An electric power image processing method and device

Abstract

The application discloses a kind of electric power image processing method and device, method includes: first image is obtained by acquisition;By SIFT algorithm, the first image is registered with reference image and is aligned processing;Difference is obtained by the first image and reference image after registration alignment, difference image;Based on the difference image, using frame strategy determines the difference area in the first image.The above scheme compares the reference image corresponding to the normal working state by determining the acquisition image, determines the electric power equipment with exception, realizes the effective monitoring of substation equipment defect, operating state and safety risk, greatly improves the inspection efficiency of substation equipment, improves the control ability of equipment quality.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and more specifically, to a method and apparatus for processing electrical images. Background Technology

[0002] Operation and maintenance of power grid equipment is fundamental to equipment management and safety assurance. Routine equipment inspections are the most crucial aspect of operation and maintenance, serving as a vital means to obtain equipment operating status and provide decision-making support. With the continuous advancement of power grid construction, the number of power grid devices is rapidly increasing, placing higher demands on equipment safety during power grid operation.

[0003] With the application of substation inspection robots and video surveillance devices, a small portion of the inspection work on substation equipment can currently be completed with the assistance of these devices. However, existing substation inspection robots and video surveillance systems have limited functionality and lack the ability to detect equipment defects or operational status. Furthermore, the images of substation equipment inspections (acquired by inspection robots or video surveillance devices) require manual verification or review, consuming significant time and manpower, severely reducing and impacting maintenance efficiency. Summary of the Invention

[0004] In view of the above, the present invention provides the following technical solution:

[0005] A power image processing method, comprising:

[0006] The first image was acquired.

[0007] The first image and the reference image are registered and aligned using the SIFT algorithm.

[0008] The difference between the first image after registration and alignment and the reference image is obtained;

[0009] Based on the difference image, the difference regions in the first image are identified and marked.

[0010] Optionally, the process of registering and aligning the first image with the reference image using the SIFT algorithm includes:

[0011] Feature points of the first image and the reference image are calculated based on the SIFT algorithm;

[0012] Based on the calculated feature points, a nearest neighbor algorithm is used to match feature points between the first image and the reference image;

[0013] The first image and the reference image after feature point matching are transformed and aligned.

[0014] Optionally, the transformation and alignment of the first image and the reference image after feature point matching includes:

[0015] The reference image is aligned with the first image by rotation transformation using a homography matrix.

[0016] Optionally, the step of subtracting the registered and aligned first image from the reference image to obtain the difference image includes:

[0017] The first image after registration and alignment is compared with the image after a sudden turn by matrix subtraction, and the absolute value of the matrix difference is determined as the difference image.

[0018] Optionally, after subtracting the registered and aligned first image from the reference image to obtain the difference image, the method further includes:

[0019] The difference image is then denoised.

[0020] Optionally, the denoising process on the difference image includes:

[0021] Edge detection processing is performed on the reference image to obtain an edge detection result image;

[0022] The edge detection result image is multiplied with the difference image to eliminate the edge noise data summarized in the difference image.

[0023] Optionally, determining and identifying the difference regions in the first image based on the difference image includes:

[0024] Extract the maximum pixel value from the difference image;

[0025] If the maximum pixel value is greater than a first threshold, the difference image is binarized to obtain a binarized image.

[0026] The regions containing pixels with a first value in the binarized image are identified.

[0027] Optionally, before marking the pixels in the binarized image with a rectangular box that have a pixel value of the first value, the method further includes:

[0028] Traverse each pixel in the binarized image whose pixel value is the first value. If the total number of pixels with the first value within a first range around the pixel is less than a second threshold, then set the pixel value of that pixel to 0.

[0029] A power image processing device, comprising:

[0030] The image acquisition module is used to acquire and obtain the first image;

[0031] The image registration module is used to register and align the first image with the reference image using the SIFT algorithm;

[0032] The difference processing module is used to calculate the difference between the registered and aligned first image and the reference image to obtain the difference image;

[0033] The difference determination module is used to determine the difference regions in the first image based on the difference image and using a drawing frame strategy.

[0034] Optionally, the image registration module includes:

[0035] The feature point determination module is used to calculate the feature points of the first image and the reference image based on the SIFT algorithm.

[0036] The feature point matching module is used to perform feature point matching between the first image and the reference image based on the calculated feature points using a nearest neighbor algorithm.

[0037] The alignment and transformation module is used to transform and align the first image and the reference image after feature point matching is completed.

[0038] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a power image processing method and apparatus. The method includes: acquiring a first image; registering and aligning the first image with a reference image using the SIFT algorithm; calculating the difference between the registered and aligned first image and the reference image to obtain a difference image; and determining the difference region in the first image based on the difference image using a drawing frame strategy. This solution, by comparing the acquired image with a reference image corresponding to the normal operating state, identifies abnormal power equipment, achieving effective monitoring of substation equipment defects, operating status, and safety risks. This greatly improves the efficiency of substation equipment inspection and enhances the ability to control equipment quality. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0040] Figure 1 This is a flowchart of a power image processing method disclosed in an embodiment of the present invention;

[0041] Figure 2 This is a flowchart of image registration and alignment processing disclosed in an embodiment of this application;

[0042] Figure 3 This is an example diagram illustrating the matching features between the reference image and the first image disclosed in an embodiment of this application.

[0043] Figure 4 This is an example image showing the difference between the reference image and the first image disclosed in the embodiments of this application;

[0044] Figure 5 This is an example image of the edge detection result of the reference image disclosed in the embodiments of this application;

[0045] Figure 6 This is an example image of the difference image after noise removal disclosed in an embodiment of this application;

[0046] Figure 7 This is a schematic diagram of the first image defect recognition process disclosed in an embodiment of this application;

[0047] Figure 8 This is an example of a binarized image disclosed in an embodiment of this application;

[0048] Figure 9 This is an example image of the first image anomaly detection result disclosed in an embodiment of this application;

[0049] Figure 10 This is a schematic diagram of the SIFT-based power image processing process disclosed in an embodiment of this application;

[0050] Figure 11 This is a schematic diagram of the structure of a power image processing device disclosed in the embodiments of this application. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0052] Figure 1 This is a flowchart illustrating a power image processing method disclosed in an embodiment of the present invention. See also... Figure 1 As shown, the power image processing method may include:

[0053] Step 101: Acquire the first image.

[0054] The first image is an image of the power grid equipment at the power grid site. It can be an overall or partial structural image of the power equipment. The overall image can be used to determine whether there is any damage to the appearance of the power equipment. Similarly, the partial structural image can be used to determine whether the important structure of the power equipment is damaged, or it can be used to determine whether there is any abnormality in the current working status of the power equipment. For example, the partial structure can be the instrument display of the power equipment, and some parameters, indicator lights, etc. can indicate the working status of the power equipment.

[0055] The first image can be acquired by an inspection robot at the power grid equipment site, or by an image acquisition device installed at a specific location at the power grid equipment site. This application does not impose any fixed restrictions on this.

[0056] Step 102: Register and align the first image with the reference image using the SIFT (Scale-invariant feature transform) algorithm.

[0057] The reference image can be an image taken in the same space as the first image but at a different time; the reference image refers to an image taken when the power equipment is normal, and the reference image was obtained earlier than the first image.

[0058] This application can calculate the feature points of the reference image and the first image based on the SIFT algorithm, and use the KNN (K-Nearest Neighbor) algorithm to complete the feature point matching between the two images; then, the transformation and alignment of the two images are completed based on the matched feature points of the reference image and the first image. The specific implementation will be described in detail in the following embodiments, and will not be explained in detail here.

[0059] Step 103: Calculate the difference between the registered and aligned first image and the reference image to obtain the difference image.

[0060] Specifically, the first image after registration and alignment and the rapidly rotating image can be subjected to matrix subtraction, and the absolute value of the obtained matrix difference can be determined as the difference image.

[0061] Step 104: Based on the difference image, determine and identify the difference regions in the first image.

[0062] After obtaining the difference image, based on a certain recognition algorithm, the regions with differences can be identified and marked. The content represented by the regions that differ from the reference image corresponds to the abnormal power equipment or the power equipment with abnormal operating conditions.

[0063] The power image processing method described in this embodiment can identify abnormal power equipment by comparing and analyzing the acquired image with the reference image corresponding to the normal working state. This enables effective monitoring of substation equipment defects, operating status, and safety risks, greatly improving the efficiency of substation equipment inspection and enhancing the ability to control equipment quality.

[0064] Figure 2 This is a flowchart illustrating the image registration and alignment process disclosed in an embodiment of this application. (In conjunction with...) Figure 2 As shown in the above embodiment, the registration and alignment process of the first image and the reference image using the SIFT algorithm may include:

[0065] Step 201: Calculate the feature points of the first image and the reference image based on the SIFT algorithm.

[0066] This step performs feature point detection. For feature point detection:

[0067] First, a scale space is constructed, and feature points are described based on this scale space. The scale space is constructed according to equations (1) and (2), as shown below:

[0068]

[0069]

[0070] In Equation (1), the variable Gaussian kernel function is given, σ represents the scale-space Gaussian blur function, and p and q represent the dimensions of the Gaussian template. In Equation (2), I(x,y) is the input original image, and (x,y) is the pixel position of the image.

[0071] A Gaussian difference image can be constructed using the difference operator in the scale space of the Gaussian difference. The detection points in the image are compared with the other eight neighboring points at the same scale and the 2x9 points at the adjacent scales above and below, so that extreme points can be detected in both the two-dimensional space and the scale space. The local maxima obtained are the feature points.

[0072] Then, a feature point descriptor can be constructed, and the gradient magnitude and direction of the pixels in a 16x16 window centered on the feature point can be calculated. The pixels in the window are divided into 16 sub-regions, and the gradient and magnitude of each sub-region are calculated in eight directions, which can form a 128-dimensional feature vector.

[0073] Step 202: Based on the calculated feature points, the nearest neighbor algorithm is used to match the feature points of the first image and the reference image.

[0074] This step is called feature point matching. Specifically, the distance between the reference image and the feature points of the first image can be calculated based on Euclidean distance, and the ratio of the nearest distance to the second nearest distance can be calculated. If the ratio is within the threshold range, the matching can be considered successful. The threshold can be set based on experience or requirements. The inventors of this application conducted multiple experiments during the implementation of the solution of this application, and found that the effect was optimal when the threshold was 0.5. Therefore, 0.5 can be used as the threshold for Euclidean distance in the solution of this application. The Euclidean distance formula used for feature point matching is shown in Equation (3).

[0075]

[0076] Where dis represents the distance between the feature points of the reference image and the first image, P is the dimension of the feature point descriptor, and D m D n These are feature point descriptors for the reference image and the first image, respectively.

[0077] Step 203: Transform and align the first image and the reference image after feature point matching is completed.

[0078] The specific implementation of this step is as follows: First, the feature points corresponding to the reference image and the first image can be determined through the aforementioned steps, such as... Figure 3 As shown; then, in order to eliminate the difference between the reference image and the first image taken from different perspectives, the homography matrix can be used to align the reference image and the first image through rotation transformation, so as to determine the difference parts of the image. The homography of a plane refers to the projection mapping from one plane to another.

[0079] It should be noted that, in order to better understand the specific implementation of this application, the accompanying drawings of the reference image and the first image in the embodiments of this application, including the related drawings of the difference images that will appear later, are all corresponding images in one example and can be referred to each other to understand the corresponding content.

[0080] In the above embodiments, the step of subtracting the registered and aligned first image and the reference image to obtain the difference image may include: performing matrix subtraction on the registered and aligned first image and the rapidly rotating image, and determining the absolute value of the obtained matrix difference as the difference image. Figure 4 This is an example image showing the difference between the reference image and the first image disclosed in the embodiments of this application, combined with... Figure 4 As shown, the difference between the first image and the reference image indicates an abnormal door closure.

[0081] In an optional implementation, to improve the accuracy of the final recognition result, after subtracting the registered and aligned first image and the reference image to obtain the difference image, the method may further include: performing noise reduction processing on the difference image.

[0082] Specifically, the denoising process may be as follows: perform edge detection processing on the reference image to obtain an edge detection result image; multiply the edge detection result image with the difference image to eliminate the edge noise data summarized in the difference image.

[0083] The purpose of edge detection is to identify points in a digital image where brightness changes significantly. Image edge detection greatly reduces the amount of data and eliminates irrelevant information while preserving important structural properties of the image. Figure 5 This is the edge detection result of the reference image.

[0084] The edge detection results of the reference image can be used to eliminate noise data in the difference image between the reference image and the first image. Specifically, each pixel value in the edge detection result image of the reference image is first detected one by one. If a pixel is an edge value, the pixel values ​​within three surrounding pixel units are set to 0. Then, the edge detection result image with the set values ​​is multiplied by the difference image to eliminate edge noise data in the difference image. The result is as follows: Figure 6 As shown.

[0085] Figure 4 and Figure 6 These are the difference images of the baseline image and the first image before and after removing edge noise, respectively. By comparison, it can be found that... Figure 3 The more obvious edge noise data in Figure 5 All characters have been removed.

[0086] In one implementation, determining and identifying the difference region in the first image based on the difference image may include: extracting the maximum pixel value of the difference image; performing binarization processing on the difference image to obtain a binarized image if the maximum pixel value is greater than a first threshold; and identifying the region where the pixel with a pixel value of a first value is located in the binarized image.

[0087] For example, firstly, the maximum pixel value n is extracted from the difference image after noise removal. If n is less than or equal to 30, the reference image and the first image are indistinguishable; if n is greater than 30, the difference image is binarized. Then, the coordinate boxes of the difference portion are calculated by combining the binarized difference image and the bounding box strategy, and the defective device in the first image is marked. The specific process is as follows: Figure 6 As shown. The first threshold of 30 can be determined through experimental calibration.

[0088] Image binarization involves setting the grayscale value of pixels in an image to 0 or 255, effectively creating a black and white image. Binarization significantly reduces the amount of data in an image, thus highlighting the outline of the target. In this invention, for example, pixel values ​​less than or equal to 0.5*n can be set to 0, and pixel values ​​greater than 0.5*n can be set to 255. The binarized difference image is shown below. Figure 8 As shown.

[0089] Before marking the pixels in the binarized image with a rectangle that have a pixel value of the first value, the method may further include: traversing each pixel in the binarized image that has a pixel value of the first value, and if the total number of pixels with a pixel value of the first value within a first range around the pixel is less than a second threshold, then setting the pixel value of the pixel to 0.

[0090] Specifically, the number of pixels with a value of 255 in the binarized difference image can be counted, with a total of N. Then, for each pixel with a value of 255, the process iterates through it. If the total number of pixels with a value of 255 within 50 units of that pixel is less than 0.1*N, then that pixel is considered noise and its pixel value is set to 0. Finally, the pixels with a value of 255 in the binarized difference image are marked with a rectangle. The portion marked by this rectangle represents the abnormal portion in the first image, such as... Figure 9 As shown.

[0091] Figure 10 This is a schematic diagram of the SIFT-based power image processing process disclosed in an embodiment of this application. (Combined with...) Figure 10 As shown, the power image processing solution of this application can comprehensively improve the multi-scene recognition and detection capabilities of substation equipment, including defect identification, operation status identification, and safety risk identification. At the same time, it can carry out autonomous collaborative inspections from multiple perspectives, multiple terminals, and multiple scenarios, realizing the automated, intelligent, and three-dimensional acquisition and analysis of substation equipment video images, greatly improving the efficiency of substation equipment inspection and enhancing the ability to control equipment quality.

[0092] For the foregoing method embodiments, in order to simplify the description, they are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0093] The methods described in the above embodiments of the present invention are detailed. The methods of the present invention can be implemented by various forms of devices. Therefore, the present invention also discloses a device. Specific embodiments are given below for detailed description.

[0094] Figure 11 This is a schematic diagram of the structure of one of the power image processing devices disclosed in the embodiments of this application. See also... Figure 11 As shown, the power image processing device 100 may include:

[0095] Image acquisition module 1101 is used to acquire the first image.

[0096] The image registration module 1102 is used to register and align the first image with the reference image using the SIFT algorithm.

[0097] The difference processing module 1103 is used to calculate the difference between the registered and aligned first image and the reference image to obtain the difference image.

[0098] The difference determination module 1104 is used to determine the difference region in the first image based on the difference image and using a drawing frame strategy.

[0099] The power image processing device described in this embodiment can identify abnormal power equipment by comparing and analyzing the acquired image with the reference image corresponding to the normal working state. This enables effective monitoring of substation equipment defects, operating status, and safety risks, greatly improving the efficiency of substation equipment inspection and enhancing the ability to control equipment quality.

[0100] In one implementation, the image registration module may include: a feature point determination module, used to calculate feature points of a first image and a reference image based on the SIFT algorithm; a feature point matching module, used to perform feature point matching between the first image and the reference image based on the calculated feature points using a nearest neighbor algorithm; and an alignment conversion module, used to convert and align the first image and the reference image after feature point matching.

[0101] Any of the power image processing devices described in the above embodiments includes a processor and a memory. The image acquisition module, image registration module, difference processing module, difference determination module, feature point determination module, feature point matching module, alignment conversion module, etc. in the above embodiments are all stored as program modules in the memory, and the processor executes the above program modules stored in the memory to realize the corresponding functions.

[0102] The processor contains a kernel, which retrieves the corresponding program modules from memory. One or more kernels can be configured, and the processing of backtracking data can be achieved by adjusting kernel parameters.

[0103] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0104] In an exemplary embodiment, a computer-readable storage medium is also provided, which can be directly loaded into the internal memory of a computer, and contains software code. After being loaded and executed by the computer, the computer program can implement the steps shown in any embodiment of the above-described power image processing method.

[0105] In an exemplary embodiment, a computer program product is also provided, which can be directly loaded into the internal memory of a computer and contains software code. After being loaded and executed by the computer, the computer program can implement the steps shown in any embodiment of the power image processing method described above.

[0106] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0107] It should also be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0108] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0109] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of electrical image processing, characterized by, include: The first image was acquired. Feature points of the first image and the reference image are calculated based on the SIFT algorithm; Based on the calculated feature points, a nearest neighbor algorithm is used to match feature points between the first image and the reference image. Specifically, the distance between feature points in the reference image and the first image is calculated based on Euclidean distance, and the ratio of the nearest distance to the second nearest distance is calculated. If the ratio is within a threshold range, a successful match is determined. The reference image and the first image are aligned using a homography matrix through rotation transformation. The difference between the registered and aligned first image and the reference image is calculated to obtain a difference image. Based on the difference image, the difference regions in the first image are determined and identified; The step of determining and identifying the difference region in the first image based on the difference image includes: extracting the maximum pixel value of the difference image; if the maximum pixel value is greater than a first threshold, performing binarization processing on the difference image to obtain a binarized image; traversing each pixel in the binarized image whose pixel value is a first value, and if the total number of pixels with a pixel value of the first value within a first range around the pixel is less than a second threshold, then setting the pixel value of the pixel to 0; and identifying the region where the pixel with a pixel value of the first value is located in the binarized image includes: calculating the coordinate box of the difference part by combining the binarized difference image and the drawing frame strategy, and marking the defective device in the first image.

2. The electric power image processing method according to claim 1, characterized by, The step of subtracting the registered and aligned first image from the reference image to obtain the difference image includes: The first image after registration and alignment is compared with the image after a sudden turn by matrix subtraction, and the absolute value of the matrix difference is determined as the difference image.

3. The electric power image processing method according to claim 1, characterized by, After subtracting the registered and aligned first image from the reference image to obtain the difference image, the method further includes: The difference image is then denoised.

4. The method of claim 3, wherein the power image is a power spectrum of the image. The denoising process for the difference image includes: Edge detection processing is performed on the reference image to obtain an edge detection result image; The edge detection result image is multiplied by the difference image to eliminate the edge noise data summarized in the difference image.

5. An electric power image processing apparatus characterized by comprising: include: The image acquisition module is used to acquire and obtain the first image; The image registration module is used to calculate the feature points of the first image and the reference image based on the SIFT algorithm; Based on the calculated feature points, a nearest neighbor algorithm is used to match feature points between the first image and the reference image. Specifically, the distance between feature points in the reference image and the first image is calculated based on Euclidean distance, and the ratio of the nearest distance to the second nearest distance is calculated. If the ratio is within a threshold range, a successful match is determined. The reference image and the first image are aligned using a homography matrix through rotation transformation. A difference processing module is used to calculate the difference between the registered and aligned first image and the reference image to obtain a difference image. The difference determination module is used to determine the difference region in the first image based on the difference image and using a drawing frame strategy; The difference determination module is specifically used for: extracting the maximum pixel value of the difference image; if the maximum pixel value is greater than a first threshold, performing binarization processing on the difference image to obtain a binarized image; traversing each pixel in the binarized image whose pixel value is a first value, and if the total number of pixels with a pixel value of the first value within a first range around the pixel is less than a second threshold, then setting the pixel value of the pixel to 0; and marking the area where the pixel with a pixel value of the first value is located in the binarized image, including: calculating the coordinate box of the difference part by combining the binarized difference image and the drawing frame strategy, and marking the defective device in the first image.

Images