An insulator testing method, apparatus, electronic device, and storage medium

Abstract

The application discloses an insulator detection method and device, electronic equipment and a storage medium, and relates to the technical field of insulator detection. The method comprises the following steps: performing convolution processing on an infrared image corresponding to an insulator to be detected to obtain a differential image to be processed, and extracting at least one feature point to be processed in the differential image to be processed; determining pixel point association information corresponding to each feature point to be processed, and determining a feature vector to be compared corresponding to each feature point to be processed based on the pixel point association information; determining an error to be determined corresponding to each feature point to be processed according to the feature vector to be compared and a target feature vector corresponding to each feature point to be processed; and determining a degraded insulating area in the insulator to be detected according to a target feature point with an error to be determined less than a preset error. The effect of accurately detecting the degraded insulating area in the insulator based on an infrared image containing the insulator without being affected by a temperature environment is achieved, the detection efficiency is improved, and the workload of the staff is reduced.

Description

Technical Field

[0001] This invention relates to the field of power grid testing technology, and in particular to an insulator testing method, device, electronic equipment, and storage medium. Background Technology

[0002] Outdoor insulators are exposed to the air and subjected to rain, snow, frost, dew, etc., and are subjected to working voltage and wire tension for a long time. This may cause the insulation to reach zero value, affecting the normal operation of the power grid.

[0003] Currently, insulator inspection typically involves either capturing images of the insulator and manually inspecting them to determine if degradation exists, or collecting infrared images of the insulator and using the temperature information within those images to assess degradation. However, manual inspection is inaccurate, and infrared image-based inspection is easily affected by external environmental factors.

[0004] To solve the above problems, the testing methods for insulators need to be improved. Summary of the Invention

[0005] This invention provides an insulator testing method, device, electronic device, and storage medium to solve the problems of low testing efficiency and inaccurate test results when testing insulators by manual inspection, or the significant impact of environmental factors on testing based on temperature information in infrared images.

[0006] In a first aspect, embodiments of the present invention provide an insulator detection method, comprising:

[0007] The infrared image corresponding to the insulator to be detected is convolved to obtain a differential image to be processed, and at least one feature point to be processed is extracted from the differential image to be processed.

[0008] Determine the pixel association information corresponding to each feature point to be processed, and determine the feature vector to be compared corresponding to the feature point to be processed based on the pixel association information; wherein, the pixel association information includes at least one associated pixel in the neighborhood of the corresponding feature point to be processed, as well as the distance and angle to be determined of each associated pixel relative to the feature point to be processed.

[0009] Based on the comparison vector of each feature point to be processed and the corresponding target feature vector, the error to be determined for each feature point to be processed is determined; wherein, the target feature vector is a predetermined vector.

[0010] Based on the target feature points whose error is less than the preset error, the deteriorated insulation area in the insulator to be tested is determined;

[0011] The target feature point is a feature point among the at least one feature point to be processed.

[0012] Secondly, embodiments of the present invention also provide an insulator detection device, comprising:

[0013] The feature point determination module is used to perform convolution processing on the infrared image corresponding to the insulator to be detected to obtain a differential image to be processed, and to extract at least one feature point to be processed from the differential image to be processed.

[0014] The feature vector determination module is used to determine the pixel association information corresponding to each feature point to be processed, and to determine the feature vector to be compared corresponding to the feature point to be processed based on the pixel association information; wherein, the pixel association information includes at least one associated pixel in the neighborhood of the corresponding feature point to be processed, and the distance and angle to be determined of each associated pixel relative to the feature point to be processed.

[0015] The undetermined error determination module is used to determine the undetermined error corresponding to the corresponding feature point to be processed based on the comparison vector of each feature point to be processed and the corresponding target feature vector; wherein, the target feature vector is a predetermined vector;

[0016] The deteriorated insulation region determination module is used to determine the deteriorated insulation region in the insulator to be tested based on the target feature points whose error to be determined is less than a preset error.

[0017] The target feature point is a feature point among the at least one feature point to be processed.

[0018] Thirdly, embodiments of the present invention also provide an electronic device, comprising:

[0019] At least one processor; and

[0020] A memory communicatively connected to the at least one processor; wherein,

[0021] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the insulator detection method according to any embodiment of the present invention.

[0022] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer instructions, which are used to cause a processor to execute and implement the insulator detection method described in any embodiment of the present invention.

[0023] The technical solution of this invention involves convolution processing of the infrared image corresponding to the insulator to be tested to obtain a differential image to be processed, and extracting at least one feature point to be processed from the differential image. It then determines the pixel association information corresponding to each feature point to be processed, and determines the corresponding comparison feature vector based on the pixel association information. Based on the comparison vector of each feature point to be processed and the corresponding target feature vector, it determines the error to be determined for each feature point to be processed. Finally, based on target feature points where the error to be determined is less than a preset error, it determines the degraded insulation region in the insulator to be tested. This solves the problems of low detection efficiency and inaccurate results when manually inspecting insulators, or the significant environmental influences when using temperature information from infrared images. It achieves accurate detection of degraded insulation regions in insulators based on infrared images containing the insulator, without being affected by temperature conditions, while also improving detection efficiency and reducing the workload of personnel.

[0024] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a flowchart of an insulator testing method provided in Embodiment 1 of the present invention;

[0027] Figure 2 This is a schematic diagram of the structure of an infrared imaging detection device according to Embodiment 1 of the present invention;

[0028] Figure 3 This is a flowchart of an insulator testing method provided in Embodiment 2 of the present invention;

[0029] Figure 4 This is a schematic diagram of determining a group of direction vectors to be processed according to Embodiment 2 of the present invention;

[0030] Figure 5 This is a schematic diagram of the angle between direction vectors to be determined according to Embodiment 2 of the present invention;

[0031] Figure 6This is a schematic diagram of the structure of an insulator testing device according to Embodiment 3 of the present invention;

[0032] Figure 7 This is a schematic diagram of the structure of an electronic device that implements the insulator detection method of this invention. Detailed Implementation

[0033] 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.

[0034] 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 sequences other than those illustrated or described herein.

[0035] Before elaborating on this technical solution, a brief introduction to its application scenarios is provided to facilitate a clearer understanding. Outdoor insulators, exposed to air, rain, snow, frost, and dew, and subjected to long-term working voltage and wire tension, may experience insulation failure. In severe cases, this can lead to zero insulation strength, resulting in single-phase grounding, flashover, breakdown, or power outages. This will seriously affect the safe operation of the power grid and cause significant economic losses. Therefore, to ensure the safe operation of the power grid, it is necessary to promptly inspect zero-value insulators, that is, to inspect the deteriorated insulation areas of the insulators.

[0036] Example 1

[0037] Figure 1 The flowchart of an insulator detection method is provided in Embodiment 1 of the present invention. This embodiment is applicable to the situation of accurately detecting whether there are deteriorated insulation areas in an insulator based on infrared images. The method can be executed by an insulator detection device, which can be implemented in hardware and / or software. The insulator detection device can be configured in a computing device that can execute the insulator detection method.

[0038] like Figure 1 As shown, the method includes:

[0039] S110. Perform convolution processing on the infrared image corresponding to the insulator to be detected to obtain the differential image to be processed, and extract at least one feature point to be processed from the differential image to be processed.

[0040] Understandably, an insulator is a special type of insulating component that plays a crucial role in overhead transmission lines. In the past, insulators were primarily used on utility poles. Gradually, they evolved into disc-shaped insulators hung at one end of high-voltage power line towers to increase creepage distance. These are typically made of glass or ceramic and are called insulators. Power grids usually include a large number of insulators to ensure their normal operation. The insulators to be inspected are those about to undergo insulation degradation testing.

[0041] In order to determine whether the insulator to be tested is a deteriorated insulator and to locate the deteriorated insulation area in the insulator to be tested, this technical solution provides a method for detecting the insulator to be tested based on infrared images.

[0042] In order to determine whether there is a degraded insulation region in the insulator to be detected based on the infrared image containing the insulator to be detected, the infrared image needs to be convolved to obtain a difference image to be processed. For example, the difference image to be processed can be an image obtained by convolving based on a Gaussian function. It should be noted that when convolving the infrared image based on a Gaussian function, using different "Gaussian kernels" can yield images with different degrees of blurring. That is to say, the difference image to be processed includes a Gaussian image in at least one scale space. The feature points to be processed in this technical solution can be understood as feature points that exist in different scale spaces. The feature points to be processed usually refer to the feature points corresponding to the insulator to be detected in the difference image to be processed.

[0043] Specifically, a large number of insulators installed in the power grid need to be inspected regularly to promptly detect insulators with deterioration. Any insulator can be used as the insulator to be inspected. In this technical solution, the insulator to be inspected is mainly detected by infrared image detection. Based on this, after determining the insulator to be inspected, it is necessary to acquire the infrared image corresponding to the insulator to be inspected. After transforming the infrared image at different scales, a differential image to be processed is obtained. Further, feature points are extracted based on the differential image to be processed. Feature points that exist in all scales are taken as the feature points to be processed in the differential image to be processed, and at least one feature point to be processed is extracted from the differential image to be processed.

[0044] Optionally, the infrared image corresponding to the insulator to be detected is convolved to obtain a differential image to be processed, and at least one feature point to be processed is extracted from the differential image to be processed, including: taking an image of the insulator to be detected based on an infrared imaging device to obtain an infrared image containing the insulator to be detected; convolving the infrared image based on a pre-constructed Gaussian kernel function to obtain a differential image to be processed; and extracting at least one feature point to be processed from the differential image to be processed based on a feature point detection algorithm.

[0045] Infrared imaging equipment can be understood as a camera device that takes pictures of the insulators to be inspected. The infrared image obtained by the infrared imaging equipment can be a grayscale image or a color image. The Gaussian kernel function, also known as the radial basis function, is a scalar function that is radially symmetric and is used to map data of a lower dimension to a higher dimension space. Feature point detection algorithms can be understood as algorithms that detect feature points in the difference image to be processed. These algorithms extract the feature points to be processed from the difference image; for example, they can be scale-invariant feature transform (SIFT) functions.

[0046] For example, when inspecting insulators in a power grid, workers can use infrared imaging equipment to photograph the insulators. Alternatively, to save manpower and reduce the workload of workers, infrared imaging equipment can be installed next to each insulator in the power grid, with the camera angle and camera area of ​​each device pre-set so that when an insulator needs to be inspected, it can be photographed by each infrared imaging device to obtain the corresponding infrared image.

[0047] Specifically, in this technical solution, to facilitate the detection of insulators to be tested in the power grid, a detection device based on infrared imaging detection is pre-constructed. For example... Figure 2As shown, the detection device mainly consists of four parts: a power supply module, a circuit overcurrent protection module, a filtering module, and an infrared image acquisition module. The power supply module uses 380V AC power, which is stabilized at 380V by a voltage regulator, and then converted to 440V AC voltage via an intermediate step-up transformer (1:2). The circuit overcurrent protection module uses a 2mH inductor L1 and a 100pF capacitor C2 connected in series, with 1000-ohm resistors R5 and R6 connected to the ground terminal to limit the circuit current. It is also connected in series with a relay; when the circuit current far exceeds the circuit's maximum withstand threshold, the relay will disconnect, achieving overcurrent protection. The filtering module consists of a 500pF capacitor C1 and a 1kΩ resistor R1 connected in series to filter the 440V voltage generated by the power supply module. The infrared image acquisition module is mainly powered by a 380V power supply connected in series with a 500Ω resistor R7. At the same time, a 24V battery is connected to the infrared imaging device, such as an infrared imager, to acquire infrared images of the insulator to be detected, thus obtaining an infrared image containing the insulator to be detected.

[0048] Furthermore, the Gaussian kernel function can be constructed based on the following formula:

[0049] D(x,y,σ)=I(x,y)*(G(x,y.kσ)-G(x,y,σ))

[0050] Where I(x,y) represents the infrared image, x represents the horizontal coordinate of the infrared image, y represents the vertical coordinate of the infrared image, σ represents the scaling factor of the Gaussian kernel function, k represents the hyperparameter, G(x,y,σ) represents the two-dimensional Gaussian kernel function, and G(x,y.kσ) represents the K-dimensional Gaussian kernel function.

[0051] Specifically, G(x,y,σ) can be represented by the following formula:

[0052]

[0053] Where x represents the horizontal coordinate in the infrared image, y represents the vertical coordinate in the infrared image, and σ represents the scaling factor of the Gaussian kernel function.

[0054] The Gaussian kernel function constructed based on the above formula is used to convolve the infrared image containing the insulator to be detected to obtain the difference image to be processed. Furthermore, by using the feature point detection algorithm to extract feature points from the difference image to be processed, at least one feature point in the difference image to be processed corresponding to the insulator to be detected can be obtained.

[0055] S120. Determine the pixel association information corresponding to each feature point to be processed, and determine the feature vector to be compared corresponding to the feature point to be processed based on the pixel association information.

[0056] The pixel association information includes at least one associated pixel in the neighborhood of the corresponding feature point to be processed, as well as the distance and angle to be determined for each associated pixel relative to the feature point to be processed.

[0057] In practical applications, after extracting each feature point to be processed, the corresponding feature vector to be compared is determined.

[0058] Specifically, taking one of the feature points to be processed as an example, when determining the feature vector to be compared for the feature point to be processed, it is first necessary to obtain at least one associated pixel point related to the feature point to be processed, and determine the distance and angle to be determined corresponding to the feature point to be processed based on each associated pixel point.

[0059] It is understandable that in the difference image to be processed, each pixel adjacent to the feature point to be processed can be taken as a related pixel, and the geometric distance between each related pixel and the feature point to be processed can be calculated. Based on each geometric distance, the spatial distance of the corresponding related pixel relative to the feature point to be processed can be determined, that is, the distance to be determined.

[0060] Simultaneously, when processing the feature points to be processed, the orientation parameter information of the feature points can also be determined. Specifically, the processing region corresponding to the feature point can be pre-determined, and weighted processing is performed based on the orientation information of each pixel within the processing region to determine the orientation parameter information of the feature point. The orientation parameter information of the feature point includes at least one primary direction and at least one secondary direction. Further, the orientation parameter information of the feature point determines the angle to be determined corresponding to the feature point.

[0061] Based on this, the feature vector to be compared is calculated according to the distance and angle to be determined of the feature point to be processed.

[0062] S130. Based on the comparison vector of each feature point to be processed and the corresponding target feature vector, determine the error to be determined for each feature point to be processed.

[0063] The target feature vector is a pre-determined vector used as a reference vector for the comparison vector of the feature point to be processed. Specifically, the target feature vector can be the feature vector corresponding to the feature point in the label image. Similar to the feature point to be processed, the target feature vector of the feature point in the label image is also determined by the corresponding feature point, as well as the distance and angle information between the corresponding pixels. The error to be determined can be understood as the vector error determined from the comparison vector and the target vector corresponding to the same feature point to be processed.

[0064] Specifically, before inspecting the insulators to be tested, a large number of insulator sample images need to be collected to accurately identify whether there are degraded insulation areas. When training a pre-built model, such as a neural network model or an image recognition model, insulator sample images with degraded insulation areas of different types, locations, and sizes are usually provided as label images. By training and recognizing a large number of insulator sample images, the goal of accurately identifying degraded insulation areas in the insulators to be tested can be achieved. Accordingly, each label image contains at least one feature point of the insulator, and the feature vector of each feature point in the label image is a pre-determined target feature vector. Therefore, when processing each feature point to be processed in the difference image, a key point matching algorithm can be used to determine the feature point corresponding to the feature point to be processed in the label image. Then, based on the target feature vector corresponding to the feature point and the comparison vector of the feature point to be processed, the error to be determined corresponding to the feature point to be processed is calculated.

[0065] Optionally, based on the comparison vector of each feature point to be processed and the corresponding target feature vector, the determination of the determination error corresponding to the feature point to be processed is determined, including: acquiring the tag image corresponding to the insulator to be detected, and determining the matched feature points of each feature point to be processed in the tag image; and determining the determination error of the corresponding feature point to be processed based on the comparison vector of each feature point to be processed and the target feature vector of the corresponding matched feature points.

[0066] Among them, the matched feature points can be understood as feature points in the label image that match the feature points to be processed. For example, after determining the feature points to be processed, the matched feature points corresponding to each feature point to be processed are detected in the label image based on key point detection technology.

[0067] For example, the number of matching vectors corresponding to each feature point to be processed corresponds to the number of associated pixels. For instance, if there are 8 associated pixels in the neighborhood of a feature point to be processed, then there are 8 distances to be determined for that feature point. In this technical solution, the direction information of the feature point to be processed is set to 8 specified directions: east, south, west, north, southeast, southwest, northwest, and northeast. Accordingly, based on the direction information of the feature point to be processed, 8 angles to be determined corresponding to the feature point can be obtained. Further, based on the 8 distances to be determined and the angles to be determined, 8 matching vectors corresponding to the feature point to be processed can be obtained. Accordingly, in the label image, the number of target feature vectors of the matched feature points corresponding to the feature point to be processed is 8.

[0068] It should be noted that, taking the current feature point as an example, when calculating the undetermined error corresponding to the current feature point based on each undetermined distance and each undetermined angle, it is necessary to establish the correspondence between each undetermined distance and each undetermined angle. Specifically, each associated pixel point corresponds to an undetermined distance, and the undetermined distances are sorted from near to far based on the distance between each associated pixel point and the current feature point. Similarly, when sorting the undetermined angles, since the direction information of the current feature point is obtained by weighting the direction information of the pixels in the processing area, after weighting the direction of the current feature point, the lengths of the direction vectors of the current feature point in the eight directions are different. The shortest direction vector is determined as the reference direction vector of the current feature point, and the undetermined angle is calculated by rotating counterclockwise based on the reference direction vector and forming the undetermined angle between each pair of adjacent direction vectors.

[0069] Furthermore, taking one of the feature points to be processed as the current feature point as an example, the alignment vector of the current feature point in the difference image A to be processed is v. A Based on the target feature vector v in the pre-set label image B B The error to be determined for the current feature point can be determined using the following function:

[0070]

[0071] Where j represents the total number of feature points to be compared corresponding to the current feature point, E j v represents the undetermined error of the j-th associated pixel corresponding to the current feature point. jA v represents the alignment vector between the current feature point and the image to be processed, determined based on the j-th associated pixel. jB This represents the target feature vector in the labeled image that corresponds to the matched feature point, determined based on the j-th associated pixel.

[0072] S140. Based on the target feature points whose error to be determined is less than the preset error, determine the deteriorated insulation area in the insulator to be tested.

[0073] To determine whether an insulator under test has deteriorated, a preset error can be set based on the error information of the feature points of the insulator under normal conditions. When the error to be determined for a feature point is less than the preset error, the corresponding area on the insulator to be tested for that feature point can be considered a deteriorated insulation area. The target feature point is a feature point among at least one feature point to be processed.

[0074] Understandably, there can be multiple feature points to be processed in the differential image. When the insulator to be detected deteriorates, the existence of a deteriorated insulation region can be determined based on the undetermined error corresponding to the feature point. In other words, for all feature points to be processed, if the undetermined error corresponding to the feature point is less than a preset error, the local area of ​​the insulator corresponding to the corresponding feature point can be determined as a deteriorated insulation region.

[0075] The advantage of this setup is that it allows for the detection of degraded insulation areas in the insulator based on whether the errors corresponding to the feature points to be processed meet preset error detection conditions. Furthermore, during the detection process, the errors corresponding to the feature points are unaffected by external conditions such as temperature; therefore, the detection results of degraded insulation areas based on the errors of each feature point are more accurate.

[0076] The technical solution of this embodiment involves convolving the infrared image corresponding to the insulator to be detected to obtain a differential image to be processed, and extracting at least one feature point to be processed from the differential image. The infrared image of the insulator to be detected is acquired using an infrared imaging device, and convolved with a Gaussian kernel function to obtain the differential image to be processed. This yields feature points that exist in all Gaussian scale spaces, allowing for the extraction of at least one feature point to be processed from the differential image based on feature point detection technology. Furthermore, pixel association information corresponding to each feature point to be processed is determined, and the corresponding feature vector to be compared is determined based on this pixel association information. The distance to be determined between each feature point to be processed and its associated pixels is calculated, and the corresponding angle to be determined is obtained based on the direction vector information of each feature point to be processed. Finally, the comparison vector is obtained based on the distances to be determined and the corresponding angles to be determined. Furthermore, based on the comparison vector and corresponding target feature vector of each feature point to be processed, the undetermined error corresponding to the feature point to be processed is determined. Matched feature points corresponding to each feature point to be processed are determined from the pre-trained label image, and the undetermined error corresponding to the feature point to be processed is calculated based on the target feature vector corresponding to each matched feature point and the corresponding comparison vector. Further, based on target feature points where the undetermined error is less than a preset error, the degraded insulation region in the insulator to be detected is determined. In practical applications, a preset error corresponding to each feature point to be processed has been pre-set. If the undetermined error is less than the preset error, it indicates that the insulator region corresponding to the corresponding feature point is a degraded insulation region. This solves the problems of low detection efficiency and inaccurate detection results when using manual detection methods for insulators, or the significant environmental influences when using temperature information from infrared images for detection. It achieves the effect of accurately detecting degraded insulation regions in insulators based on infrared images containing insulators, without being affected by temperature conditions, while also improving detection efficiency and reducing the workload of personnel.

[0077] Example 2

[0078] Figure 3 The flowchart of an insulator detection method provided in Embodiment 2 of the present invention includes, optionally, determining the pixel association information corresponding to each feature point to be processed, and refining the feature vector to be compared corresponding to the feature point to be processed based on the pixel association information.

[0079] like Figure 3 As shown, the method includes:

[0080] S210. Perform convolution processing on the infrared image corresponding to the insulator to be detected to obtain the differential image to be processed, and extract at least one feature point to be processed from the differential image to be processed.

[0081] S220. For each feature point to be processed, determine at least one associated pixel in the neighborhood of the current feature point, and determine the distance to be determined between each associated pixel and the current feature point based on the distance function.

[0082] The distance determination function can be understood as a function used to determine the geometric distance between the associated pixel and the current feature point. The distance to be determined here refers to the geometric distance between the associated pixel and the current feature point.

[0083] For example, the current feature point is taken as the neighborhood center, and the 8 related pixels (x1, y1), (x2, y2) ... (x8, y8) closest to the neighborhood center in the difference image to be processed are determined, and the geometric distances d1, d2, ... d8 between each related pixel and the neighborhood center are determined respectively.

[0084] The geometric distance d can be calculated based on the following formula:

[0085] d i =((xx) i ) 2 +(yy i ) 2 ) 1 / 2

[0086] Where, d i Let x represent the geometric distance between the i-th associated pixel and the neighborhood center, x represent the x-coordinate of the neighborhood center in the difference image to be processed, and y represent the y-coordinate of the neighborhood center in the difference image to be processed. i Let y represent the x-coordinate of the i-th associated pixel in the difference image to be processed. i This represents the ordinate of the i-th associated pixel in the difference image to be processed.

[0087] S230. Determine the region to be processed corresponding to the current feature point, and perform weighted processing on the direction parameters of each pixel to be processed within the region to obtain a group of direction vectors to be processed corresponding to the current feature point.

[0088] In this technical solution, when determining the angle to be determined for the current feature point, the processing region corresponding to the current feature point must first be determined. The processing direction vector group includes a primary direction vector and at least one secondary direction vector, and among the at least one secondary direction vector, the shortest direction vector is determined as the reference direction vector for the current feature point.

[0089] It should be noted that by weighting the directional information of each pixel within the processing area, the directional vectors of the current feature point in each direction can be obtained. Each directional vector has a corresponding vector length. The directional vector with the longest length is taken as the principal directional vector of the current feature point, and directional vectors whose length is greater than 80% of the principal directional vector are taken as auxiliary directional vectors. Here, 80% is an example and does not represent an actual proportion. That is, after determining the principal directional vectors corresponding to the current feature point from 8 directions, the directional vectors in the remaining 7 directions can all be used as auxiliary directional vectors, provided their lengths reach a preset proportion of the principal directional vectors. Furthermore, by comparing the auxiliary directional vectors, the shortest auxiliary directional vector is taken as the reference directional vector of the current feature point.

[0090] Optionally, a processing region corresponding to the current feature point is determined, and the orientation parameters of each pixel in the processing region are weighted to obtain a processing orientation vector group corresponding to the current feature point. This includes: determining the radius to be used based on the scale transformation factor in the Gaussian kernel function, and constructing the processing region of the current feature point based on the radius to be used; wherein the center point of the processing region is the current feature point; and weighting at least one pixel in the processing region based on the Gaussian weighting algorithm to obtain the processing orientation vector group of the current feature point.

[0091] For example, in this technical solution, such as Figure 4 As shown, the region to be processed can be a circular template based on the radius to be used. This template is divided into four sector regions, and the radius is set according to the scaling factor of the Gaussian kernel function, such as 3σ. The region to be processed contains a large number of pixels, each containing corresponding directional information. Without processing, the direction of each pixel can correspond to any direction within 360°. To reduce the amount of directional information described, eight directions can be pre-set, and the directional information of each pixel in the region to be processed is weighted using a Gaussian weighted algorithm. The final weighted result serves as the directional feature description of the current feature point. The group of directional vectors corresponding to the current feature point includes the direction vectors of the current feature point in each direction.

[0092] The advantage of this setup is that it reduces the number of directional feature vectors corresponding to the feature points to be processed to eight, which greatly reduces the amount of computation and improves the matching speed between feature points. At the same time, it also reduces the vector dimension of the comparison vectors of the feature points to be processed, avoids excessive feature selection, and improves the accuracy of feature point matching.

[0093] S240. Determine the angle to be determined for the current feature point based on each direction vector in the direction vector group to be processed.

[0094] In practical applications, the angle to be determined between each pair of adjacent direction vectors in the group of direction vectors to be processed for the current feature point is determined starting from the reference direction vector and rotating counterclockwise. For each angle to be determined, the angle to be determined corresponding to the current angle is determined based on the vector association information corresponding to the current angle to be determined.

[0095] Among them, the vector association information is the vector length information corresponding to the direction vector associated with the included angle to be determined.

[0096] For example, if the direction vector group to be used contains one main direction vector and five auxiliary direction vectors, see [reference needed]. Figure 5 The direction vector formed between the pixel to be processed (x3, y3) and the current feature point (x, y) is the main direction vector. The auxiliary direction vectors corresponding to the current feature point (x, y) are the other pixel to be processed (x1, y1), (x2, y2), (x4, y4), (x5, y5), and (x6, y6). Among the auxiliary direction vectors, the shortest direction vector is taken as the reference vector, i.e., (x1, y1).

[0097] Correspondingly, the angles to be determined between each pair of adjacent direction vectors are obtained by rotating the reference direction vector counterclockwise, namely θ1, θ2, θ3, θ4, θ5, and θ6. These angles can be understood as the angular information corresponding to each angle; for example, the angle to be determined corresponding to θ1 is 30°.

[0098] Specifically, when determining the angle to be determined corresponding to each included angle, the following formula can be used:

[0099]

[0100] Where, θ i This represents the angle to be determined corresponding to the i-th angle to be determined. This represents the vector length between the i-th direction vectors; This represents the vector length between the (i+1)th direction vectors. This represents the length of the vector between the i-th direction vector and the (i+1)-th direction vector.

[0101] It should be noted that the i-th direction vector is a direction vector that is rotated counterclockwise based on the reference direction vector.

[0102] S250. Based on the distance and angle to be determined corresponding to the current feature point, determine the comparison vector corresponding to the current feature point.

[0103] Specifically, for each distance and angle to be determined for the current feature point, taking the current feature point as an example, at least one distance to be determined corresponding to the current feature point is normalized to obtain the corresponding first parameter to be used; at least one angle to be determined corresponding to the current feature point is normalized to obtain the corresponding second parameter to be used; based on each first parameter to be used and the corresponding first weight, and each second parameter to be used and the corresponding second weight, the comparison vector corresponding to the current feature point is obtained.

[0104] In this technical solution, in order to facilitate the determination of the comparison vector corresponding to the current feature point based on the distance and angle to be determined, it is necessary to normalize each distance and angle to be determined.

[0105] Taking the determination of the alignment vector based on one set of distances and angles to be determined as an example, the normalization of the distances to be determined can be processed based on the following formula:

[0106]

[0107] Where, r i d represents the first parameter to be used corresponding to the i-th distance to be determined. i This represents the distance to be determined between the i-th associated pixel and the current feature point.

[0108] For example, d1 represents the distance to be determined between the first associated pixel and the current feature point, and d2 represents the distance to be determined between the second associated pixel and the current feature point.

[0109] When normalizing a given angle, the following formula can be used:

[0110]

[0111] Among them, a i θ represents the second parameter to be used corresponding to the i-th angle to be determined. i This represents the i-th angle to be determined.

[0112] Furthermore, by processing the distance and angle to be determined using the following formula, the comparison vector can be obtained:

[0113] v i =w i r i +ω i a i

[0114] Among them, v i Let r represent the i-th vector to be compared. iw represents the first parameter to be used corresponding to the i-th distance to be determined. i a represents the first weight to be used corresponding to the i-th first parameter to be used. i ω represents the second parameter to be used corresponding to the i-th angle to be determined. i This represents the second weight to be used corresponding to the i-th second parameter to be used.

[0115] S260. Based on the comparison vector of each feature point to be processed and the corresponding target feature vector, determine the error to be determined for each feature point to be processed.

[0116] S270. Based on the target feature points whose error to be determined is less than the preset error, determine the deteriorated insulation area in the insulator to be tested.

[0117] The technical solution of this embodiment, for each feature point to be processed, determines at least one associated pixel in the neighborhood of the current feature point, and determines the distance to be determined between each associated pixel and the current feature point based on a distance function. Pixels adjacent to the feature point in the difference image to be processed are identified as associated pixels, and the distance to be determined between each associated pixel and the feature point to be processed is calculated. Simultaneously, a region to be processed corresponding to the current feature point is determined, and the direction parameters of each pixel to be processed within the region are weighted to obtain a group of direction vectors to be processed corresponding to the current feature point. A circular template with a radius determined by the scaling factor of the Gaussian kernel function is used to determine the region to be processed corresponding to each feature point. Based on the direction information of all pixels within the region to be processed, a group of direction vectors to be processed for the corresponding feature point is determined, wherein the group of direction vectors to be processed includes a main direction vector and at least one auxiliary direction vector. Further, based on each direction vector in the group of direction vectors to be processed, the angle to be determined for the current feature point is determined, and based on the distance to be determined and the angle to be determined for the current feature point, a comparison vector corresponding to the current feature point is determined. The advantage of this setup is that it allows for more accurate calculation of the corresponding comparison vector based on the distance and angle to be determined for each feature point, without needing to consider environmental factors such as temperature. This makes the determination error obtained based on the comparison vector and the corresponding target feature vector more accurate, which is beneficial for detecting degraded insulation areas in the insulator. Furthermore, when a degraded insulation area is detected in the insulator, it can be located based on the coordinate information of the corresponding feature points to be processed, facilitating subsequent treatment of the degraded insulator.

[0118] Example 3

[0119] Figure 6 This is a schematic diagram of an insulator testing device provided in Embodiment 3 of the present invention. Figure 6 As shown, the device includes: a feature point determination module 310, a feature vector determination module 320, an error determination module 330, and a degraded insulation region determination module 340.

[0120] The feature point determination module 310 is used to perform convolution processing on the infrared image corresponding to the insulator to be detected, to obtain the differential image to be processed, and to extract at least one feature point to be processed from the differential image to be processed.

[0121] The feature vector determination module 320 is used to determine the pixel association information corresponding to each feature point to be processed, and to determine the feature vector to be compared corresponding to the feature point to be processed based on the pixel association information; wherein, the pixel association information includes at least one associated pixel in the neighborhood of the corresponding feature point to be processed, as well as the distance and angle to be determined of each associated pixel relative to the feature point to be processed.

[0122] The undetermined error determination module 330 is used to determine the undetermined error corresponding to the corresponding feature point based on the comparison vector of each feature point to be processed and the corresponding target feature vector; wherein, the target feature vector is a pre-determined vector;

[0123] The deteriorated insulation region determination module 340 is used to determine the deteriorated insulation region in the insulator to be tested based on the target feature points whose error to be determined is less than the preset error.

[0124] The target feature point is a feature point among at least one feature point to be processed.

[0125] The technical solution of this invention involves convolution processing of the infrared image corresponding to the insulator to be tested to obtain a differential image to be processed, and extracting at least one feature point to be processed from the differential image. It then determines the pixel association information corresponding to each feature point to be processed, and determines the corresponding comparison feature vector based on the pixel association information. Based on the comparison vector of each feature point to be processed and the corresponding target feature vector, it determines the error to be determined for each feature point to be processed. Finally, based on target feature points where the error to be determined is less than a preset error, it determines the degraded insulation region in the insulator to be tested. This solves the problems of low detection efficiency and inaccurate results when manually inspecting insulators, or the significant environmental influences when using temperature information from infrared images. It achieves accurate detection of degraded insulation regions in insulators based on infrared images containing the insulator, without being affected by temperature conditions, while also improving detection efficiency and reducing the workload of personnel.

[0126] Optionally, the feature point determination module includes: an infrared image determination submodule, used to take a picture of the insulator to be detected based on an infrared imaging device to obtain an infrared image containing the insulator to be detected;

[0127] The differential image determination submodule is used to perform convolution processing on the infrared image based on a pre-built Gaussian kernel function to obtain the differential image to be processed;

[0128] The feature point determination submodule is used to extract at least one feature point from the difference image to be processed based on the feature point detection algorithm.

[0129] Optionally, the feature vector determination module includes: a distance determination submodule, used to determine at least one associated pixel in the neighborhood of the current feature point for each feature point to be processed, and to determine the distance to be determined between each associated pixel and the current feature point based on a distance function;

[0130] The direction vector group determination submodule is used to determine the region to be processed corresponding to the current feature point, and to perform weighted processing on the direction parameters of each pixel to be processed within the region to obtain the direction vector group to be processed corresponding to the current feature point; wherein, the direction vector group to be processed includes a main direction vector and at least one auxiliary direction vector, and the shortest direction vector among the at least one auxiliary direction vector is the reference direction vector of the current feature point;

[0131] The undetermined angle determination submodule is used to determine the undetermined angle of the current feature point based on each direction vector in the unprocessed direction vector group;

[0132] The submodule for determining the alignment vector is used to determine the alignment vector corresponding to the current feature point based on the distance and angle to be determined for the current feature point.

[0133] Optionally, the direction vector group determination submodule includes: a region to be processed determination unit, used to determine the radius to be used based on the scaling factor in the Gaussian kernel function, and to construct the region to be processed for the current feature point based on the radius to be used; wherein, the center point of the region to be processed is the current feature point;

[0134] The direction vector group determination unit is used to perform weighted processing on at least one pixel point to be processed within the processing area based on the Gaussian weighted algorithm to obtain the direction vector group to be used for the current feature point.

[0135] Optionally, the submodule for determining the angle to be determined includes: a unit for determining the included angle, used to determine the included angle between every two adjacent direction vectors in the group of direction vectors to be processed for the current feature point, starting from the reference direction vector and rotating counterclockwise;

[0136] The angle determination unit is used to determine the angle to be determined for each angle to be determined based on the vector association information corresponding to the current angle to be determined; wherein, the vector association information includes two direction vectors adjacent to the angle to be determined, and the vector formed between the two adjacent direction vectors.

[0137] Optionally, the submodule for determining the vector to be compared includes: a first parameter to be used determination unit, used to normalize at least one distance to be determined corresponding to the current feature point for each feature point to be processed, so as to obtain the corresponding first parameter to be used;

[0138] The second parameter determination unit is used to normalize at least one angle to be determined corresponding to the current feature point to obtain the corresponding second parameter to be used.

[0139] The comparison vector determination unit is used to obtain the comparison vector corresponding to the current feature point based on each first parameter to be used and the corresponding first weight, and each second parameter to be used and the corresponding second weight.

[0140] Optionally, the error determination module includes: a matched feature point determination submodule, used to acquire the tag image corresponding to the insulator to be detected, and determine the matched feature points of each feature point to be processed in the tag image;

[0141] The undetermined error determination submodule is used to determine the undetermined error of the corresponding feature point based on the comparison vector of each feature point to be processed and the target feature vector of the corresponding matched feature points.

[0142] The insulator testing device provided in this embodiment of the invention can execute the insulator testing method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0143] Example 4

[0144] Figure 7 A schematic diagram of the structure of an electronic device 10 according to an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0145] like Figure 7As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0146] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0147] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as insulator detection methods.

[0148] In some embodiments, the insulator detection method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the insulator detection method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the insulator detection method by any other suitable means (e.g., by means of firmware).

[0149] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0150] Computer programs for implementing the insulator detection method of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer programs can be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0151] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0152] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0153] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0154] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0155] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0156] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method of insulator detection, characterized by, The method comprises the following steps: performing convolution processing on an infrared image corresponding to the to-be-detected insulator to obtain a to-be-processed difference image, and extracting at least one to-be-processed feature point in the to-be-processed difference image; determining pixel point association information corresponding to each to-be-processed feature point, and determining a to-be-compared feature vector corresponding to the corresponding to-be-processed feature point based on the pixel point association information; wherein the pixel point association information comprises at least one associated pixel point in the neighborhood of the corresponding to-be-processed feature point, and a to-be-determined distance and a to-be-determined angle of each associated pixel point relative to the to-be-processed feature point; determining a to-be-determined error corresponding to the corresponding to-be-processed feature point according to the to-be-compared vector of each to-be-processed feature point and a corresponding target feature vector; wherein the target feature vector is a predetermined vector; determining a degraded insulating region in the to-be-detected insulator according to the target feature point with a to-be-determined error less than a preset error; wherein the target feature point is a feature point in the at least one to-be-processed feature point.

2. The method of claim 1, wherein, The method comprises the following steps: performing convolution processing on an infrared image corresponding to the to-be-detected insulator to obtain a to-be-processed difference image, and extracting at least one to-be-processed feature point in the to-be-processed difference image; capturing the to-be-detected insulator based on an infrared imaging device to obtain an infrared image containing the to-be-detected insulator; performing convolution processing on the infrared image based on a pre-constructed Gaussian kernel function to obtain a to-be-processed difference image; 3. The method of claim 1, wherein, extracting at least one to-be-processed feature point in the to-be-processed difference image based on a feature point detection algorithm. The method comprises the following steps: for each to-be-processed feature point, determining at least one associated pixel point in the neighborhood of the current feature point, and determining a to-be-determined distance between each associated pixel point and the current feature point based on a distance function; determining a to-be-processed region corresponding to the current feature point, and performing weighted processing on the direction parameters of each to-be-processed pixel point in the to-be-processed region to obtain a to-be-processed direction vector group corresponding to the current feature point; wherein the to-be-processed direction vector group comprises a main direction vector and at least one auxiliary direction vector, and the shortest direction vector in the at least one auxiliary direction vector is the reference direction vector of the current feature point; determining a to-be-determined angle of the current feature point according to each direction vector in the to-be-processed direction vector group; 4. The method of claim 3, wherein, determining a to-be-compared vector corresponding to the current feature point based on the to-be-determined distance and the to-be-determined angle corresponding to the current feature point. The method comprises the following steps: determining a to-be-used radius according to a scale transformation factor in the Gaussian kernel function, and constructing a to-be-processed region of the current feature point according to the to-be-used radius; wherein the center point of the to-be-processed region is the current feature point; Based on the Gaussian weighted algorithm, at least one pixel in the region to be processed is weighted to obtain the direction vector group to be used for the current feature point.

5. The method of claim 3, wherein, The step of determining the angle to be determined for the current feature point based on each direction vector in the group of direction vectors to be processed includes: In the group of direction vectors to be processed for the current feature point, the angle to be determined between every two adjacent direction vectors starting from the reference direction vector and rotating counterclockwise; For each angle to be determined, the angle to be determined corresponding to the current angle to be determined is determined based on the vector association information corresponding to the current angle to be determined.

6. The method of claim 3, wherein, The step of determining the comparison vector corresponding to the current feature point based on the distance and angle to be determined corresponding to the current feature point includes: Normalize at least one distance to be determined corresponding to the current feature point to obtain the corresponding first parameter to be used; Normalize at least one angle to be determined corresponding to the current feature point to obtain the corresponding second parameter to be used. Based on each of the first parameters to be used and the corresponding first weight, and each of the second parameters to be used and the corresponding second weight, a comparison vector corresponding to the current feature point is obtained.

7. The method of claim 1, wherein, The step of determining the undetermined error corresponding to the respective feature point to be processed based on the comparison vector of each feature point to be processed and the corresponding target feature vector includes: Acquire a tag image corresponding to the insulator to be detected, and determine the matched feature points of each feature point to be processed in the tag image; Based on the comparison vector of each feature point to be processed and the target feature vector of the corresponding matched feature points, the determination error of the corresponding feature point to be processed is determined.

8. An insulator detection device, characterized by include: The feature point determination module is used to perform convolution processing on the infrared image corresponding to the insulator to be detected to obtain a differential image to be processed, and to extract at least one feature point to be processed from the differential image to be processed. The feature vector determination module is used to determine the pixel association information corresponding to each feature point to be processed, and to determine the feature vector to be compared corresponding to the feature point to be processed based on the pixel association information; wherein, the pixel association information includes at least one associated pixel in the neighborhood of the corresponding feature point to be processed, and the distance and angle to be determined of each associated pixel relative to the feature point to be processed. The undetermined error determination module is used to determine the undetermined error corresponding to the corresponding feature point to be processed based on the comparison vector of each feature point to be processed and the corresponding target feature vector; wherein, the target feature vector is a predetermined vector; The deteriorated insulation region determination module is used to determine the deteriorated insulation region in the insulator to be tested based on the target feature points whose error to be determined is less than a preset error. The target feature point is a feature point among the at least one feature point to be processed.

9. An electronic device, comprising: The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the insulator detection method in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for enabling the processor to implement the insulator detection method in any one of claims 1-7 when executed.

Images