Insulator contamination level assessment method based on infrared images

Abstract

The application provides an insulator contamination grade evaluation method based on an infrared image, and comprises the following steps: S1, acquiring an infrared image of a sample insulator by using an infrared imaging device, wherein the image is acquired under different temperatures, different humidities and different contamination grades; S2, preprocessing the infrared image; S3, extracting a temperature characteristic parameter from the preprocessed image; S4, constructing a probabilistic neural network, inputting the temperature characteristic parameter, the environmental temperature and the environmental humidity into the probabilistic neural network as input features, and training the probabilistic neural network; S4, acquiring an infrared image of a to-be-tested insulator in real time, inputting the temperature characteristic parameter of the to-be-tested insulator and the environmental temperature and the environmental humidity of the to-be-tested insulator into the trained probabilistic neural network after processing through steps S2 and S3, and outputting a contamination grade of the to-be-tested insulator.

Description

Technical Field

[0001] This invention relates to a method for evaluating power equipment, and more particularly to a method for evaluating the pollution level of insulators based on infrared images. Background Technology

[0002] Insulators operate outdoors year-round, and dust particles in the air continuously accumulate on their surfaces, eventually forming a pollution layer. This can easily lead to flashover accidents during operation, severely affecting the stability of the power system.

[0003] To reduce the occurrence of insulator flashover accidents, it is necessary to decontaminate insulators in operation. Therefore, before decontamination, it is necessary to assess the degree of contamination on the insulators to provide accurate contamination information for the decontamination process. In the existing technology, the main methods for assessing the contamination degree of insulators include spark gap method, leakage current detection method, pulse current detection method, distributed voltage method, and ultraviolet imaging method. These assessment methods are all contact-based detection methods, which have low efficiency and high cost. More importantly, these methods are often susceptible to electromagnetic interference, resulting in low accuracy of the assessment.

[0004] Therefore, in order to solve the above-mentioned technical problems, it is urgent to propose a new technical approach. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a method for assessing the pollution level of insulators based on infrared images. This method acquires infrared images of insulators using infrared equipment and utilizes the characteristic that the operating temperature of insulators changes with the degree of pollution to construct a corresponding probabilistic neural network. The probabilistic neural network is trained based on the temperature characteristics of the insulators obtained from the infrared images, as well as ambient temperature, to obtain the pollution level of the insulators. This method effectively improves monitoring efficiency and reduces monitoring costs. Furthermore, it takes into account the influence of ambient temperature and humidity during the assessment process, thereby effectively ensuring the accuracy of the final results.

[0006] This invention provides a method for assessing the pollution level of insulators based on infrared images, comprising the following steps:

[0007] S1. Infrared images of the sample insulators are acquired using infrared imaging equipment, wherein the acquisition of images is carried out under different temperature, different humidity and different pollution levels.

[0008] S2. Preprocess the infrared image;

[0009] S3. Extract temperature feature parameters from the preprocessed image;

[0010] S4. Construct a probabilistic neural network and input the temperature feature parameters, ambient temperature, and ambient humidity as input features into the probabilistic neural network to train the probabilistic neural network;

[0011] S4. Acquire the infrared image of the insulator under test in real time, and after processing through steps S2 and S3, input the temperature characteristic parameters of the insulator under test, along with the ambient temperature and humidity of the insulator under test, into the trained probabilistic neural network to output the pollution level of the insulator under test.

[0012] Furthermore, in step S2, the preprocessing of the infrared image specifically includes:

[0013] S21. Convert the infrared image to the HSV color space and extract the S component information;

[0014] S22. Perform bilateral filtering on the S-component information;

[0015] S23. The S-component information processed in step S22 is reconstructed using a morphological reconstruction opening and closing operation algorithm to obtain a reconstructed image;

[0016] S24. Calculate the local maxima of the reconstructed image and use the local maxima as the foreground markers of the reconstructed image;

[0017] The reconstructed image is converted into a binary image, and a distance transformation is performed on the binary image to obtain the vector distance matrix of the binary image. The vector distance matrix is ​​then subjected to a watershed transformation to obtain the boundary line between the foreground and background regions of the reconstructed image, and the boundary line is used as the background marker.

[0018] S25. Convert the infrared image into the RGB color space, and use the Sobel operator to calculate the gradients of the R component in the RGB color space and the S component in the HSV color space respectively.

[0019] S26. Compare the gradient values ​​of the corresponding pixels of the R component and the S component, select the larger gradient value of the two components as the gradient value of the current pixel, and synthesize the RS component gradient map.

[0020] S27. Use the imimposemin function to correct the RS component gradient map so that the corrected RS component gradient map has a minimum value at the foreground and background markers, thus obtaining the corrected RS component gradient map.

[0021] A watershed transform is performed on the corrected RS component gradient map to segment the insulator region in the infrared image.

[0022] Furthermore, step S3 specifically includes:

[0023] S31: Determine the target region from the segmented insulator regions and convert the target region into a binary matrix;

[0024] S32. Extract the temperature matrix of the insulator region from the infrared image, and the temperature matrix and the binary matrix have the same order;

[0025] S33. Perform a dot product between the binary matrix and the temperature matrix to obtain the temperature rise matrix of the target region, and extract temperature feature parameters from the temperature rise matrix, including: maximum value, minimum value, average value, median value, mode, root mean square, range, variance, standard deviation and kurtosis.

[0026] The beneficial effects of this invention are as follows: This invention acquires infrared images of insulators using infrared equipment and utilizes the characteristic that the operating temperature of insulators changes with the degree of pollution to construct a corresponding probabilistic neural network. The probabilistic neural network is trained based on the temperature characteristics of the insulators and the ambient temperature obtained from the infrared images of the insulators, and then the degree of pollution of the insulators is obtained. On the one hand, this can effectively improve monitoring efficiency and reduce monitoring costs. On the other hand, the influence of ambient temperature and humidity is taken into account during the evaluation process, thereby effectively ensuring the accuracy of the final result. Attached Figure Description

[0027] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0028] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0029] The present invention will be further described in detail below:

[0030] This invention provides a method for assessing the pollution level of insulators based on infrared images, comprising the following steps:

[0031] S1. Infrared images of the sample insulators are acquired using infrared imaging equipment, wherein the acquisition of images is carried out under different temperature, different humidity and different pollution levels.

[0032] S2. Preprocess the infrared image;

[0033] S3. Extract temperature feature parameters from the preprocessed image;

[0034] S4. Construct a probabilistic neural network and input the temperature feature parameters, ambient temperature, and ambient humidity into the probabilistic neural network to train it. These input parameters need to be normalized and then formed into a matrix before being input into the probabilistic neural network.

[0035] S4. Infrared images of the insulator under test are acquired in real time. After processing through steps S2 and S3, the temperature characteristic parameters of the insulator, along with the ambient temperature and humidity, are input into the trained probabilistic neural network to output the pollution level of the insulator. This method acquires infrared images of the insulator using infrared equipment and utilizes the characteristic that the insulator's operating temperature changes with pollution level to construct a corresponding probabilistic neural network. The network is trained based on the insulator's temperature characteristics obtained from the infrared images and the ambient temperature to obtain the pollution level. This approach effectively improves monitoring efficiency, reduces monitoring costs, and ensures the accuracy of the final results by considering the influence of ambient temperature and humidity during the evaluation process.

[0036] In the above, the sample insulators are set up in a laboratory. A specific type of insulator is selected as the sample insulator, such as a double-umbrella porcelain insulator. Then, different ambient temperatures, humidity levels, and pollution levels are simulated in the laboratory.

[0037] The ambient temperature is generally set to 5℃-25℃. This temperature range can be divided into multiple temperature intervals, and corresponding experiments are carried out in each temperature interval. Generally, dividing it into two intervals, 5℃-15℃ and 15℃-25℃, can meet the needs of sample data.

[0038] Humidity refers to relative humidity, which is generally 60% to 100%. It is generally sufficient to divide the data into three ranges to meet the needs of the sample data: 65%-75%, 75%-85%, and 85%-95%.

[0039] The pollution levels are divided into five grades: Grade 0, Grade I, Grade II, Grade III, and Grade IV. 0.03 mg / cm², 0.05 mg / cm², 0.08 mg / cm², 0.15 mg / cm², and 0.25 mg / cm² represent the equivalent salt density for each of the five pollution levels (Grade 0, Grade I, Grade II, Grade III, and Grade IV), respectively. The ratio of ash density to salt density is 6:1. Sodium chloride, diatomaceous earth, and silica are used to simulate the pollution.

[0040] A DC voltage is applied to the insulator under the above conditions to obtain several infrared images under different conditions. These images are then used to train a probabilistic neural network, which can be trained using existing technologies.

[0041] In this embodiment, step S2, the preprocessing of the infrared image specifically includes:

[0042] S21. Convert the infrared image to the HSV color space and extract the S component information;

[0043] S22. Perform bilateral filtering on the S-component information to filter out noise interference, wherein:

[0044] The formula for calculating bilateral filtering is as follows:

[0045]

[0046] Among them: u(x), u BF (x) represents the image matrices before and after bilateral filtering, respectively, w d (x, y) are filter coefficients determined by geometric spatial distance, w r (x, y) are the filter coefficients determined by the pixel gray-level difference, C d，r The normalization coefficient refers to the number of pixels in the calculation area each time.

[0047] S23. The S component information processed in step S22 is reconstructed using a morphological reconstruction opening and closing operation algorithm to obtain a reconstructed image. This step effectively removes pseudo-local extrema caused by fine textures and noise interference, preserves the complete contour information of the insulator, and ensures the accuracy of subsequent processing. The morphological reconstruction opening and closing operation algorithm is an existing technology and will not be described in detail here.

[0048] S24. Calculate the local maxima of the reconstructed image and use the local maxima as the foreground markers of the reconstructed image;

[0049] The reconstructed image is converted into a binary image, and a distance transform is performed on the binary image to obtain a vector distance matrix. A watershed transform is then applied to the vector distance matrix to obtain the boundary between the foreground and background regions of the reconstructed image, and this boundary is used as the background marker. Euclidean distance is used to perform the distance transform on the binary image, where D is the distance between each pixel and the nearest non-zero pixel. The calculation formula is as follows:

[0050]

[0051] In the formula: p(x1, y1) and q(x2, y2) are the coordinates of the foreground and background pixel values, respectively. The watershed transformation is based on existing watershed algorithms, which will not be elaborated here;

[0052] S25. Convert the infrared image into the RGB color space, and use the Sobel operator to calculate the gradients of the R component in the RGB color space and the S component in the HSV color space respectively; the Sobel operator and its gradient calculation process are existing technologies and will not be described in detail here.

[0053] S26. Compare the gradient values ​​of corresponding pixels in the R component and S component, select the larger gradient value of the two components as the gradient value of the current pixel, and synthesize the RS component gradient map; for example: the gradient of the R component and the gradient of the S component both describe pixels. For example, if the image has 3*3 pixels, that is... Then m 11 If the gradient value of the R component of a pixel is greater than the gradient value of the S component, then the gradient value of this pixel is the gradient value of the R component. If m 12 If the gradient value of the R component of a pixel is less than the gradient value of the S component, then the gradient value of this pixel is the gradient value of the S component, and so on, thus merging the R component gradient map and the S component gradient map into a single RS component gradient map.

[0054] S27. The imimposemin function is used to correct the RS component gradient map so that the corrected RS component gradient map has minimum values ​​at the foreground and background markers, thus obtaining the corrected RS component gradient map. By using the above method, the interference of pseudo-min values ​​on image segmentation can be reduced, and the accuracy of the final result can be improved.

[0055] A watershed transform is performed on the corrected RS component gradient map to segment the insulator region in the infrared image.

[0056] In this embodiment, step S3 specifically includes:

[0057] S31: Determine the target region from the segmented insulator regions and convert the target region into a binary matrix;

[0058] S32. Extract the temperature matrix of the insulator region from the infrared image, and the temperature matrix and the binary matrix have the same order;

[0059] S33. Perform a dot product between the binary matrix and the temperature matrix to obtain the temperature rise matrix of the target region, and extract temperature feature parameters from the temperature rise matrix, including: maximum value, minimum value, average value, median value, mode, root mean square, range, variance, standard deviation and kurtosis.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for assessing the pollution level of insulators based on infrared images, characterized in that: Includes the following steps: S1. Infrared images of the sample insulators are acquired using infrared imaging equipment, wherein the acquisition of images is carried out under different temperature, different humidity and different pollution levels. S2. Preprocess the infrared image; S3. Extract temperature feature parameters from the preprocessed image; S4. Construct a probabilistic neural network and input the temperature feature parameters, ambient temperature, and ambient humidity as input features into the probabilistic neural network to train the probabilistic neural network; S4. Acquire the infrared image of the insulator under test in real time, and after processing through steps S2 and S3, input the temperature characteristic parameters of the insulator under test, along with the ambient temperature and humidity of the insulator under test, into the trained probabilistic neural network to output the pollution level of the insulator under test. Step S2, the preprocessing of the infrared image specifically includes: S21. Convert the infrared image to the HSV color space and extract the S component information; S22. Perform bilateral filtering on the S-component information; S23. The S-component information processed in step S22 is reconstructed using a morphological reconstruction opening and closing operation algorithm to obtain a reconstructed image; S24. Calculate the local maxima of the reconstructed image and use the local maxima as the foreground markers of the reconstructed image; The reconstructed image is converted into a binary image, and a distance transformation is performed on the binary image to obtain the vector distance matrix of the binary image. The vector distance matrix is ​​then subjected to a watershed transformation to obtain the boundary line between the foreground and background regions of the reconstructed image, and the boundary line is used as the background marker. S25. Convert the infrared image into the RGB color space, and use the Sobel operator to calculate the gradients of the R component in the RGB color space and the S component in the HSV color space respectively. S26. Compare the gradient values ​​of the corresponding pixels of the R component and the S component, select the larger gradient value of the two components as the gradient value of the current pixel, and synthesize the RS component gradient map. S27. Use the imimposemin function to correct the RS component gradient map so that the corrected RS component gradient map has a minimum value at the foreground and background markers, thus obtaining the corrected RS component gradient map. A watershed transform is performed on the corrected RS component gradient map to segment the insulator region in the infrared image. Step S3 specifically includes: S31: Determine the target region from the segmented insulator regions and convert the target region into a binary matrix; S32. Extract the temperature matrix of the insulator region from the infrared image, and the temperature matrix and the binary matrix have the same order; S33. Perform a dot product between the binary matrix and the temperature matrix to obtain the temperature rise matrix of the target region, and extract temperature feature parameters from the temperature rise matrix, including: maximum value, minimum value, average value, median value, mode, root mean square, range, variance, standard deviation and kurtosis.

Images