first adder module is connected to the input of the second deconvolution module, and the output of the second deconvolution module is connected to the first input of the second adder module. The second input of the second adder module is used to receive the multiple first insulator images, and the output of the second adder module is used to output the multiple second insulator images.

[0040] Using the above methods, a novel denoising convolutional neural network was constructed. By combining convolution and deconvolution, and upsampling and downsampling, the image denoising process was fitted into an image mapping process from a noisy image to a noise-free image. The network parameters were set according to the requirements of the image denoising problem and the network was trained. Noise was filtered out through feature learning, and the denoising effect was obvious, while preserving the edge and texture details of the image.

[0041] It should be noted that the above-mentioned denoising convolutional neural network first uses two convolutional modules to perform convolution processing on the compensated insulator image. The first and second convolutional modules each include 64 3×3 convolutional kernels to continuously extract target features. These features are then input into two downsampling modules, which convert the 64×256×256 feature map into a 128×64×64 feature map. The 128×64×64 feature map then passes through two residual modules to further extract target features and remove noise. Next, through two upsampling modules, the feature map is restored to a 64×256×256 feature map. Each upsampling doubles the length and width of the feature map. Finally, through two addition modules and two deconvolution modules, an approximation of the target image is generated, resulting in the denoised insulator image, which serves as the second insulator image. In the denoising convolutional neural network, convolutional and deconvolutional modules are used to extract and restore the features of the target image and remove noise. Upsampling and downsampling modules are used to reduce the feature dimension and computational cost, and then restore the image size. Residual modules and two addition modules are used to fuse deep and shallow features, which better preserves image information and reduces the loss of useful information during the denoising process. This helps the deconvolutional layer to recover a clearer insulator image. Therefore, it can improve the overall denoising performance in complex environments and improve the accuracy of subsequent insulator snow cover identification.

[0042] Step S108: Input the above multiple images of the second insulator into the pre-constructed initial insulator snow cover recognition model for training to obtain the target insulator snow cover recognition model.

[0043] Optionally, the above-mentioned initial insulator snow-covered identification model is constructed based on the single-stage target detection YOLOv5 algorithm.

[0044] It's important to note that YOLO is, overall, an end-to-end network model. From a network design perspective, it revolutionizes the region proposal box-based detection framework. R-CNN series models all require generating proposal boxes and performing classification and regression on them, but these proposal boxes overlap, leading to significant duplication of work. YOLO divides the entire image into S*S grids, with each grid responsible for detecting objects centered within that grid. It solves the problem in one shot by predicting the bounding boxes, location confidence, and probability vectors for all objects in all grids simultaneously. In other words, YOLO training and detection are performed within a single network. Furthermore, YOLO treats object detection as a regression problem; after a single inference iteration, it obtains the location, category, and corresponding confidence probability of all objects in the image. In contrast, R-CNN series models require solving the detection results in two parts: object category (classification problem) and object location (bounding box).

[0045] Step S110: The above-mentioned target insulator snow cover recognition model is used to recognize the image of the insulator to be recognized, and the snow cover recognition result corresponding to the above-mentioned insulator image to be recognized is obtained.

[0046] Optionally, an image of the insulator to be identified is acquired in real time through an image acquisition device, and the image is input into the target insulator snow cover recognition model for recognition. The snow cover status of the insulator in the image is determined based on the obtained snow cover recognition result.

[0047] Through the above steps S102 to S110, the goal of accurately identifying and detecting snow cover on insulators in complex environments can be achieved, thereby improving the accuracy of snow cover identification on transmission line insulators and timely identifying potential fault risks of transmission lines. This solves the technical problem of low identification accuracy and inability to timely identify potential fault risks of transmission lines in related technologies.

[0048] Based on the above embodiments and optional embodiments, the present invention proposes an optional implementation method, which specifically includes the following steps:

[0049] Step S1: Under high-altitude micro-meteorological conditions, acquire multiple images of insulators, including images of insulators covered in snow and images of insulators without snow.

[0050] Step S2: Set an image enhancement method to perform image enhancement processing on multiple insulator images, obtaining compensated multiple insulator images as multiple first insulator images. Specifically, this includes the following steps:

[0051] Step S21: Perform grayscale processing on multiple insulator images to obtain multiple third insulator images. Based on pixel intensity values, divide the pixels in the multiple third insulator images into three sets: a first pixel set A1, a second pixel set A2, and a third pixel set A3, where:

[0052]

[0053] In the formula, x i This represents the pixel intensity value of the i-th pixel.

[0054] Step S22: In set A3, add a first preset intensity value α to each pixel, and determine the number N of pixels whose pixel intensity value exceeds a third preset pixel intensity value of 255. Preferably, α = 25.

[0055] Step S23: In set A3, determine the pixels whose pixel intensity values ​​are less than the second preset pixel intensity value (255-α), and add an intensity compensation value φ to each pixel to obtain the compensated set A3 as the fifth pixel set, where:

[0056]

[0057] In the formula, N A3 This represents the total number of pixels in set A3.

[0058] Step S24: In sets A1 and A2, reduce the intensity compensation value φ for each pixel to obtain the compensated set A1 as the sixth pixel set and the compensated set A2 as the seventh pixel set.

[0059] Step S25: Reconstruct the image using the compensated pixel sets (i.e., the fifth pixel set, the sixth pixel set, and the seventh pixel set) corresponding to the multiple third insulator images respectively, and obtain the compensated multiple third insulator images as multiple first insulator images.

[0060] The pixel intensity values ​​of the multiple insulator images acquired using the above method are positive integers between 0 and 255. Since the target area image is typically higher than the background image (i.e., the snow-covered insulator is higher than the background image), areas with higher pixel intensity values ​​are selected for image enhancement processing to further increase their intensity value, thereby reducing the intensity value of the background image. This highlights the target, reduces image noise, and improves the accuracy of snow-covered insulator identification in complex environments. Simultaneously, an adaptive intensity compensation value is set, which adjusts the value based on the number of pixels corresponding to large pixel intensity values ​​to prevent pixel distortion and loss of target information.

[0061] Step S3: Construct a denoising convolutional neural network, denoise multiple first insulator images, and obtain multiple second insulator images.

[0062] It should be noted that under high-altitude micro-meteorological conditions, the environment is extremely complex. During the acquisition of insulator images, they will inevitably be significantly affected by noise, leading to a decrease in image quality, the addition of noise, and an impact on the accuracy of subsequent recognition. Therefore, this embodiment of the invention constructs a novel denoising convolutional neural network. By combining convolution and deconvolution, and upsampling and downsampling, the image denoising process is fitted into an image mapping process from a noisy image to a noise-free image. The network parameters are set according to the requirements of the image denoising problem, and the network is trained. Noise is filtered out through feature learning, resulting in a significant denoising effect while preserving the edge and texture details of the image. Step S3 specifically includes the following steps:

[0063] Step S31: Construct a denoising convolutional neural network by combining convolution and deconvolution, upsampling and downsampling, and deep features and shallow features.

[0064] As shown in Figure 2, the denoising convolutional neural network includes a first convolution module, a second convolution module, a first downsampling module, a second downsampling module, a first residual module, a second residual module, a first upsampling module, a second upsampling module, a first deconvolution module, a second deconvolution module, a first addition module, and a second addition module. The input of the first convolution module receives the compensated insulator image. The output of the first convolution module is connected to the input of the second convolution module. The output of the second convolution module is connected to the input of the first downsampling module. The output of the first downsampling module is connected to the input of the second downsampling module. The output of the second downsampling module is connected to the input of the first residual module. The output of the first residual module is connected to the input of the second residual module. The output of the second residual module is connected to the input of the first upsampling module. The output of the first upsampling module is connected to the input of the second upsampling module. The output of the second upsampling module is connected to the input of the first deconvolution module. The output of the first deconvolution module is connected to the first input of the first adder module. The second input of the first adder module is connected to the output of the first convolution module. The output of the first adder module is connected to the input of the second deconvolution module. The output of the second deconvolution module is connected to the first input of the second adder module. The second input of the second adder module receives the compensated insulator image. The output of the second adder module outputs the denoised insulator image.

[0065] The denoising convolutional neural network first uses two convolutional modules to process the compensated insulator image. The first and second convolutional modules each contain 64 3×3 convolutional kernels to continuously extract target features. These features are then input into two downsampling modules, transforming the 64×256×256 feature map into a 128×64×64 feature map. The 128×64×64 feature map is then processed by two residual modules to extract target features in depth and remove noise. Next, two upsampling modules restore the feature map to a 64×256×256 feature map. Each upsampling doubles the length and width of the feature map. Finally, two addition modules and two deconvolution modules generate an approximation of the target image, resulting in the denoised insulator image.

[0066] Step S32: Input the compensated insulator image into the denoising convolutional neural network for denoising processing to obtain the denoised insulator image.

[0067] In the denoising convolutional neural network, convolutional and deconvolutional modules are used to extract and restore the features of the target image and remove noise. Upsampling and downsampling modules are used to reduce the feature dimension and computational cost, and then restore the image size. Residual modules and two addition modules are used to fuse deep and shallow features, which better preserves image information and reduces the loss of useful information during the denoising process. This helps the deconvolutional layer to recover a clearer insulator image. Therefore, it can improve the overall denoising performance in complex environments and improve the accuracy of subsequent insulator snow cover identification.

[0068] Step S4: Input multiple images of the second insulator into the initial insulator snow cover recognition model pre-built using the YOLOv5 algorithm for training, and obtain the target insulator snow cover recognition model.

[0069] Step S5: Obtain an image of the insulator to be identified in real time through the image acquisition device, input the image of the insulator to be identified into the target insulator snow cover recognition model for recognition, and determine the snow cover status of the insulator in the image of the insulator to be identified based on the obtained snow cover recognition result.

[0070] The embodiments of the present invention can achieve at least the following technical effects: (1) Constructing an image enhancement method and a denoising convolutional neural network to perform image enhancement and denoising processing on insulator images, effectively improving the recognition accuracy of insulator snow cover. (2) Adaptively adjusting the intensity compensation value according to the number of pixels corresponding to large pixel intensity values, enhancing the area with larger pixel intensity values, reducing the intensity value of the background image, highlighting the target, reducing image noise, and improving the recognition accuracy of insulator snow cover in subsequent complex environments. (3) Using convolution and deconvolution modules to extract and restore target image features and remove noisy images; using upsampling and downsampling modules to restore the image size after reducing the feature dimension and computational load; using residual modules and two addition modules to fuse deep features and shallow features, better preserving image information during denoising, reducing the loss of useful information, and helping the deconvolution layer to recover a clearer insulator image, improving the overall denoising performance in complex environments, and improving the recognition accuracy of insulator snow cover in subsequent environments.

[0071] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0072] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0073] This embodiment also provides a snow-covered identification device for transmission line insulators. This device is used to implement the above embodiments and preferred embodiments, and will not be repeated as already described. As used below, the terms "module" and "device" can refer to a combination of software and / or hardware that performs a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0074] According to an embodiment of the present invention, an apparatus embodiment for implementing the above-described method for identifying snow-covered insulators of transmission lines is also provided. Figure 3 is a schematic structural diagram of an apparatus for identifying snow-covered insulators of transmission lines according to an embodiment of the present invention. As shown in Figure 3, the above-described apparatus for identifying snow-covered insulators of transmission lines includes: an acquisition module 300, a first image processing module 302, a second image processing module 304, a training module 306, and an image recognition module 308, wherein:

[0075] The aforementioned acquisition module 300 is used to acquire multiple insulator images, including snow-covered and snow-free insulator images acquired under various acquisition environments.

[0076] The first image processing module 302 is connected to the acquisition module 300 and is used to perform image enhancement processing on the multiple insulator images using an image enhancement method to obtain multiple first insulator images.

[0077] The second image processing module 304 is connected to the first image processing module 302 and is used to perform denoising processing on the multiple first insulator images using a pre-built denoising convolutional neural network to obtain multiple second insulator images.

[0078] The training module 306 is connected to the second image processing module 304 and is used to input the multiple second insulator images into the pre-constructed initial insulator snow cover recognition model for training, so as to obtain the target insulator snow cover recognition model.

[0079] The image recognition module 308 is connected to the training module 306 and is used to recognize the image of the insulator to be recognized using the target insulator snow cover recognition model, so as to obtain the snow cover recognition result corresponding to the image of the insulator to be recognized.

[0080] In this embodiment of the invention, the acquisition module 300 is configured to acquire multiple insulator images, including snow-covered and snow-free images of insulators acquired under various acquisition environments. The first image processing module 302, connected to the acquisition module 300, is used to perform image enhancement processing on the multiple insulator images using an image enhancement method to obtain multiple first insulator images. The second image processing module 304, connected to the first image processing module 302, is used to perform denoising processing on the multiple first insulator images using a pre-constructed denoising convolutional neural network to obtain multiple second insulator images. The training module 306 is connected to the second image processing module 304. The training module 306 is used to input the aforementioned multiple images of the second insulator into a pre-constructed initial insulator snow cover recognition model for training, thereby obtaining a target insulator snow cover recognition model. The image recognition module 308, connected to the training module 306, is used to recognize the insulator image to be recognized using the target insulator snow cover recognition model, thereby obtaining the snow cover recognition result corresponding to the insulator image to be recognized. This achieves the goal of accurately identifying and detecting the snow cover situation of insulators in complex environments, thus realizing the technical effect of improving the accuracy of snow cover recognition of transmission line insulators and timely identifying potential fault risks of transmission lines. This solves the technical problem of low recognition accuracy and inability to timely identify potential fault risks of transmission lines in related technologies.

[0081] It should be noted that the above modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0082] It should be noted that the acquisition module 300, the first image processing module 302, the second image processing module 304, the training module 306, and the image recognition module 308 mentioned above correspond to steps S102 to S110 in the embodiments. The instances and application scenarios implemented by the above modules and their corresponding steps are the same, but they are not limited to the content disclosed in the above embodiments. It should be noted that the above modules, as part of the device, can run on a computer terminal.

[0083] It should be noted that the optional or preferred implementation methods of this embodiment can be found in the relevant descriptions in the embodiments, and will not be repeated here.

[0084] The aforementioned snow-covered insulator identification device for transmission lines may also include a processor and a memory. The aforementioned acquisition module 300, first image processing module 302, second image processing module 304, training module 306, image recognition module 308, etc., are all stored in the memory as program modules, and the processor executes the aforementioned program modules stored in the memory to realize the corresponding functions.

[0085] The processor contains a core that retrieves the corresponding program modules from memory. One or more cores may be configured. Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory includes at least one memory chip.

[0086] According to an embodiment of this application, an embodiment of a non-volatile storage medium is also provided. Optionally, in this embodiment, the non-volatile storage medium includes a stored program, wherein, when the program is running, it controls the device containing the non-volatile storage medium to execute any of the above-mentioned transmission line insulator snow cover identification methods.

[0087] Optionally, in this embodiment, the non-volatile storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals, and the non-volatile storage medium includes stored programs.

[0088] Optionally, during program execution, the device containing the non-volatile storage medium performs the following functions: acquiring multiple insulator images, including snow-covered and snow-free images of insulators acquired under various acquisition environments; performing image enhancement processing on the multiple insulator images using an image enhancement method to obtain multiple first insulator images; performing denoising processing on the multiple first insulator images using a pre-constructed denoising convolutional neural network to obtain multiple second insulator images; inputting the multiple second insulator images into a pre-constructed initial insulator snow-covered recognition model for training to obtain a target insulator snow-covered recognition model; and using the target insulator snow-covered recognition model to recognize the insulator image to be recognized to obtain the snow-covered recognition result corresponding to the insulator image to be recognized.

[0089] According to an embodiment of this application, an embodiment of a processor is also provided. Optionally, in this embodiment, the processor is used to run a program, wherein the program executes any of the above-described methods for identifying snow cover on transmission line insulators.

[0090] According to an embodiment of this application, an embodiment of a computer program product is also provided, which, when executed on a data processing device, is adapted to execute a program that initializes the snow cover identification method steps for transmission line insulators having any of the above-described steps.

[0091] Optionally, when the aforementioned computer program product is executed on a data processing device, it is suitable to execute an initialization program with the following steps: acquiring multiple insulator images, wherein the multiple insulator images include snow-covered insulator images and snow-free insulator images acquired under various acquisition environments; performing image enhancement processing on the multiple insulator images using an image enhancement method to obtain multiple first insulator images; performing denoising processing on the multiple first insulator images using a pre-constructed denoising convolutional neural network to obtain multiple second insulator images; inputting the multiple second insulator images into a pre-constructed initial insulator snow-covered recognition model for training to obtain a target insulator snow-covered recognition model; and using the target insulator snow-covered recognition model to recognize the insulator image to be recognized to obtain the snow-covered recognition result corresponding to the insulator image to be recognized.

[0092] According to an embodiment of this application, an embodiment of a processor is also provided. Optionally, in this embodiment, the processor is used to run a program, wherein the program executes any of the above-described methods for identifying snow cover on transmission line insulators.

[0093] As shown in Figure 4, this embodiment of the invention provides an electronic device 10, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it performs the following steps: acquiring multiple insulator images, wherein the multiple insulator images include snow-covered insulator images and snow-free insulator images acquired under various acquisition environments; performing image enhancement processing on the multiple insulator images using an image enhancement method to obtain multiple first insulator images; performing denoising processing on the multiple first insulator images using a pre-constructed denoising convolutional neural network to obtain multiple second insulator images; inputting the multiple second insulator images into a pre-constructed initial insulator snow-covered recognition model for training to obtain a target insulator snow-covered recognition model; and using the target insulator snow-covered recognition model to recognize the insulator image to be recognized to obtain the snow-covered recognition result corresponding to the insulator image to be recognized.

[0094] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0095] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0096] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of modules described above can be a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between modules, and may be electrical or other forms.

[0097] The modules described above as separate components may or may not be physically separate. Similarly, the components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple modules. Some or all of the modules can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0098] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0099] If the aforementioned integrated modules are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable non-volatile storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a non-volatile storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned non-volatile storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0100] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for identifying snow cover on transmission line insulators, characterized in that, include: Multiple insulator images are acquired, including snow-covered and snow-free images of insulators acquired under various acquisition environments. Image enhancement methods are used to enhance each of the multiple insulator images, resulting in multiple first insulator images. A pre-constructed denoising convolutional neural network is used to denoise each of the multiple first insulator images, resulting in multiple second insulator images. These second insulator images are then input into a pre-constructed initial insulator snow-covered recognition model for training, resulting in a target insulator snow-covered recognition model. The target insulator snow-covered recognition model is then used... A snow recognition model identifies the insulator image to be identified, obtaining the snow cover recognition result corresponding to the insulator image. Specifically, an image enhancement method is used to perform image enhancement processing on the multiple insulator images to obtain multiple first insulator images, including: performing grayscale processing on the multiple insulator images to obtain multiple third insulator images; and dividing the multiple third insulator images into pixels based on the pixel intensity values ​​corresponding to each of the multiple third insulator images to obtain a first number of pixel sets corresponding to each of the multiple third insulator images, wherein the first number of pixel sets... Corresponding to different pixel intensity ranges, the first set of pixels includes a first set of pixels, a second set of pixels, and a third set of pixels. The first set of pixels corresponds to a first pixel intensity range, the second set of pixels corresponds to a second pixel intensity range, and the third set of pixels corresponds to a third pixel intensity range. The upper limit of the first pixel intensity range is less than or equal to the lower limit of the second pixel intensity range, and the upper limit of the second pixel intensity range is less than or equal to the lower limit of the third pixel intensity range. A first preset pixel intensity value is added to each pixel in the third set to obtain a fourth set of pixels. A pixel intensity compensation value is added to each pixel in the fourth set that is less than a second preset pixel intensity value to obtain a fifth set of pixels. The pixel intensity compensation value is decreased for each pixel in the first set to obtain a sixth set of pixels. The pixel intensity compensation value is decreased for each pixel in the second set to obtain a seventh set of pixels. Image reconstruction is performed based on the fifth set of pixels, the sixth set of pixels, and the seventh set of pixels to obtain the plurality of first insulator images. The pixel intensity compensation value is obtained in the following manner: ;in, This represents the pixel intensity compensation value, where N represents the number of pixels in the third pixel set that exceed a third preset pixel intensity value, and the third preset pixel intensity value is 255. This represents the first preset pixel intensity value. This indicates the total number of pixels included in the third pixel set, and the second preset pixel intensity value is 255- 。 2. The method according to claim 1, characterized in that, The method involves using a pre-constructed denoising convolutional neural network to denoise the multiple first insulator images to obtain multiple second insulator images. This includes: constructing the denoising convolutional neural network using a combination of convolution and deconvolution, upsampling and downsampling, and deep features and shallow features; and using the denoising convolutional neural network to denoise the multiple first insulator images to obtain the multiple second insulator images.

3. The method according to claim 2, characterized in that, The denoising convolutional neural network includes a first convolution module, a second convolution module, a first downsampling module, a second downsampling module, a first residual module, a second residual module, a first upsampling module, a second upsampling module, a first deconvolution module, a second deconvolution module, a first addition module, and a second addition module. The input of the first convolution module receives the multiple first insulator images. The output of the first convolution module is connected to the input of the second convolution module. The output of the second convolution module is connected to the input of the first downsampling module. The output of the first downsampling module is connected to the input of the second downsampling module. The output of the second downsampling module is connected to the input of the first residual module. The output of the first residual module is connected to the input of the second residual module. The input terminal of the first upsampling module is connected to the input terminal of the second residual module. The output terminal of the second residual module is connected to the input terminal of the first upsampling module. The output terminal of the first upsampling module is connected to the input terminal of the second upsampling module. The output terminal of the second upsampling module is connected to the input terminal of the first deconvolution module. The output terminal of the first deconvolution module is connected to the first input terminal of the first addition module. The second input terminal of the second addition module is used to receive the multiple first insulator images. The output terminal of the second addition module is used to output the multiple second insulator images.

4. The method according to any one of claims 1 to 3, characterized in that, The method further includes: the initial insulator snow-covered identification model is constructed based on the single-stage target detection YOLOv5 algorithm.

5. A snow-covered identification device for transmission line insulators, characterized in that, include: The system comprises: an acquisition module for acquiring multiple insulator images, including snow-covered and snow-free images of insulators acquired under various acquisition environments; a first image processing module for performing image enhancement processing on the multiple insulator images using image enhancement methods to obtain multiple first insulator images; a second image processing module for performing denoising processing on the multiple first insulator images using a pre-constructed denoising convolutional neural network to obtain multiple second insulator images; and a training module for inputting the multiple second insulator images into a pre-constructed initial insulator snow-covered image. The recognition model is trained to obtain a target insulator snow cover recognition model; the image recognition module is used to recognize the image of the insulator to be recognized using the target insulator snow cover recognition model to obtain the snow cover recognition result corresponding to the image of the insulator to be recognized; wherein, the first image processing module is further used to perform grayscale processing on the multiple insulator images to obtain multiple third insulator images; based on the pixel intensity values ​​corresponding to the multiple third insulator images respectively, the multiple third insulator images are divided into pixels to obtain a first number of pixel sets corresponding to the multiple third insulator images respectively, wherein the first number The first set of pixels corresponds to different pixel intensity ranges. This first set of pixels includes a first set, a second set, and a third set. The first set corresponds to a first pixel intensity range, the second set corresponds to a second pixel intensity range, and the third set corresponds to a third pixel intensity range. The upper limit of the first pixel intensity range is less than or equal to the lower limit of the second pixel intensity range, and the upper limit of the second pixel intensity range is less than or equal to the lower limit of the third pixel intensity range. A first preset pixel intensity value is added to each pixel in the third set to obtain a fourth set of pixels. A pixel intensity compensation value is added to each pixel in the fourth set that has an intensity less than a second preset pixel intensity value to obtain a fifth set of pixels. The pixel intensity compensation value is decreased for each pixel in the first set to obtain a sixth set of pixels. The pixel intensity compensation value is decreased for each pixel in the second set to obtain a seventh set of pixels. Image reconstruction is performed based on the fifth set, the sixth set, and the seventh set of pixels to obtain the multiple first insulator images. The pixel intensity compensation value is obtained in the following manner: ;in, This represents the pixel intensity compensation value, where N represents the number of pixels in the third pixel set that exceed a third preset pixel intensity value, and the third preset pixel intensity value is 255. This represents the first preset pixel intensity value. This indicates the total number of pixels included in the third pixel set, and the second preset pixel intensity value is 255- 。 6. A non-volatile storage medium, characterized in that, The non-volatile storage medium stores multiple instructions, which are adapted to be loaded by a processor and executed by the snow-covered identification method for transmission line insulators according to any one of claims 1 to 4.

7. An electronic device, characterized in that, The method includes one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method for identifying snow cover on transmission line insulators as described in any one of claims 1 to 4.

Images