Remote monitoring method of power equipment status in substation based on image analysis

Abstract

The present invention belongs to the field of image processing technology, and specifically relates to a method for remotely monitoring the status of power equipment in a substation based on image analysis. The method comprises: dividing a monitoring image of the power equipment in the substation into multiple regions of equal size, dividing each region into four neighboring image blocks and one intermediate image block, and performing denoising processing on each image block in the region using Wiener filtering. The method comprises: iteratively adjusting the side length of the blur kernel to determine the target size of the blur kernel for the neighboring image blocks by adjusting the side length of the blur kernel multiple times, including: determining the side length of the blur kernel after the next iterative adjustment based on the rate of change of the noise level and the degree of attenuation of the rate of change of the noise level during each iterative adjustment, as well as the noise level and the side length of the blur kernel of the neighboring image blocks after each iterative adjustment; and using the average of the target sizes of the blur kernels of the neighboring image blocks as the target size of the blur kernel for the intermediate image block. The present invention improves the efficiency of the entire denoising process.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and more particularly to a method for remotely monitoring the status of power equipment in a substation based on image analysis. Background Art

[0002] With the rapid development of power systems, substations, as key nodes of power transmission, are of vital importance for their safe and stable operation.

[0003] Remote status monitoring of power equipment is an important means to ensure the reliable operation of substations. By deploying various sensors and monitoring equipment in substations, the operating data and status information of power equipment can be collected in real time, realizing remote monitoring and diagnosis.

[0004] However, the quality of monitoring images has a direct impact on the accuracy of inspection results. In actual applications, due to environmental factors (such as bad weather and lighting changes), equipment performance limitations (such as equipment aging, equipment overheating, and lens distortion), and transmission interference (such as signal attenuation and electromagnetic interference), the collected monitoring images often contain noise.

[0005] These noises will reduce the clarity of the image, obscure the detailed features of the equipment, make it difficult for monitoring personnel to accurately judge the status of the equipment, and may even lead to safety hazards due to misjudgment. Therefore, improving the quality of monitoring images and effectively removing noise are of great significance to ensuring the accuracy and reliability of remote monitoring of the status of substation power equipment.

[0006] As an existing denoising method, Wiener filtering usually adopts a strategy of iteratively adjusting the blur kernel size, and the adjustment amount is the same in each iteration. Therefore, the blur kernel and denoised image need to be calculated multiple times, resulting in a huge amount of calculation, significantly increasing processing time, and affecting real-time performance and efficiency. Summary of the Invention

[0007] In order to solve the technical problem that the adjustment amount of the Wiener filter is the same in each iteration, resulting in a large amount of calculation and affecting real-time performance and efficiency, the present invention provides a remote monitoring method for the status of power equipment in a substation based on image analysis, comprising: dividing the monitoring image of the power equipment in the substation into multiple areas of the same size; dividing each area into four neighborhood image blocks of the same size and one neighborhood image block of the same size; an intermediate image block; performing denoising processing on each image block in the area by Wiener filtering: determining a target size of the blur kernel of the neighboring image block by iteratively adjusting the side length of the blur kernel multiple times, and taking the average of the target sizes of the blur kernels of the four neighboring image blocks as the target size of the blur kernel of the intermediate image block; the iteratively adjusting the side length of the blur kernel multiple times comprises: calculating the noise level of the neighboring image block after each iterative adjustment based on high-frequency coefficients in a DCT coefficient matrix obtained by discrete cosine transforming the neighboring image block after each iterative adjustment; calculating the changing speed of the noise level during each iterative adjustment based on the difference between the noise levels of the neighboring image block after each iterative adjustment and the initial neighboring image block, and the side length of the blur kernel after each iterative adjustment; calculating the attenuation speed of the changing speed of the noise level during each iterative adjustment based on the changing speed of the noise level during each iterative adjustment and the previous iterative adjustment; and determining the side length of the blur kernel after the next iterative adjustment based on the changing speed of the noise level and the attenuation degree of the changing speed of the noise level during each iterative adjustment, the noise level of the neighboring image block after each iterative adjustment, and the side length of the blur kernel.

[0008] The present invention takes the average of the target sizes of the blur kernels of the four neighboring image blocks as the target size of the blur kernel of the intermediate image block, reduces the number of image blocks that require iterative adjustment processing, predicts the optimal blur kernel size by analyzing the noise level and its changing speed and attenuation speed, reduces the number of iterations, and thus speeds up the iterative adjustment speed; in summary, the present invention significantly improves the efficiency of the entire denoising process by reducing the number of image blocks that require iterative adjustment processing and speeding up the iterative adjustment speed, making it more suitable for status monitoring scenarios of substation power equipment with high timeliness requirements.

[0009] Preferably, the feature is that the monitoring image is divided into a plurality of regions of the same size, including: dividing the monitoring image into a plurality of regions, requiring that the sizes of all regions are the same and equal to , The value range is and is an integer.

[0010] Preferably, each region is divided into 4 neighborhood image blocks of the same size and 1 neighborhood image block of the same size. The intermediate image block includes: 4 neighborhood image blocks of the same size: upper neighborhood image block, lower neighborhood image block, left neighborhood image block, right neighborhood image block, and the sizes of the upper neighborhood image block and the lower neighborhood image block are equal to , the size of the neighboring image block and the right neighboring image block is equal to .

[0011] The present invention divides the monitoring image into multiple areas, and further divides each area into four neighborhood image blocks of the same size and one intermediate image block of 5×5 size. While retaining the local features of the image, the present invention reduces the number of image blocks that need to be iteratively adjusted to determine the target size of the blur kernel, thereby reducing the amount of calculation and improving processing efficiency.

[0012] Preferably, the method for calculating the noise level of the iteratively adjusted neighborhood image block / the initial neighborhood image block is: performing discrete cosine transform on the iteratively adjusted neighborhood image block / the initial neighborhood image block to obtain a DCT coefficient matrix of the iteratively adjusted neighborhood image block / the initial neighborhood image block; Where, is the noise level of the iteratively adjusted neighborhood image block / initial neighborhood image block, is the size of the iteratively adjusted neighborhood image block / initial neighborhood image block, is the coordinate of the DCT coefficient in the DCT coefficient matrix, The coordinates are The DCT coefficients of Indicates rounding up.

[0013] Preferably, the calculation of the speed of change of the noise level during each iterative adjustment includes: calculating the The noise level of the neighborhood image block after the iterative adjustment The noise level of the initial neighborhood image block The difference ; Calculate the difference With the The side length of the blur kernel after iterative adjustment Ratio , as the first The speed of change of noise level during the iterative adjustment .

[0014] By calculating the changing speed of the noise level, the present invention can dynamically monitor the changing trend of the noise during the iteration process, provide a basis for the subsequent adjustment of the side length of the blur kernel, and ensure that each adjustment can effectively reduce the noise.

[0015] Preferably, the calculation of the decay rate of the noise level change rate during each iterative adjustment includes: calculating the The speed of change of noise level during the iterative adjustment With the The speed of change of noise level during the iterative adjustment The difference , as the first The decay rate of the noise level during the iterative adjustment .

[0016] By calculating the attenuation rate, the present invention can predict the convergence trend of noise changes, thereby judging when to stop iteration, providing a more accurate basis for adjusting the side length of the blur kernel and avoiding over-adjustment.

[0017] Preferably, the determining of the side length of the blur kernel after the next iterative adjustment comprises: if If it is greater than or equal to 0, stop the iteration and The size of the blur kernel adjusted after the iteration is used as the target size of the blur kernel of the neighborhood image block; if If it is less than 0, calculate the The side length of the blur kernel after iterative adjustment : Where, For the The side length of the blur kernel after iterative adjustment, For the The noise level of the neighborhood image block after the iterative adjustment, Iteration stop parameter, For the The speed of change of noise level during the iterative adjustment, For the The attenuation degree of the speed of change of the noise level during the iterative adjustment, Indicates rounding up.

[0018] The present invention combines the noise level of the neighborhood image block after the previous iterative adjustment, the changing speed of the noise level during the previous iterative adjustment, and the attenuation degree of the changing speed of the noise level to accurately predict the optimal blur kernel size, reduce unnecessary iterations, and significantly improve processing efficiency.

[0019] Preferably, the method of adjusting the side length of the blur kernel by multiple iterations to determine the target size of the blur kernel of the neighborhood image block further includes: initially setting an initial blur kernel: if the neighborhood image block is an upper neighborhood image block or a lower neighborhood image block, setting the size of the initial blur kernel to , if the neighborhood image block is the left neighborhood image block or the right neighborhood image block, then the size of the initial blur kernel is set to , is the initial side length of the blur kernel, Indicates rounding up; during the first iteration, the initial side length of the initial blur kernel is Increase by 1.

[0020] Preferably, the method of adjusting the side length of the blur kernel by multiple iterations to determine the target size of the blur kernel of the neighborhood image block further includes: The change rate of the noise level of the neighborhood image block after the iterative adjustment is greater than or equal to 0, or The noise level of the neighborhood image block after the first iteration adjustment is the same as that of the The ratio of the noise level of the neighborhood image block after the iteration adjustment is greater than When , stop the iteration and set the The size of the blur kernel after the iteration adjustment is used as the target size of the blur kernel of the neighborhood image block. Iteration stop parameter.

[0021] Preferably, the denoising process is performed on each image block in the region by using Wiener filtering, including: setting a target blur kernel of the image block according to a target size of the blur kernel of the image block; applying the target blur kernel to a Wiener filtering formula, and denoising the image block using the Wiener filtering formula to obtain a denoised image block.

[0022] The beneficial effects of the present invention are:

[0023] The present invention takes the average of the target sizes of the blur kernels of the four neighboring image blocks as the target size of the blur kernel of the intermediate image block, reduces the number of image blocks that require iterative adjustment processing, predicts the optimal blur kernel size by analyzing the noise level and its changing speed and attenuation speed, reduces the number of iterations, and thus speeds up the iterative adjustment speed; in summary, the present invention significantly improves the efficiency of the entire denoising process by reducing the number of image blocks that require iterative adjustment processing and speeding up the iterative adjustment speed, making it more suitable for status monitoring scenarios of substation power equipment with high timeliness requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a flow chart schematically illustrating a method for remote monitoring of power equipment status in a substation based on image analysis in the present invention;

[0025] Figure 2 is a schematic diagram schematically illustrating dividing each region into four neighborhood image blocks of equal size and one intermediate image block of size equal to 5×5;

[0026] Figure 3 is a schematic diagram schematically illustrating an inspection image containing noise;

[0027] Figure 4 This is a schematic diagram showing the method of this embodiment. Figure 3 Schematic diagram of the denoised monitoring image obtained by performing Wiener filtering. DETAILED DESCRIPTION

[0028] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work shall fall within the scope of protection of the present invention.

[0029] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0030] The embodiment of the present invention discloses a remote monitoring method for power equipment status of a substation based on image analysis, referring to Figure 1 , including steps S1 to S3:

[0031] S1. Collect monitoring images of power equipment in substations.

[0032] Specifically, fixed cameras are installed around key equipment and important areas in the substation, including but not limited to transformers, circuit breakers, mutual inductors, etc.; monitoring images of power equipment are collected through fixed cameras.

[0033] It should be noted that in actual applications, due to environmental factors (such as bad weather, changes in lighting), equipment performance limitations (such as equipment aging, equipment overheating, lens distortion), and transmission interference (such as signal attenuation, electromagnetic interference), the collected monitoring images often contain noise, which reduces the clarity of the monitoring images, obscures the detailed features of the equipment, and affects the monitoring personnel's judgment of the equipment status.

[0034] It should be further explained that Wiener filtering is a commonly used denoising method, in which the blur kernel is a key factor affecting the blur effect. Therefore, the blur kernel size is usually adjusted iteratively. However, in conventional Wiener filtering, the adjustment amount is the same in each iterative adjustment. This results in the need to calculate the blur kernel and denoised image multiple times to find the optimal blur kernel size, which significantly increases processing time and affects real-time performance and efficiency.

[0035] S2, divide the monitoring image into multiple regions of the same size, and divide each region into 4 neighborhood image blocks of the same size and 1 neighborhood image block of the same size. The intermediate image block.

[0036] It should be noted that, on the one hand, this embodiment divides the monitoring image into regions and image blocks, determines the target size of the blur kernel of the neighborhood image blocks through multiple iterative adjustments, and determines the target size of the blur kernel of the intermediate image block based on the target sizes of the blur kernels of the four neighborhood image blocks, thereby reducing the number of image blocks that need to be iteratively adjusted and processed, and can significantly improve the efficiency of the entire denoising process.

[0037] Specifically, the monitoring image is divided into multiple regions, and all regions are required to have the same size and be equal to , The value range is and is an integer.

[0038] Furthermore, each region is divided into 4 neighborhood image blocks of the same size and 1 neighborhood image block of the same size. The four neighborhood image blocks of the same size are: upper neighborhood image block, lower neighborhood image block, left neighborhood image block, right neighborhood image block, and the sizes of the upper neighborhood image block and the lower neighborhood image block are equal to , the size of the left and right neighbor image blocks is equal to .

[0039] It should be noted that the power equipment monitoring image of the substation is divided into multiple areas of the same size to facilitate subsequent unified processing and analysis; each area is further divided into four neighborhood image blocks of the same size and one intermediate image block of size 5×5, and when the target size of the blur kernel of the image block is subsequently determined, the average of the target sizes of the blur kernels of the four neighborhood image blocks is used as the target size of the blur kernel of the intermediate image block. This division method can retain the local features of the image while reducing the number of image blocks that need to be iteratively adjusted to determine the target size of the blur kernel, thereby reducing the amount of calculation and improving processing efficiency.

[0040] In one embodiment, each region is divided into 4 neighborhood image blocks of equal size and 1 neighborhood image block of equal size. The schematic diagram of the intermediate image block is as follows Figure 2 As shown in (1), the coordinates of the upper right corner vertex of the upper neighbor image block are the same as the coordinates of the upper left corner vertex of the right neighbor image block, the coordinates of the lower right corner vertex of the right neighbor image block are the same as the coordinates of the upper right corner vertex of the lower neighbor image block, and the coordinates of the lower left corner vertex of the lower neighbor image block are the same as the coordinates of the lower right corner vertex of the left neighbor image block; in another embodiment, each area is divided into 4 neighbor image blocks of the same size and 1 neighbor image block of the same size. The schematic diagram of the intermediate image block is as follows Figure 2 As shown in (2), the coordinates of the lower right corner vertex of the upper neighbor image block are the same as the upper right corner vertex of the right neighbor image block, the coordinates of the lower left corner vertex of the right neighbor image block are the same as the lower right corner vertex of the lower neighbor image block, and the coordinates of the upper left corner vertex of the lower neighbor image block are the same as the lower left corner vertex of the left neighbor image block.

[0041] In addition, in the size of the region / image block, " "The front represents the length of the side parallel to the horizontal direction," "The length of the side parallel to the vertical direction is represented by the following; for example: the size of the area is equal to , " "The front Indicates the length of the side parallel to the horizontal direction, "The back Represents the length of the side parallel to the vertical direction; the size of the upper and lower neighborhood image blocks is equal to , " "The front Indicates the length of the side parallel to the horizontal direction, "The back Indicates the length of the side parallel to the vertical direction.

[0042] S3, divide the monitoring image of the power equipment of the substation into multiple areas of the same size; divide each area into 4 neighborhood image blocks of the same size and 1 neighborhood image block of the same size. The intermediate image block in the region is denoised by Wiener filtering. The side length of the blur kernel is adjusted through multiple iterations to determine the target size of the blur kernel of the neighboring image block. The average of the target sizes of the blur kernels of the four neighboring image blocks is used as the target size of the blur kernel of the intermediate image block.

[0043] It should be noted that, on the other hand, this embodiment predicts the optimal blur kernel size of the neighborhood image block by analyzing the noise level, its changing speed and attenuation speed, thereby reducing the number of iterations, speeding up the iterative adjustment speed, and significantly improving the efficiency of the entire denoising process.

[0044] Specifically, each image block in the area is denoised by Wiener filtering: the side length of the blur kernel is adjusted through multiple iterations to determine the target size of the blur kernel of the neighborhood image block; the target size of the blur kernel of the intermediate image block is determined according to the target sizes of the blur kernels of the four neighborhood image blocks; the target blur kernel of the neighborhood image block / intermediate image block is set according to the target size of the blur kernel of the neighborhood image block / intermediate image block; the target blur kernel of the neighborhood image block / intermediate image block is applied to the Wiener filtering formula, and the neighborhood image block / intermediate image block is denoised using the Wiener filtering formula to obtain the denoised neighborhood image block / intermediate image block.

[0045] Among them, in each area, the target size of the blur kernel of the intermediate image block is determined according to the target sizes of the blur kernels of the four neighboring image blocks. The specific method is: calculate the average of the target sizes of the blur kernels of the four neighboring image blocks, round up the average of the target sizes of the blur kernels of the four neighboring image blocks, and use the rounded-up result as the target size of the blur kernel of the intermediate image block.

[0046] Furthermore, for each region, four denoised neighborhood image blocks and one denoised intermediate image block are obtained, and the four denoised neighborhood image blocks and one denoised intermediate image block constitute a denoised region; all denoised regions constitute a denoised monitoring image.

[0047] The specific process of iteratively adjusting the side length of the blur kernel is as follows:

[0048] 1. Calculate the noise level of the initial neighborhood image block.

[0049] Specifically, the initial neighborhood image block is subjected to discrete cosine transform to obtain the DCT coefficient matrix of the initial neighborhood image block; according to the high-frequency coefficients in the DCT coefficient matrix of the initial neighborhood image block, the noise level of the initial neighborhood image block is calculated and recorded as .

[0050] It should be noted that when the image blocks are transformed using discrete cosine transform, the high-frequency coefficients in the DCT coefficient matrix usually reflect the details and noise information in the image. Therefore, they can be used as an indicator of the noise level and provide a quantitative basis for the subsequent calculation of the noise change rate.

[0051] 2. Initially, set an initial blur kernel; apply the initial blur kernel to the Wiener filter formula, use the Wiener filter formula to denoise the initial neighborhood image block, and obtain the denoised neighborhood image block , calculate the denoised neighborhood image block The noise level is recorded as .

[0052] Specifically, if the neighborhood image block is an upper neighborhood image block or a lower neighborhood image block, the size of the initial blur kernel is set to , if the neighborhood image block is the left neighborhood image block or the right neighborhood image block, then the size of the initial blur kernel is set to , is the initial side length of the blur kernel, The specific value of can be set according to the actual application scenario and requirements, and the value range is [3,7]. Set to 3.

[0053] 3. When making the first iterative adjustment, the initial side length of the initial blur kernel is Increase by 1 as the side length of the blur kernel after the first iteration adjustment, and then obtain the blur kernel after the first iteration adjustment; apply the blur kernel after the first iteration adjustment to the Wiener filter formula, and use the Wiener filter formula to denoise the neighborhood image block Perform denoising to obtain the denoised neighborhood image block , and record it as the neighborhood image block after the first iteration adjustment; calculate the neighborhood image block after denoising The noise level is recorded as the noise level of the neighborhood image block after the first iteration adjustment. .

[0054] 4. Carry out the During the iterative adjustment, , according to The difference between the noise level of the neighborhood image block after the first iteration and the initial neighborhood image block, and The side length of the blur kernel after the iteration adjustment is calculated. The speed of change of noise level during the first iteration adjustment; The first iteration adjustment and The speed of change of noise level during the iterative adjustment is calculated. The decay rate of the noise level change rate during the first iteration adjustment; The speed of change of noise level and the degree of attenuation of the speed of change of noise level during the first iteration adjustment, as well as the The noise level of the neighborhood image block and the side length of the blur kernel after the first iteration are adjusted to determine the The side length of the blur kernel after the first iteration is adjusted, and then the The blur kernel after the iterative adjustment; The blur kernel after the iteration adjustment is applied to the Wiener filter formula, and the Wiener filter formula is used to filter the denoised neighborhood image blocks. Perform denoising to obtain the denoised neighborhood image block , and record it as The neighborhood image block after the iteration adjustment; calculate the denoised neighborhood image block The noise level is recorded as The noise level of the neighborhood image block after the iterative adjustment .

[0055] (1) wherein discrete cosine transform is performed on the iteratively adjusted neighborhood image block / initial neighborhood image block to obtain a DCT coefficient matrix of the iteratively adjusted neighborhood image block / initial neighborhood image block; and the noise level of the iteratively adjusted neighborhood image block / initial neighborhood image block is calculated based on the high-frequency coefficients in the DCT coefficient matrix. The specific calculation method is:

[0056] ;

[0057] Where, is the noise level of the iteratively adjusted neighborhood image block / initial neighborhood image block, is the size of the iteratively adjusted neighborhood image block / initial neighborhood image block, is the coordinate of the DCT coefficient in the DCT coefficient matrix, The coordinates are DCT coefficients; in the DCT coefficient matrix, exist Within the range and exist DCT coefficients in the range is the high frequency coefficient, Indicates rounding up.

[0058] (2) According to The difference between the noise level of the neighborhood image block after the first iteration and the initial neighborhood image block, and The side length of the blur kernel after the iteration adjustment is calculated. The speed of change of noise level during the iterative adjustment includes: calculating the The noise level of the neighborhood image block after the iterative adjustment The noise level of the initial neighborhood image block The difference ; Calculate the difference With the The side length of the blur kernel after iterative adjustment Ratio , as the first The speed of change of noise level during the iterative adjustment .

[0059] It should be noted that when When The speed of change of noise level during the iterative adjustment, including: calculating the neighborhood image block after denoising Noise level The noise level of the initial neighborhood image block The difference ; Calculate the difference and the initial edge length of the blur kernel Ratio , as the speed of change of the noise level during the 0th iteration adjustment .

[0060] It should be noted that as the number of iterations increases, the side length of the blur kernel continues to increase, and correspondingly the noise level of the neighborhood image block gradually decreases.

[0061] (3) According to The first iteration adjustment and The speed of change of noise level during the iterative adjustment is calculated. The decay rate of the noise level change rate during the iterative adjustment, including: calculating the The speed of change of noise level during the iterative adjustment With the The speed of change of noise level during the iterative adjustment The difference , as the first The decay rate of the noise level during the iterative adjustment .

[0062] It should be noted that, as the number of iterations increases, the rate of change of the noise level of the neighborhood image blocks gradually decreases.

[0063] (4) According to The speed of change of noise level and the degree of attenuation of the speed of change of noise level during the first iteration adjustment, as well as the The noise level of the neighborhood image block and the side length of the blur kernel after the first iteration are adjusted to determine the The side length of the blur kernel after the iteration adjustment, including:

[0064] if If it is greater than or equal to 0, stop the iteration and The size of the blur kernel adjusted after the iteration is used as the target size of the blur kernel of the neighborhood image block.

[0065] if If it is less than 0, calculate the The side length of the blur kernel after iterative adjustment :

[0066] ;

[0067] Where, For the The side length of the blur kernel after iterative adjustment, For the The noise level of the neighborhood image block after the iterative adjustment, Iteration stop parameter, For the The speed of change of noise level during the iterative adjustment, For the The attenuation degree of the speed of change of the noise level during the iterative adjustment, Indicates rounding up.

[0068] It should be noted that by calculating the The attenuation degree of the speed of change of the noise level during the iterative adjustment , predict the convergence trend of noise changes, and then predict the The speed of change of the noise level during the iterative adjustment, that is, The condition for stopping the iteration is that the speed of change of the noise level of the neighborhood image block after the next iteration is greater than or equal to 0, or the ratio of the noise level of the neighborhood image block after the previous iteration to the noise level of the neighborhood image block after the next iteration is greater than Therefore, when predicting The speed of change of noise level during the iterative adjustment Or equal to 0, stop iteration, and directly The size of the blur kernel adjusted after the first iteration is used as the target size of the blur kernel of the neighborhood image block; or The noise level is equal to When , the iteration stops, during which the noise level changes , then the change in noise level With the predicted The speed of change of noise level during the iterative adjustment The ratio of is the adjustment amount of the side length of the blur kernel; The side length of the blur kernel after iterative adjustment Add the adjustment amount on the basis of The side length of the blur kernel after the iteration adjustment can achieve the purpose of predicting the optimal blur kernel size and reducing the number of iterations.

[0069] 5. And so on until the The change rate of the noise level of the neighborhood image block after the iterative adjustment is greater than or equal to 0, or The noise level of the neighborhood image block after the first iteration adjustment is the same as that of the The ratio of the noise level of the neighborhood image block after the iteration adjustment is greater than When , stop the iteration and set the The size of the blur kernel after the iteration adjustment is used as the target size of the blur kernel of the neighborhood image block. Iteration stop parameter.

[0070] Among them, the iteration stop parameter The specific value of can be set according to the actual application scenario and requirements, and the iteration stop parameter The value range is [0.8,1), and the present invention will iterate the stopping parameter Set to 0.85.

[0071] For example, the schematic diagram of the inspection image containing noise is as follows: Figure 3 As shown, the method of this embodiment is used to Figure 3 Perform Wiener filtering to obtain the schematic diagram of the denoised monitoring image. Figure 4 shown.

[0072] It should be noted that this embodiment predicts the optimal blur kernel size of the neighborhood image block by analyzing the noise level, its changing speed and attenuation speed, which can reduce the number of iterations and speed up the iterative adjustment speed. On this basis, the target size of the blur kernel of the intermediate image block is determined directly according to the target size of the blur kernel of the four neighborhood image blocks, which can reduce the number of image blocks that need iterative adjustment processing; by speeding up the iterative adjustment speed and reducing the number of image blocks that need iterative adjustment processing, the efficiency of the entire denoising process can be significantly improved, ensuring the timeliness of the status monitoring of the substation power equipment.

Claims

1. A remote monitoring method for substation power equipment status based on image analysis, characterized in that: include: Divide the monitoring image of the power equipment of the substation into multiple areas of equal size; Each region is divided into 4 neighborhood image blocks of the same size and 1 block of size equal to The intermediate image block in the region is de-noised by using Wiener filtering. The target size of the blur kernel of the neighboring image block is determined by adjusting the side length of the blur kernel through multiple iterations. The average of the target sizes of the blur kernels of the four neighboring image blocks is used as the target size of the blur kernel of the intermediate image block. The target blur kernel of the image block is set according to the target size of the blur kernel of the image block. Applying the target blur kernel to the Wiener filter formula, performing denoising on the image block using the Wiener filter formula, and obtaining a denoised image block; The multiple iterations of adjusting the side length of the blur kernel include: calculating the noise level of the neighborhood image block after each iteration adjustment based on high-frequency coefficients in a DCT coefficient matrix obtained by performing discrete cosine transform on the neighborhood image block after each iteration adjustment; The changing speed of the noise level in each iterative adjustment is calculated based on the difference between the noise levels of the neighborhood image block after each iterative adjustment and the initial neighborhood image block, as well as the side length of the blur kernel after each iterative adjustment; the decaying speed of the changing speed of the noise level in each iterative adjustment is calculated based on the changing speed of the noise level in each iterative adjustment and the previous iterative adjustment; the side length of the blur kernel after the next iterative adjustment is determined based on the changing speed of the noise level and the decaying speed of the changing speed of the noise level in each iterative adjustment, as well as the noise level of the neighborhood image block after each iterative adjustment and the side length of the blur kernel.

2. The remote monitoring method for substation power equipment status based on image analysis according to claim 1 is characterized in that: Dividing the monitoring image into a plurality of regions of equal size, including: The monitoring image is divided into multiple regions, requiring all regions to have the same size and be equal to , The value range is and is an integer.

3. The remote monitoring method for substation power equipment status based on image analysis according to claim 1 is characterized in that: Each region is divided into 4 neighborhood image blocks of the same size and 1 block of size equal to The intermediate image block includes: The four neighborhood image blocks of the same size are: upper neighborhood image block, lower neighborhood image block, left neighborhood image block, and right neighborhood image block, and the sizes of the upper neighborhood image block and the lower neighborhood image block are equal to , the size of the neighboring image block and the right neighboring image block is equal to .

4. The method for remote monitoring of substation power equipment status based on image analysis according to claim 1, characterized in that: The method for calculating the noise level of the iteratively adjusted neighborhood image block / initial neighborhood image block is: Performing discrete cosine transform on the iteratively adjusted neighborhood image block / the initial neighborhood image block to obtain a DCT coefficient matrix of the iteratively adjusted neighborhood image block / the initial neighborhood image block; ； Where, is the noise level of the iteratively adjusted neighborhood image block / initial neighborhood image block, is the size of the iteratively adjusted neighborhood image block / initial neighborhood image block, is the coordinate of the DCT coefficient in the DCT coefficient matrix, The coordinates are The DCT coefficients of Indicates rounding up.

5. The remote monitoring method for substation power equipment status based on image analysis according to claim 1 is characterized in that: The calculation of the change rate of the noise level during each iterative adjustment includes: Calculate the The noise level of the neighborhood image block after the iterative adjustment The noise level of the initial neighborhood image block The difference ; Calculate the difference With the The side length of the blur kernel after iterative adjustment Ratio , as the first The speed of change of noise level during the iterative adjustment .

6. The method for remote monitoring of substation power equipment status based on image analysis according to claim 1, characterized in that: The calculation of the attenuation rate of the change rate of the noise level during each iterative adjustment includes: Calculate the The speed of change of noise level during the iterative adjustment With the The speed of change of noise level during the iterative adjustment The difference , as the first The decay rate of the noise level during the iterative adjustment .

7. The method for remote monitoring of substation power equipment status based on image analysis according to claim 1, characterized in that: The determining of the side length of the blur kernel after the next iterative adjustment includes: if If it is greater than or equal to 0, stop the iteration and The size of the blur kernel adjusted after the iteration is used as the target size of the blur kernel of the neighborhood image block; if If it is less than 0, calculate the The side length of the blur kernel after iterative adjustment : ； Where, For the The side length of the blur kernel after iterative adjustment, For the The noise level of the neighborhood image block after the iterative adjustment, Iteration stop parameter, For the The speed of change of noise level during the iterative adjustment, For the The decay rate of the noise level change rate during the iterative adjustment, Indicates rounding up.

8. The remote monitoring method for substation power equipment status based on image analysis according to claim 1 is characterized in that: The step of adjusting the side length of the blur kernel by multiple iterations to determine the target size of the blur kernel of the neighborhood image block further includes: Initially, set an initial blur kernel: if the neighborhood image block is the upper neighborhood image block or the lower neighborhood image block, set the size of the initial blur kernel to , if the neighborhood image block is the left neighborhood image block or the right neighborhood image block, then the size of the initial blur kernel is set to , is the initial side length of the blur kernel, Indicates rounding up; During the first iteration, the initial side length of the initial blur kernel is Increase by 1.

9. The method for remote monitoring of substation power equipment status based on image analysis according to claim 1, characterized in that: The step of adjusting the side length of the blur kernel by multiple iterations to determine the target size of the blur kernel of the neighborhood image block further includes: When The change rate of the noise level of the neighborhood image block after the iterative adjustment is greater than or equal to 0, or The noise level of the neighborhood image block after the first iteration adjustment is the same as that of the The ratio of the noise level of the neighborhood image block after the iteration adjustment is greater than When , stop the iteration and set the The size of the blur kernel after the iteration adjustment is used as the target size of the blur kernel of the neighborhood image block. Iteration stop parameter.

Images