Power transmission line icing analysis method and system based on image recognition

Abstract

The invention provides a power transmission line icing analysis method and system based on image recognition, and relates to the technical field of image processing. The method comprises the following steps: acquiring a historical image set about power transmission line icing detection, determining a target icing area of each line icing image, extracting icing visual features of each target icing area, constructing an icing data set, and training to obtain an icing analysis model; performing icing feature clustering on the plurality of line icing images to generate a plurality of icing feature class clusters; the method comprises the following steps: after a to-be-analyzed icing image of a power transmission line is collected, extracting a target icing feature of the to-be-analyzed icing image, generating a first icing analysis result according to a plurality of icing feature clusters, and processing the target icing feature through an icing analysis model to generate a second icing analysis result, and fusing the first icing analysis result and the second icing analysis result to obtain a target icing analysis result. According to the invention, the precision and reliability of icing detection are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a method and system for analyzing ice coating on a power transmission line based on image recognition. Background Art

[0002] Detection of ice on transmission lines is an important part of ensuring the safe operation of power systems. Traditional ice detection methods rely on sensors, radars, infrared thermal imaging and other equipment, combined with meteorological data and sensor information for analysis. These methods usually require high costs, and in actual applications may be affected by sensor failures, signal interference and climatic conditions, resulting in poor adaptability and inability to complete large-scale ice monitoring in a timely and accurate manner.

[0003] With the development of computer vision technology, the potential of image analysis in various industrial applications has been gradually explored. In the field of ice detection on transmission lines, the relevant images collected by camera equipment can reflect the characteristics of the ice layer from multiple dimensions. The information such as surface texture and brightness difference can indirectly convey the characteristics of ice thickness, adhesion and type. In-depth analysis of ice images can help analyze the characteristics of ice thickness and type, and provide a basis for comprehensive evaluation of ice. Therefore, in the field of ice image analysis of transmission lines, how to efficiently mine the surface features of images such as texture and brightness to model and analyze adhesion, thickness and other information is the key to improving the accuracy of ice analysis, which is of great significance to improving the accuracy and reliability of ice detection. It is urgent to propose a comprehensive technical solution for analyzing ice images to improve the accuracy and reliability of ice detection analysis. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes a method and system for analyzing ice coating on power transmission lines based on image recognition, which combines visual feature analysis and cluster analysis of images to optimize the analysis process of ice coating type and thickness, and can better identify the type and thickness of ice coating, thereby improving the accuracy and reliability of ice coating detection.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows: A first aspect of the present invention provides a method for analyzing ice coating on a power transmission line based on image recognition, comprising: Acquire a historical image set about ice detection of power transmission lines collected by a camera device, wherein the historical image set includes a plurality of line ice images and ice annotation data of each line ice image; Determine the target ice-covered area of ​​each line ice-covered image according to the ice-covered annotation data, perform local brightness difference analysis on multiple target ice-covered areas to extract thickness features of each target ice-covered area, and perform local texture analysis on multiple target ice-covered areas to extract texture features of each target ice-covered area; The thickness features and texture features of the target ice-covered area are spliced ​​to generate ice-covered visual features of each target ice-covered area. An ice-covered dataset is constructed based on multiple ice-covered visual features and ice-covered annotation data. An ice-covered analysis model is obtained through training with the ice-covered dataset. Clustering the ice-covered images of multiple lines according to the multiple ice-covered visual features to generate multiple ice-covered feature clusters, including determining multiple initial cluster centers based on ice-covered annotation data, using a K-means clustering algorithm based on cluster center restriction optimization, and clustering the multiple ice-covered images of multiple lines according to the multiple ice-covered visual features to obtain multiple ice-covered feature clusters; After collecting the ice-covered image to be analyzed of the transmission line, the target ice-covered features of the ice-covered image to be analyzed are extracted, and a first ice-covered analysis result of the ice-covered image to be analyzed is determined according to multiple ice-covered feature clusters. The target ice-covered features are processed by an ice-covered analysis model to generate a second ice-covered analysis result. The first ice-covered analysis result and the second ice-covered analysis result are fused to obtain the target ice-covered analysis result of the ice-covered image to be analyzed.

[0006] Preferably, performing local brightness difference analysis on multiple target ice-covered areas to extract thickness features of each target ice-covered area, and performing local texture analysis on multiple target ice-covered areas to extract texture features of each target ice-covered area include: The grayscale image corresponding to each target ice-covered area is extracted, and sliding window analysis is performed on multiple grayscale images to extract the local grayscale mean and local contrast of each window area. The global grayscale mean and global contrast of the target ice-covered area are calculated based on multiple local grayscale means and local contrasts, and the global grayscale standard deviation of the target ice-covered area is calculated based on the global grayscale mean. Based on the edge detection operator, a gradient analysis is performed on multiple grayscale images to obtain a gradient image of each grayscale image. Multiple local gradient peaks are extracted from the gradient image through sliding window analysis, and the global gradient peak of each target ice-covered area is calculated to obtain the thickness characteristics of each target ice-covered area. Sliding window analysis is performed on multiple grayscale images respectively, and each window area is processed by Gabor filter to obtain multiple Gabor filter responses of each window area. Multiple local texture feature parameters of each window area are extracted according to the multiple Gabor filter responses. Multiple global texture feature parameters of each grayscale image are generated according to the multiple local texture features to obtain the texture features of each target ice-covered area.

[0007] Preferably, performing ice-covered feature clustering on a plurality of line ice-covered images according to a plurality of ice-covered visual features to generate a plurality of ice-covered feature clusters includes: Determine the sample ice thickness range according to the ice annotation data in the historical image set, divide the sample ice thickness range into multiple thickness intervals based on the preset thickness interval, and pre-classify multiple line ice images according to the ice annotation data to obtain a sub-sample set for each thickness interval; An initial cluster center of each thickness interval is determined according to multiple sub-sample sets, and a K-means clustering algorithm is used to cluster ice features of multiple line ice images based on a preset thickness range according to the multiple initial cluster centers, including in each iteration process, if the cluster centers corresponding to multiple samples assigned to the current ice feature cluster calculated deviate from the preset thickness range of the thickness interval to which the ice feature cluster belongs, then according to the preset thickness range of the thickness interval to which the ice feature cluster belongs, the multiple samples assigned to the current ice feature cluster are screened, and multiple target samples whose ice thickness belongs to the preset thickness range among the multiple samples are determined, and the cluster center of the ice feature cluster is determined according to the multiple target samples. After the preset cluster center fluctuation condition is met, the clustering of the multiple line ice images is completed to obtain multiple ice feature clusters.

[0008] Preferably, an icing dataset is constructed according to a plurality of icing visual features and icing annotation data, and an icing analysis model is obtained by training the icing dataset, including: The ice thickness label and ice type label of each line ice image are extracted from the ice annotation data, and an ice dataset is constructed based on multiple ice visual features. The corresponding ice thickness label and ice type label are added to each line ice image in the ice dataset, and the ice analysis model is trained through the ice dataset. Among them, the icing analysis model is a multi-task deep neural network, including an input layer, a shared feature extraction layer, an ice thickness regression branch and an ice type classification branch. The total loss function of the icing analysis model is constructed according to the loss functions corresponding to the ice thickness regression branch and the ice type classification branch respectively. During the model training process, multiple icing visual features in the icing dataset are used as input, the ice thickness label corresponding to the line icing image is used as the training target of the ice thickness regression branch, and the ice type label corresponding to the line icing image is used as the training target of the icing type classification branch. The icing analysis model is obtained by training the icing dataset.

[0009] Preferably, determining a first ice-covering analysis result of the ice-covering image to be analyzed according to a plurality of ice-covering feature clusters includes: Determine the thickness cluster center of each ice feature cluster, match the target ice feature of the ice image to be analyzed with multiple thickness cluster centers, determine the thickness cluster center closest to the target ice feature, and generate a first ice thickness analysis result of the ice image to be analyzed; Multiple line icing images in the icing feature cluster are classified based on icing type to obtain multiple sub-icing type clusters, the type cluster center of each sub-icing type cluster is determined, the target icing feature of the icing image to be analyzed is matched with the multiple type cluster centers, the type cluster center closest to the target icing feature is determined, and a first icing type analysis result of the icing image to be analyzed is generated; and a first icing analysis result of the icing image to be analyzed is determined according to the first icing thickness analysis result and the first icing type analysis result.

[0010] Preferably, the first ice covering analysis result and the second ice covering analysis result are fused to obtain a target ice covering analysis result of the ice covering image to be analyzed, including: Determine a second ice thickness analysis result and a second ice type analysis result generated by the ice analysis model in the second ice analysis result, and match the first ice thickness analysis result with the second ice thickness analysis result; If the first ice thickness analysis result matches the second ice thickness analysis result successfully, the second ice thickness analysis result is used as the target ice thickness analysis result of the ice image to be analyzed; Otherwise, a first reference cluster is determined according to the type cluster center closest to the target icing feature, a first similarity between the target icing feature and the first reference cluster is calculated, a second reference cluster associated with the second icing analysis result is determined, and a second similarity between the target icing feature and the second reference cluster is calculated; The third similarity between the target icing feature and the most matching sub-icing type cluster in the first reference cluster is calculated. After correcting the first similarity based on the third similarity, the icing analysis result corresponding to the maximum value between the corrected first similarity and the second similarity is selected as the target icing analysis result of the icing image to be analyzed.

[0011] A second aspect of the present invention provides a power transmission line icing analysis system based on image recognition, which is used to implement the above-mentioned power transmission line icing analysis method based on image recognition, comprising: An ice data acquisition module is used to acquire a historical image set about ice detection of power transmission lines collected by a camera device, wherein the historical image set includes a plurality of line ice images and ice annotation data of each line ice image; An ice feature extraction module is used to determine the target ice area of ​​each line ice image according to the ice annotation data, perform local brightness difference analysis on multiple target ice areas to extract the thickness feature of each target ice area, and perform local texture analysis on multiple target ice areas to extract the texture feature of each target ice area; The icing analysis model training module is used to generate icing visual features of each target icing area by combining the thickness features and texture features of the target icing area, construct an icing dataset based on multiple icing visual features and icing annotation data, and obtain an icing analysis model through icing dataset training; An ice feature clustering analysis module is used to cluster ice features of multiple line ice images according to multiple ice visual features to generate multiple ice feature clusters, including determining multiple initial cluster centers based on ice annotation data, using a K-means clustering algorithm based on cluster center restriction optimization, and clustering multiple line ice images based on multiple ice visual features to obtain multiple ice feature clusters; The icing feature analysis module is used to extract the target icing features of the icing image to be analyzed after collecting the icing image to be analyzed of the transmission line, determine the first icing analysis result of the icing image to be analyzed according to multiple icing feature clusters, process the target icing features through the icing analysis model to generate a second icing analysis result, and fuse the first icing analysis result and the second icing analysis result to obtain the target icing analysis result of the icing image to be analyzed.

[0012] Preferably, the ice feature clustering analysis module performs ice feature clustering on multiple line ice images according to multiple ice visual features to generate multiple ice feature clusters, including: Determine the sample ice thickness range according to the ice annotation data in the historical image set, divide the sample ice thickness range into multiple thickness intervals based on the preset thickness interval, and pre-classify multiple line ice images according to the ice annotation data to obtain a sub-sample set for each thickness interval; An initial cluster center of each thickness interval is determined according to multiple sub-sample sets, and a K-means clustering algorithm is used to cluster ice features of multiple line ice images based on a preset thickness range according to the multiple initial cluster centers, including in each iteration process, if the cluster centers corresponding to multiple samples assigned to the current ice feature cluster calculated deviate from the preset thickness range of the thickness interval to which the ice feature cluster belongs, then according to the preset thickness range of the thickness interval to which the ice feature cluster belongs, the multiple samples assigned to the current ice feature cluster are screened, and multiple target samples whose ice thickness belongs to the preset thickness range among the multiple samples are determined, and the cluster center of the ice feature cluster is determined according to the multiple target samples. After the preset cluster center fluctuation condition is met, the clustering of the multiple line ice images is completed to obtain multiple ice feature clusters.

[0013] The present invention has the following beneficial effects: The present invention analyzes the historical image set of transmission line icing detection, extracts the global thickness characteristics and texture characteristics of ice images of different lines to obtain the icing visual characteristics of the images, characterizes the different ice layer characteristics of icing, clusters multiple icing images based on icing thickness based on the icing visual characteristics, analyzes the differences in visual characteristics of ice layer characteristics under different icing thicknesses with icing thickness as the core, assists in realizing preliminary icing characteristic analysis of the images to be analyzed through multiple icing feature clusters obtained by clustering, constructs a data set based on multiple icing visual features and trains to obtain an icing analysis model for further analysis of icing characteristics, realizes icing analysis of transmission lines from the perspective of machine learning, and finally organically integrates the icing characteristic analysis results generated based on image optical group characteristics and machine learning models, which can more accurately identify the thickness and type of icing images, and provide a solid data foundation for the analysis and monitoring of ice layers. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 The present invention is a flowchart of a method for analyzing ice coating on a power transmission line based on image recognition provided in one embodiment of the present invention.

[0015] Figure 2 The present invention is a schematic structural diagram of a power transmission line icing analysis system based on image recognition provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0016] In order to enable those skilled in the art to better understand the technical solutions in 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. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. It should also be noted that, for ease of description, only the parts related to the present application are shown in the accompanying drawings, not all.

[0017] See also Figure 1 , which shows a schematic flow chart of a method for analyzing ice coating on a power transmission line based on image recognition provided in one embodiment of the present invention, the method specifically comprises the following steps: Step S10: Acquire a historical image set collected by a camera device regarding ice detection of power transmission lines, wherein the historical image set includes a plurality of line ice images and ice annotation data of each line ice image.

[0018] In this embodiment, the historical image set can be a historical data set about the transmission line collected by relevant camera equipment. At present, some ice monitoring devices that can be installed on the transmission line can remotely monitor the transmission line in a complex environment, collect images through high-definition cameras, and can also collect line tension data based on other sensor equipment to conduct a comprehensive analysis of the ice condition of the transmission line. The historical image set contains rich feature information under different ice conditions. Combined with relevant ice annotation data, such as manually annotated information about the thickness and type of ice, it can provide a large amount of basic data for ice detection and analysis based on image recognition.

[0019] Step S20, determining the target ice-covered area of ​​each line ice-covered image according to the ice-covered annotation data, performing local brightness difference analysis on multiple target ice-covered areas to extract thickness characteristics of each target ice-covered area, and performing local texture analysis on multiple target ice-covered areas to extract texture characteristics of each target ice-covered area.

[0020] In this embodiment, the target ice-covered area in each image is first determined by analyzing the ice-covered annotation data in the image. For a historical image set containing multiple line ice-covered images, the target ice-covered area can be obtained by manual annotation during the manual analysis process, or can be automatically analyzed using relevant image segmentation techniques, such as threshold segmentation, edge detection, etc., to determine the area actually covered by ice in the image.

[0021] After determining the target ice-covered area of ​​each line ice-covered image, a more detailed analysis of each area is required, including local brightness difference analysis. By comparing and analyzing the brightness change characteristics of different parts in the target area, the thickness characteristics of the ice can be indirectly characterized. Thicker ice layers usually lead to more obvious brightness differences, while thinner ice layers may show more uniform brightness. Therefore, the thickness characteristics of each target ice-covered area can be extracted through local brightness difference analysis.

[0022] In addition, local texture analysis is also performed. By analyzing the changes in the texture of the ice surface, texture features are further extracted. Different types of ice, such as rime, rime, and rime, will show different texture features in the image. Using image processing methods such as Gabor filtering and gray-level co-occurrence matrix (GLCM), information reflecting the surface structure of the ice layer can be effectively extracted to obtain the texture features of each target ice-covered area. This can be used for subsequent in-depth analysis of the ice-covered type.

[0023] Step S30, feature stitching the thickness features and texture features of the target ice-covered area to generate ice-covered visual features of each target ice-covered area, construct an ice-covered dataset based on multiple ice-covered visual features and ice-covered annotation data, and obtain an ice-covered analysis model through training the ice-covered dataset.

[0024] In this embodiment, the thickness features and texture features of the target ice-covered area extracted in step S20 are spliced ​​to form an ice-covered visual feature of each target ice-covered area, which contains multiple attributes of the target area, such as thickness, texture, brightness and other information. These spliced ​​features will be used to construct a subsequent ice-covered dataset. Based on the ice-covered visual features and annotation data in the historical image set, a standardized ice-covered dataset is constructed, in which each sample includes the visual features of the ice-covered area and the corresponding annotation data such as ice thickness, type, etc. An ice-covered analysis model capable of processing and analyzing ice-covered images is obtained through training with this dataset.

[0025] Step S40, clustering the ice-covered images of the multiple lines according to the multiple ice-covered visual features to generate multiple ice-covered feature clusters, including determining multiple initial cluster centers based on the ice-covered annotation data, adopting a K-means clustering algorithm based on cluster center restriction optimization, and clustering the multiple ice-covered images of the multiple lines according to the multiple ice-covered visual features to obtain multiple ice-covered feature clusters.

[0026] In this embodiment, the K-means clustering algorithm is used to cluster the ice visual features of multiple line ice images, and the restricted optimization of the cluster center is combined in the process to improve the clustering effect. The purpose of clustering is to divide the image into multiple different clusters according to the ice visual features in the image, and each cluster represents a class of similar ice features. In order to make the clustering more accurate, it is first necessary to determine the initial cluster center of the cluster. These cluster centers can be initialized by ice annotation data. For example, based on different ice thickness ranges or ice types, multiple initial cluster centers can be determined to guide the clustering process. During the clustering iteration process, each image will be assigned to the cluster center with the smallest distance according to its visual features, and the position of the cluster center will be updated. Through multiple iterations, multiple line ice images can eventually be divided into multiple ice feature clusters.

[0027] Step S50: after collecting the ice-covered image to be analyzed of the transmission line, extracting the target ice-covered features of the ice-covered image to be analyzed, determining a first ice-covered analysis result of the ice-covered image to be analyzed according to a plurality of ice-covered feature clusters, processing the target ice-covered features through an ice-covered analysis model to generate a second ice-covered analysis result, and fusing the first ice-covered analysis result and the second ice-covered analysis result to obtain the target ice-covered analysis result of the ice-covered image to be analyzed.

[0028] In this embodiment, when a new transmission line ice-covered image is collected, the target ice-covered features of the ice-covered image to be analyzed are first extracted. Specifically, the same image analysis method as in step S20 can be used to extract the thickness features and texture features of the ice-covered parts in the image to be analyzed to obtain the target ice-covered features of the image. Then, based on the multiple specification ice-covered feature clusters obtained by clustering in the above steps, the first ice-covered analysis result of the image to be analyzed is determined by calculating the distance between the target ice-covered features of the image to be analyzed and the center of each cluster, so as to preliminarily determine the ice-covered thickness range and type corresponding to the image to be analyzed. For the trained ice-covered analysis model, the target ice-covered features are input into the model for analysis, and the model generates a second ice-covered analysis result of information such as ice-covered thickness and type of the ice-covered image to be analyzed. The first ice-covered analysis result obtained by clustering is fused with the second ice-covered analysis result generated by the ice-covered analysis model to obtain a target ice-covered analysis result that can more accurately characterize the ice-covered feature related information of the ice-covered image to be analyzed, so as to provide a comprehensive and accurate analysis basis for ice-covered detection of transmission lines.

[0029] In one implementation process, for step S20, in order to comprehensively analyze multiple target ice-covered areas, the thickness characteristics and texture characteristics of each target ice-covered area are extracted. Specifically, a local brightness difference analysis is performed on multiple target ice-covered areas to extract the thickness characteristics of each target ice-covered area, and a local texture analysis is performed on multiple target ice-covered areas to extract the texture characteristics of each target ice-covered area, including: For the thickness feature, the grayscale image corresponding to each target ice-covered area is extracted, and sliding window analysis is performed on multiple grayscale images to extract the local grayscale mean and local contrast of each window area. The global grayscale mean and global contrast of the target ice-covered area are calculated based on multiple local grayscale means and local contrasts, and the global grayscale standard deviation of the target ice-covered area is calculated based on the global grayscale mean.

[0030] In this embodiment, for each target ice-covered area, its corresponding grayscale image is extracted and brightness difference analysis is performed. The core goal of brightness difference analysis is to evaluate the degree of grayscale change in each area, so as to reflect the thickness characteristics of the ice layer. In this process, a sliding window analysis is performed on the grayscale image of each target area, and a fixed-size sliding window is used to traverse the entire image. The size of the sliding window can be reasonably set based on the size of the grayscale image of the target area, and is not specifically limited here. In the process of traversing the grayscale image through the sliding window, the local grayscale mean and local contrast in each window area are calculated to reflect the local brightness difference in the area. The ice layer often has different refractive index, light transmittance and surface roughness. The greater the local brightness difference, the greater the thickness of the ice layer.

[0031] Specifically, the local grayscale mean can describe the brightness level in the area and reflect the overall brightness of the target ice layer. The local contrast specifically represents the difference between the extreme values ​​of the grayscale in the area. The larger the local contrast, the more significant the brightness difference in the area. The higher local contrast may indicate that the thickness of the ice layer is large, or the ice surface has strong reflectivity. Finally, the local grayscale means and local contrasts of multiple sliding window areas are summarized to calculate the global grayscale mean and global contrast of the entire target ice area. These global statistics can summarize the overall brightness characteristics of the target area and provide a basis for further analysis of the ice thickness. Then, the global grayscale mean is used to further analyze the multiple local grayscale means in the sliding window analysis, and the global grayscale standard deviation corresponding to the multiple local grayscale means is calculated to reflect the discrete degree of the grayscale distribution of the entire target area. A larger grayscale standard deviation usually corresponds to a thicker ice layer, especially when the ice surface is uneven.

[0032] Based on the edge detection operator, gradient analysis is performed on multiple grayscale images to obtain the gradient image of each grayscale image. Multiple local gradient peaks are extracted from the gradient image through sliding window analysis, and the global gradient peak of each target ice-covered area is calculated to obtain the thickness characteristics of each target ice-covered area.

[0033] In this embodiment, the grayscale image is also subjected to gradient analysis based on the edge detection operator to further extract the thickness characteristics of the ice-covered area. Specifically, for example, the Sobel operator can be used for edge detection and gradient analysis to obtain the gradient image of each grayscale image, which reflects the most significant grayscale change in the image. Usually, these areas correspond to the edge or depth change of the ice layer, which is particularly important for judging the edge and thickness change of the ice layer. In the gradient image, the sliding window traverses each area to extract the local gradient peak value in each window area, such as the maximum gradient value of each window area. These local gradient peak values ​​represent the most significant grayscale change in the image, which can reflect the sharpness of the edge of the ice layer and the thickness of the ice layer to a certain extent. Then, multiple local gradient peak values ​​are aggregated to calculate the global gradient peak value of each target ice-covered area to better summarize the edge characteristics of the area, which is helpful to accurately judge the thickness of the ice layer in the target ice-covered area. A larger gradient peak value usually means that the ice layer has greater unevenness or more significant thickness changes. The thickness characteristics of the target ice-covered area are characterized by the global features of grayscale, contrast and gradient values ​​extracted above.

[0034] For texture features, sliding window analysis is performed on multiple grayscale images respectively, and each window area is processed by a Gabor filter to obtain multiple Gabor filter responses of each window area. Multiple local texture feature parameters of each window area are extracted based on the multiple Gabor filter responses. Multiple global texture feature parameters of each grayscale image are generated based on the multiple local texture features to obtain the texture features of each target ice-covered area.

[0035] In this embodiment, local texture analysis can help extract the surface structural features of the target area, which is of great significance for identifying the type, roughness and other surface properties of the ice layer. Gabor filter is a tool commonly used for texture analysis. It can perform multi-scale and multi-directional filtering on the image to capture texture information of different directions and frequencies in the image. The grayscale image of each target area is subjected to sliding window analysis by Gabor filter, and each window area is subjected to multi-scale and multi-directional filtering processing by Gabor filter. Gabor filter can extract the frequency domain features of the image, and is particularly effective in capturing local texture information. By generating multiple Gabor filter responses, multiple texture features of each window area can be obtained, including but not limited to information such as directionality and frequency response, so as to effectively describe the surface texture structure of the area. For example, the surface of the ice layer may have regular texture structures, which appear as strong frequency components in the filter response.

[0036] Finally, multiple local texture features are fused, for example, the average of multiple texture features is taken to obtain multiple global texture feature parameters of each grayscale image, so as to globally characterize the texture structure information of the target ice-covered area and reflect the texture properties of the entire target area, such as roughness, directionality and other characteristics. The roughness of the texture is usually closely related to the surface morphology of the ice layer, while the complexity and directionality of the texture may be related to the formation process and thickness of the ice layer.

[0037] The thickness features and texture features extracted by the local brightness difference analysis and local texture analysis can comprehensively evaluate the ice layer characteristics of each target ice-covered area, such as identifying ice layers of different thicknesses and judging the state of the ice layer.

[0038] In one implementation process, for step S30, in order to accurately analyze the ice features in the line ice image, an ice dataset is constructed based on the thickness features and texture features of the target ice area, and an ice analysis model is obtained by training the ice dataset to analyze the ice image. Specifically, an ice dataset is constructed based on multiple ice visual features and ice annotation data, and an ice analysis model is obtained by training the ice dataset, including the following contents: The ice thickness label and ice type label of each line ice image are extracted from the ice annotation data. An ice dataset is constructed based on multiple ice visual features. The corresponding ice thickness label and ice type label are added to each line ice image in the ice dataset. The ice analysis model is trained using the ice dataset.

[0039] In this embodiment, the ice-covered visual features are used as the main features of each line ice-covered image, and the label information of each ice-covered visual feature is determined based on the ice-covered annotation data, including determining the ice-covered thickness label and ice-covered type label of each line ice-covered image, wherein the ice-covered thickness label is used to indicate the actual ice-covered thickness of each line ice-covered image, and the ice-covered type label is used to indicate the appearance state of the ice, such as rime, sleet, hoarfrost, etc., and finally an ice-covered data set for training the ice-covered analysis model is constructed.

[0040] In this embodiment, the icing analysis model is a multi-task deep neural network used to perform icing analysis tasks. The goal is to complete two tasks at the same time, including ice thickness regression and ice type classification. They have different goals but rely on the same input data. The multi-task deep neural network can share the feature extraction layer and design special branches for different tasks, which can well implement the icing analysis task.

[0041] Among them, the ice analysis model includes an input layer, a shared feature extraction layer, an ice thickness regression branch, and an ice type classification branch. The input layer is used to receive multiple ice visual features in the ice dataset, that is, the comprehensive feature data extracted in the above steps, including the grayscale information, texture information, gradient information, etc. of the image. The shared feature extraction layer is used to extract high-order features from the input ice visual features. The extracted features are shared by two tasks, namely thickness regression and type classification, so they can be processed by a common network layer, thereby reducing the complexity of the model. Its structure can be multiple convolutional layers for feature extraction. The ice thickness regression branch is used to predict the ice thickness of each line ice image. The goal of the regression problem is to predict a continuous value, namely the ice thickness, through the input ice visual features. It can be composed of a fully connected layer, and finally output a scalar value to represent the predicted ice thickness. The linear activation function is used to realize the prediction of continuous numerical values. The ice type classification branch is used to predict the ice type of each line ice image. According to the visual features of ice cover in the image, this branch will output a category label to indicate which type of ice cover the image belongs to. It can be composed of a fully connected layer and a Softmax activation function. The output is converted into a probability value for each category through the Softmax function, and the number of neurons in the output layer is determined according to the actual number of ice cover types.

[0042] In multi-task learning, the model needs to optimize the loss functions of multiple tasks at the same time. In order to achieve this goal, it is necessary to define the loss function of each task and combine them into a total loss function according to the weighted importance of each task, that is, to construct the total loss function of the ice analysis model according to the loss functions corresponding to the ice thickness regression branch and the ice type classification branch. In this embodiment, for the regression task, the mean square error (MSE) function is used as the loss function to calculate the square difference between the predicted value and the true label. The smaller the MSE, the more accurate the model's prediction. For the classification task, the cross entropy loss is used as the loss function to measure the difference between the probability distribution output by the classification model and the actual label. The total loss function is the weighted sum of the loss functions of the two tasks, and the influence of the regression task and the classification task are balanced by two weight parameters respectively.

[0043] During the model training process, multiple ice visual features in the ice dataset are used as input, the ice thickness label corresponding to the line ice image is used as the training target of the ice thickness regression branch, and the ice type label corresponding to the line ice image is used as the training target of the ice type classification branch to optimize the model parameters. Finally, the ice analysis model is obtained through training the ice dataset. In practical applications, automatic analysis of line ice images can be realized, including predicting ice thickness and classifying different types of ice layers.

[0044] In one implementation process, for step S40, in order to further perform group feature analysis on multiple line ice images according to multiple ice visual features, ice thickness data, pre-classification of sample sets and K-means clustering algorithm are combined to perform ice feature clustering on multiple line ice images to generate multiple ice feature clusters. The specific implementation process is as follows: The sample ice thickness range is determined according to the ice annotation data in the historical image set, and the sample ice thickness range is divided into multiple thickness intervals based on a preset thickness interval. Multiple line ice images are pre-classified according to the ice annotation data to obtain a sub-sample set for each thickness interval.

[0045] In this embodiment, the ice-covered annotation data in the historical image set is used to determine the ice-covered thickness range, providing the necessary thickness reference for the clustering process, so as to reasonably classify ice-covered samples of different thicknesses. The ice-covered annotation data in the historical image set includes the actual ice-covered thickness information corresponding to each image, so as to determine the ice-covered thickness range of the sample, and then the ice-covered thickness range is divided into multiple thickness intervals, each thickness interval represents an ice-covered layer of different thicknesses, such as a series of intervals such as 0mm-5mm, 5mm-10mm, and 10mm-15mm. After determining multiple thickness intervals, multiple line ice-covered images are pre-classified based on the ice-covered thickness data, and each line ice-covered image is classified into a specific thickness interval according to its thickness, so as to generate a corresponding sub-sample set for each thickness interval, and each sub-sample set contains line ice-covered image data within the thickness range.

[0046] The initial cluster center of each thickness interval is determined according to multiple sub-sample sets, and the K-means clustering algorithm is used to cluster the ice features of multiple line ice images based on the preset thickness range.

[0047] In this embodiment, the K value of K-means clustering is determined according to the number of sub-sample sets. The selection of the initial cluster center when performing K-means clustering is a crucial step, which directly affects the quality of the clustering results. In order to ensure that the clustering process can reflect the ice characteristics of different thickness intervals, the initial cluster center of each thickness interval is determined based on the data in the sub-sample set. For example, the center point of each thickness interval is selected as the reference point, and the ice visual features of the ice thickness data closest to the reference point in the sub-sample set are selected as the initial cluster center of the thickness interval. Then, based on the multiple initial cluster centers, the K-means clustering algorithm is used to cluster the ice features of multiple line ice images based on the preset thickness range.

[0048] In the clustering process, the K-means clustering algorithm will assign samples to their respective clusters according to the current cluster center in each iteration, and update the cluster center according to the characteristics of the samples. In order to enhance the clustering effect and limit the clustering of each cluster to the ice thickness as the main factor and the ice type as a secondary factor, a restriction on the ice thickness range is added in the iteration process.

[0049] Specifically, in the iterative process, after multiple samples are assigned to different clusters, for multiple samples in a certain ice-covered feature cluster, if the cluster centers corresponding to the multiple samples assigned to the current ice-covered feature cluster are calculated to deviate from the preset thickness range of the thickness interval, the cluster center of the current ice-covered feature cluster is optimized, and the multiple samples assigned to the current ice-covered feature cluster are sampled, and a new cluster center is determined based on the multiple samples obtained by screening. In the screening process, the objects whose ice-covered thicknesses belong to the preset thickness range among the multiple samples are recorded as target samples, and the cluster center of the ice-covered feature cluster is determined based on the multiple target samples. In this way, the cluster center of each ice-covered feature cluster in the clustering process is limited based on the ice-covered thickness, ensuring that the cluster center of each ice-covered feature cluster has a reference to the ice-covered thickness. In this process, the preset thickness range can be set according to the thickness interval corresponding to the ice-covered feature cluster. For example, if the thickness interval is 5mm-10mm, the preset thickness range takes a certain proportion of the thickness interval, for example, 60% of the thickness interval, i.e., 6mm-9mm.

[0050] After multiple iterations, if the preset cluster center fluctuation condition is met, for example, the position of each cluster center no longer changes significantly, clustering ends. Multiple ice feature clusters are obtained by clustering the ice thickness and visual features of the samples. Each cluster represents a specific group of ice features, in which samples have similar visual features and thickness features. Multiple ice feature clusters can represent the distribution of ice features in different ice thickness ranges. The samples in each cluster can represent some similar visual features between different ice types in a specific ice thickness scenario, providing a data basis for the analysis and monitoring of the ice layer.

[0051] In one implementation process, for step 50, in order to accurately analyze the ice-covered features of the ice-covered image to be analyzed, a first ice-covered analysis result corresponding to the ice-covered image is determined according to a plurality of ice-covered feature clusters. This process includes ice-covered thickness analysis and ice-covered type analysis. The ice-covered image to be analyzed is deeply analyzed according to the previously constructed ice-covered feature clusters to generate an analysis result. Specifically, determining the first ice-covered analysis result of the ice-covered image to be analyzed according to the plurality of ice-covered feature clusters includes the following contents: Determine the thickness cluster center of each ice feature cluster, match the target ice feature of the ice image to be analyzed with multiple thickness cluster centers, determine the thickness cluster center closest to the target ice feature, and generate a first ice thickness analysis result for the ice image to be analyzed.

[0052] In this embodiment, the thickness cluster center of each ice-covered feature cluster may be determined by the final iteration in the clustering process, and describes typical visual features of the image in the cluster, such as grayscale, texture, gradient, etc. The distance between the target ice-covered feature of the ice-covered image to be analyzed and the centers of multiple thickness clusters is calculated for matching, for example, the ice-covered feature cluster that best matches the feature of the image to be analyzed is identified by the Euclidean distance, thereby preliminarily determining the ice-covered thickness related information of the ice-covered image to be analyzed, that is, the thickness interval corresponding to the ice-covered feature cluster, and obtaining the first ice-covered thickness analysis result of the ice-covered image to be analyzed.

[0053] Then, the icing type of the icing image to be analyzed is further analyzed, including classification of multiple line icing images in the icing feature cluster closest to the target icing feature based on the icing type, and obtaining multiple sub-icing type clusters, that is, multiple line icing images belonging to the same icing type in the icing feature cluster belong to the same sub-icing type cluster, and the type cluster center of each sub-icing type cluster can be calculated by multiple samples in the cluster, for example, the feature mean of multiple samples is calculated as the type cluster center of the cluster. Then, the target icing feature of the icing image to be analyzed is matched with multiple type cluster centers, and a distance measurement formula such as Euclidean distance is used to determine the type cluster center closest to the target icing feature, and the icing type corresponding to the type cluster center is used as the icing type of the icing image to be analyzed, and the first icing type analysis result of the icing image to be analyzed is obtained, and the first icing thickness analysis result and the first icing type analysis result are summarized to obtain the first icing analysis result of the icing image to be analyzed, so as to realize the icing characteristic analysis of the icing image to be analyzed from the perspective of cluster group feature similarity.

[0054] In one implementation process, for step 50, in order to further improve the accuracy of ice image analysis, the fusion of the second ice analysis result and the first ice analysis result is an important step. By reasonably combining the outputs of the two analysis methods, a more accurate and reliable target ice analysis result can be obtained. Specifically, the first ice analysis result and the second ice analysis result are fused to obtain the target ice analysis result of the ice image to be analyzed, including the following contents: A second ice coating thickness analysis result and a second ice coating type analysis result generated by the ice coating analysis model in the second ice coating analysis result are determined, and the first ice coating thickness analysis result is matched with the second ice coating thickness analysis result.

[0055] In this embodiment, the purpose of matching the first ice coating thickness analysis result with the second ice coating thickness analysis result is to determine whether the second ice coating thickness analysis result belongs to the thickness interval corresponding to the first ice coating thickness analysis result, wherein the second ice coating thickness analysis result and the second ice coating type analysis result respectively include the ice coating thickness prediction value and the ice coating type prediction value obtained by the ice coating analysis model for the ice coating image to be analyzed.

[0056] If the match is successful, that is, the second ice thickness analysis result belongs to the thickness interval corresponding to the first ice thickness analysis result, since the generation of ice feature clusters is dominated by ice thickness, it means that the detailed ice feature information generated by the ice analysis model has a strong reference value, and the second ice result is directly used as the target ice analysis result of the ice image to be analyzed; If the matching fails, that is, the second ice thickness analysis result does not belong to the thickness interval corresponding to the first ice thickness analysis result, the first reference cluster is determined according to the type cluster center closest to the target ice feature, and the first similarity between the target ice feature and the first reference cluster is calculated. The similarity calculation formula such as cosine similarity can be used, and the absolute value of the cosine similarity is taken as the first similarity.

[0057] Then, the second reference cluster associated with the second icing analysis result is determined, specifically, the icing feature cluster corresponding to the thickness interval to which the second icing thickness analysis result belongs, and the second similarity between the target icing feature and the second reference cluster is calculated. At the same time, the first reference cluster associated with the target icing feature is further analyzed, and the third similarity between the target icing feature and the most matching sub-icing type cluster in the first reference cluster is calculated. The calculation method of the multiple similarities is specifically calculated using the calculation method of the first similarity.

[0058] Finally, the first similarity is corrected by the third similarity, and the icing analysis result corresponding to the maximum value between the corrected first similarity and the second similarity is selected as the target icing analysis result of the icing image to be analyzed, that is, the result obtained by analyzing the icing characteristics of the icing image to be analyzed from the perspective of cluster group feature similarity, and a more in-depth assessment of the overall feature similarity is conducted from the multiple sub-icing type clusters contained therein. After comprehensively considering the icing type, a more reference-worthy result is determined from the first icing analysis result and the second icing analysis result to obtain the final target icing analysis result. The icing feature analysis results from multiple angles are fused in the above manner, and the advantages of different schemes are combined to obtain high-precision icing analysis results, thereby improving the accuracy and reliability of icing detection, and finally providing a comprehensive and accurate analysis basis for icing detection of transmission lines.

[0059] See also Figure 2, which is a structural schematic diagram of a transmission line icing analysis system based on image recognition provided in one embodiment of the present invention, comprising: An ice data acquisition module is used to acquire a historical image set about ice detection of power transmission lines collected by a camera device, wherein the historical image set includes a plurality of line ice images and ice annotation data of each line ice image; An ice feature extraction module is used to determine the target ice area of ​​each line ice image according to the ice annotation data, perform local brightness difference analysis on multiple target ice areas to extract the thickness feature of each target ice area, and perform local texture analysis on multiple target ice areas to extract the texture feature of each target ice area; The icing analysis model training module is used to generate icing visual features of each target icing area by combining the thickness features and texture features of the target icing area, construct an icing dataset based on multiple icing visual features and icing annotation data, and obtain an icing analysis model through icing dataset training; Specifically, an icing dataset is constructed based on multiple icing visual features and icing annotation data, and an icing analysis model is obtained by training the icing dataset, including: The ice thickness label and ice type label of each line ice image are extracted from the ice annotation data, and an ice dataset is constructed based on multiple ice visual features. The corresponding ice thickness label and ice type label are added to each line ice image in the ice dataset, and the ice analysis model is trained through the ice dataset. Among them, the icing analysis model is a multi-task deep neural network, including an input layer, a shared feature extraction layer, an ice thickness regression branch and an ice type classification branch. The total loss function of the icing analysis model is constructed according to the loss functions corresponding to the ice thickness regression branch and the ice type classification branch respectively. During the model training process, multiple icing visual features in the icing dataset are used as input, the ice thickness label corresponding to the line icing image is used as the training target of the ice thickness regression branch, and the ice type label corresponding to the line icing image is used as the training target of the icing type classification branch. The icing analysis model is obtained by training the icing dataset.

[0060] An ice feature clustering analysis module is used to cluster ice features of multiple line ice images according to multiple ice visual features to generate multiple ice feature clusters, including determining multiple initial cluster centers based on ice annotation data, using a K-means clustering algorithm based on cluster center restriction optimization, and clustering multiple line ice images based on multiple ice visual features to obtain multiple ice feature clusters; Specifically, performing ice-covered feature clustering on multiple line ice-covered images according to multiple ice-covered visual features to generate multiple ice-covered feature clusters includes: Determine the sample ice thickness range according to the ice annotation data in the historical image set, divide the sample ice thickness range into multiple thickness intervals based on the preset thickness interval, and pre-classify multiple line ice images according to the ice annotation data to obtain a sub-sample set for each thickness interval; An initial cluster center of each thickness interval is determined according to multiple sub-sample sets, and a K-means clustering algorithm is used to cluster ice features of multiple line ice images based on a preset thickness range according to the multiple initial cluster centers, including in each iteration process, if the cluster centers corresponding to multiple samples assigned to the current ice feature cluster calculated deviate from the preset thickness range of the thickness interval to which the ice feature cluster belongs, then according to the preset thickness range of the thickness interval to which the ice feature cluster belongs, the multiple samples assigned to the current ice feature cluster are screened, and multiple target samples whose ice thickness belongs to the preset thickness range among the multiple samples are determined, and the cluster center of the ice feature cluster is determined according to the multiple target samples. After the preset cluster center fluctuation condition is met, the clustering of the multiple line ice images is completed to obtain multiple ice feature clusters.

[0061] The icing feature analysis module is used to extract the target icing features of the icing image to be analyzed after collecting the icing image to be analyzed of the transmission line, determine the first icing analysis result of the icing image to be analyzed according to multiple icing feature clusters, process the target icing features through the icing analysis model to generate a second icing analysis result, and fuse the first icing analysis result and the second icing analysis result to obtain the target icing analysis result of the icing image to be analyzed.

[0062] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. Parts not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. A method for analyzing ice coating on power transmission lines based on image recognition, characterized in that: include: Acquire a historical image set about ice detection of power transmission lines collected by a camera device, wherein the historical image set includes a plurality of line ice images and ice annotation data of each line ice image; Determine the target ice-covered area of ​​each line ice-covered image according to the ice-covered annotation data, perform local brightness difference analysis on multiple target ice-covered areas to extract thickness features of each target ice-covered area, and perform local texture analysis on multiple target ice-covered areas to extract texture features of each target ice-covered area; The thickness features and texture features of the target ice-covered area are spliced ​​to generate ice-covered visual features of each target ice-covered area. An ice-covered dataset is constructed based on multiple ice-covered visual features and ice-covered annotation data. An ice-covered analysis model is obtained through training with the ice-covered dataset. Clustering the ice-covered images of multiple lines according to the multiple ice-covered visual features to generate multiple ice-covered feature clusters, including determining multiple initial cluster centers based on ice-covered annotation data, using a K-means clustering algorithm based on cluster center restriction optimization, and clustering the multiple ice-covered images of multiple lines according to the multiple ice-covered visual features to obtain multiple ice-covered feature clusters; After collecting the ice-covered image to be analyzed of the transmission line, the target ice-covered features of the ice-covered image to be analyzed are extracted, and a first ice-covered analysis result of the ice-covered image to be analyzed is determined according to multiple ice-covered feature clusters. The target ice-covered features are processed by an ice-covered analysis model to generate a second ice-covered analysis result. The first ice-covered analysis result and the second ice-covered analysis result are fused to obtain the target ice-covered analysis result of the ice-covered image to be analyzed.

2. The method for analyzing ice coating on power transmission lines based on image recognition according to claim 1, characterized in that: Perform local brightness difference analysis on multiple target ice-covered areas to extract thickness features of each target ice-covered area, and perform local texture analysis on multiple target ice-covered areas to extract texture features of each target ice-covered area, including: The grayscale image corresponding to each target ice-covered area is extracted, and sliding window analysis is performed on multiple grayscale images to extract the local grayscale mean and local contrast of each window area. The global grayscale mean and global contrast of the target ice-covered area are calculated based on multiple local grayscale means and local contrasts, and the global grayscale standard deviation of the target ice-covered area is calculated based on the global grayscale mean. Based on the edge detection operator, a gradient analysis is performed on multiple grayscale images to obtain a gradient image of each grayscale image. Multiple local gradient peaks are extracted from the gradient image through sliding window analysis, and the global gradient peak of each target ice-covered area is calculated to obtain the thickness characteristics of each target ice-covered area. Sliding window analysis is performed on multiple grayscale images respectively, and each window area is processed by Gabor filter to obtain multiple Gabor filter responses of each window area. Multiple local texture feature parameters of each window area are extracted according to the multiple Gabor filter responses. Multiple global texture feature parameters of each grayscale image are generated according to the multiple local texture features to obtain the texture features of each target ice-covered area.

3. The method for analyzing ice coating on power transmission lines based on image recognition according to claim 1, characterized in that: Based on multiple ice visual features, multiple line ice images are clustered to generate multiple ice feature clusters, including: Determine the sample ice thickness range according to the ice annotation data in the historical image set, divide the sample ice thickness range into multiple thickness intervals based on the preset thickness interval, and pre-classify multiple line ice images according to the ice annotation data to obtain a sub-sample set for each thickness interval; An initial cluster center of each thickness interval is determined according to multiple sub-sample sets, and a K-means clustering algorithm is used to cluster ice features of multiple line ice images based on a preset thickness range according to the multiple initial cluster centers, including in each iteration process, if the cluster centers corresponding to multiple samples assigned to the current ice feature cluster calculated deviate from the preset thickness range of the thickness interval to which the ice feature cluster belongs, then according to the preset thickness range of the thickness interval to which the ice feature cluster belongs, the multiple samples assigned to the current ice feature cluster are screened, and multiple target samples whose ice thickness belongs to the preset thickness range among the multiple samples are determined, and the cluster center of the ice feature cluster is determined according to the multiple target samples. After the preset cluster center fluctuation condition is met, the clustering of the multiple line ice images is completed to obtain multiple ice feature clusters.

4. The method for analyzing ice coating on power transmission lines based on image recognition according to claim 1, characterized in that: An icing dataset is constructed based on multiple icing visual features and icing annotation data. An icing analysis model is obtained by training the icing dataset, including: The ice thickness label and ice type label of each line ice image are extracted from the ice annotation data, and an ice dataset is constructed based on multiple ice visual features. The corresponding ice thickness label and ice type label are added to each line ice image in the ice dataset, and the ice analysis model is trained through the ice dataset. Among them, the icing analysis model is a multi-task deep neural network, including an input layer, a shared feature extraction layer, an ice thickness regression branch and an ice type classification branch. The total loss function of the icing analysis model is constructed according to the loss functions corresponding to the ice thickness regression branch and the ice type classification branch respectively. During the model training process, multiple icing visual features in the icing dataset are used as input, the ice thickness label corresponding to the line icing image is used as the training target of the ice thickness regression branch, and the ice type label corresponding to the line icing image is used as the training target of the icing type classification branch. The icing analysis model is obtained by training the icing dataset.

5. The method for analyzing ice coating on power transmission lines based on image recognition according to claim 4, characterized in that: Determining a first ice-covering analysis result of the ice-covering image to be analyzed according to the plurality of ice-covering feature clusters includes: Determine the thickness cluster center of each ice feature cluster, match the target ice feature of the ice image to be analyzed with multiple thickness cluster centers, determine the thickness cluster center closest to the target ice feature, and generate a first ice thickness analysis result of the ice image to be analyzed; Multiple line icing images in the icing feature cluster are classified based on icing type to obtain multiple sub-icing type clusters, the type cluster center of each sub-icing type cluster is determined, the target icing feature of the icing image to be analyzed is matched with the multiple type cluster centers, the type cluster center closest to the target icing feature is determined, and a first icing type analysis result of the icing image to be analyzed is generated; and a first icing analysis result of the icing image to be analyzed is determined according to the first icing thickness analysis result and the first icing type analysis result.

6. The method for analyzing ice coating on power transmission lines based on image recognition according to claim 5, characterized in that: The first ice analysis result and the second ice analysis result are merged to obtain a target ice analysis result of the ice image to be analyzed, including: Determine a second ice thickness analysis result and a second ice type analysis result generated by the ice analysis model in the second ice analysis result, and match the first ice thickness analysis result with the second ice thickness analysis result; If the first ice thickness analysis result matches the second ice thickness analysis result successfully, the second ice thickness analysis result is used as the target ice thickness analysis result of the ice image to be analyzed; Otherwise, a first reference cluster is determined according to the type cluster center closest to the target icing feature, a first similarity between the target icing feature and the first reference cluster is calculated, a second reference cluster associated with the second icing analysis result is determined, and a second similarity between the target icing feature and the second reference cluster is calculated; The third similarity between the target icing feature and the most matching sub-icing type cluster in the first reference cluster is calculated. After correcting the first similarity based on the third similarity, the icing analysis result corresponding to the maximum value between the corrected first similarity and the second similarity is selected as the target icing analysis result of the icing image to be analyzed.

7. A transmission line icing analysis system based on image recognition, characterized in that: The system is used to implement the method for analyzing ice coating on a power transmission line based on image recognition as described in any one of claims 1 to 6, comprising: An ice data acquisition module is used to acquire a historical image set about ice detection of power transmission lines collected by a camera device, wherein the historical image set includes a plurality of line ice images and ice annotation data of each line ice image; An ice feature extraction module is used to determine the target ice area of ​​each line ice image according to the ice annotation data, perform local brightness difference analysis on multiple target ice areas to extract the thickness feature of each target ice area, and perform local texture analysis on multiple target ice areas to extract the texture feature of each target ice area; The icing analysis model training module is used to generate icing visual features of each target icing area by combining the thickness features and texture features of the target icing area, construct an icing dataset based on multiple icing visual features and icing annotation data, and obtain an icing analysis model through icing dataset training; An ice feature clustering analysis module is used to cluster ice features of multiple line ice images according to multiple ice visual features to generate multiple ice feature clusters, including determining multiple initial cluster centers based on ice annotation data, using a K-means clustering algorithm based on cluster center restriction optimization, and clustering multiple line ice images based on multiple ice visual features to obtain multiple ice feature clusters; The icing feature analysis module is used to extract the target icing features of the icing image to be analyzed after collecting the icing image to be analyzed of the transmission line, determine the first icing analysis result of the icing image to be analyzed according to multiple icing feature clusters, process the target icing features through the icing analysis model to generate a second icing analysis result, and fuse the first icing analysis result and the second icing analysis result to obtain the target icing analysis result of the icing image to be analyzed.

8. The power transmission line icing analysis system based on image recognition according to claim 7, characterized in that: The ice feature clustering analysis module performs ice feature clustering on multiple line ice images according to multiple ice visual features to generate multiple ice feature clusters, including: Determine the sample ice thickness range according to the ice annotation data in the historical image set, divide the sample ice thickness range into multiple thickness intervals based on the preset thickness interval, and pre-classify multiple line ice images according to the ice annotation data to obtain a sub-sample set for each thickness interval; An initial cluster center of each thickness interval is determined according to multiple sub-sample sets, and a K-means clustering algorithm is used to cluster ice features of multiple line ice images based on a preset thickness range according to the multiple initial cluster centers, including in each iteration process, if the cluster centers corresponding to multiple samples assigned to the current ice feature cluster calculated deviate from the preset thickness range of the thickness interval to which the ice feature cluster belongs, then according to the preset thickness range of the thickness interval to which the ice feature cluster belongs, the multiple samples assigned to the current ice feature cluster are screened, and multiple target samples whose ice thickness belongs to the preset thickness range among the multiple samples are determined, and the cluster center of the ice feature cluster is determined according to the multiple target samples. After the preset cluster center fluctuation condition is met, the clustering of the multiple line ice images is completed to obtain multiple ice feature clusters.

Images