Relay protection device constant value automatic checking method, device, equipment and storage medium

Through automated image processing and comparison methods, the problems of low fixed value verification efficiency and poor accuracy of the power grid relay protection device are solved, and efficient and accurate fixed value verification is achieved, ensuring the safe and stable operation of the power grid.

CN119027439AInactive Publication Date: 2024-11-26CHAOYANG POWER SUPPLY COMPANY OF STATE GRID LIAONING ELECTRIC POWER SUPPLY
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
CN202410963531.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-18
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

With the expansion of the power grid and the increase of relay protection devices in substations, fixed-value data becomes dense and large in number, and it is difficult to meet the demand by relying on manual verification, resulting in large workload, low efficiency and accuracy of verification results being affected, threatening the power grid. safe and stable operation.

Method used

Using a relay protection device automatic fixed value verification method, by obtaining the fixed value table image, converting it into a grayscale diagram and performing adaptive threshold binarization, detecting edges and corner points, performing perspective transformation correction, segmenting cells, and identifying The text information is compared with the set value list issued by the power grid dispatching agency, and the consistency of the value is automatically judged. If it is inconsistent, it will be marked and prompted to perform manual review.

Benefits of technology

The accuracy of the fixed value verification results is improved, and compared with traditional methods, the errors in manual verification are reduced, efficiency is improved, the accuracy of the fixed value setting of the relay protection device is ensured, and the safe and stable operation of the power grid is ensured.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119027439A_ABST
    Figure CN119027439A_ABST
Patent Text Reader

Abstract

The invention provides a relay protection device constant value automatic checking method and device, equipment and a storage medium. According to the method, a constant value table image is converted into a grey-scale image through an image processing and recognition technology, then the grey-scale image is converted into a black-and-white image, and line discontinuity is processed. Then, the edge of the image is detected, angular points are found through a Harris algorithm, and perspective transformation is carried out to correct the image; then, cells containing the constant value are segmented, and character information is recognized and compared with a constant value sheet of a power grid dispatching mechanism; if yes, storing the information; and if not, marking and prompting manual re-checking. According to the method, the accuracy of the checking result is improved, and automatic checking of the fixed value of the relay protection device is realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application belongs to the field of relay safety, and in particular, relates to a method, device, equipment and storage medium for automatically checking the set values ​​of a relay protection device. Background Art

[0002] One of the core elements of relay protection devices is the accuracy of internal settings, which is the cornerstone for ensuring the reliable operation of the device. The power grid dispatching agency is responsible for accurately calculating and issuing the setting notification sheet of the relay protection device (referred to as the "setting sheet"). Subsequently, professional relay protection personnel will download and print this setting sheet in the system, and then go to the site to solidify it into the actual setting value of the protection device. After completing this step, the device will print out the current setting status (referred to as the "device setting sheet").

[0003] To ensure accuracy, the relay protection personnel and equipment operation and maintenance personnel need to compare the contents of the setting value sheet and the device setting value sheet item by item. Only when the two are completely consistent can the setting value of the relay protection device be confirmed to be correct. In addition, to ensure the consistency and accuracy of the setting value, the protection device must also undergo regular setting value verification.

[0004] However, with the expansion of the power grid, the increase in relay protection devices in substations and the refinement of equipment management, the set value data has become dense and large in quantity. It is difficult to meet the needs by relying solely on manual verification. Not only is the workload large and the efficiency low, it may also affect the accuracy of the verification results, thereby causing false operation of the protection device and threatening the safe and stable operation of the power grid. Summary of the invention

[0005] The purpose of the present application is to overcome the above-mentioned prior art and provide a method, device, equipment and storage medium for automatically checking the setting values ​​of a relay protection device.

[0006] The present application provides a method for automatically checking the setting value of a relay protection device, comprising:

[0007] Get the setting table image of the relay protection device;

[0008] Converting the fixed value table image into a grayscale image;

[0009] The grayscale image is converted into a black-and-white image using an adaptive threshold binarization method based on the Wellner algorithm;

[0010] reconnecting discontinuous line portions in the black-and-white image to generate a first processed image;

[0011] Detecting the edge of the first processed image using a Canny operator;

[0012] Acquire four corner points of the outline of the table in the first processed image by using Harris corner detection algorithm;

[0013] Based on the edges and corner points, performing tilt correction on the first processed image by perspective transformation to obtain a second processed image;

[0014] segmenting cells containing a constant value in the second image;

[0015] Identify text information in the cell;

[0016] The text information is compared with the fixed value information in the setting fixed value list issued by the power grid dispatching organization to determine whether the fixed values ​​are consistent: if the fixed values ​​are consistent, the text information is stored; if the fixed values ​​are inconsistent, it is marked and the user is prompted to perform manual review.

[0017] Optionally, the image is converted into a grayscale image, and the conversion expression is as follows: including:

[0018]

[0019] Where H and W are the height and width of the input image I; I(r, c) represents the grayscale value of the rth row and cth column of the image; I min is the minimum grayscale in I; I max is the maximum grayscale in I; min , O max ] is the grayscale range of the image to be output.

[0020] Optionally, an adaptive threshold binarization method based on the Wellner algorithm is used to convert the grayscale image into a black and white image. The conversion expression is as follows:

[0021]

[0022] Where n is the number of pixels, P n is a pixel, T(n) is a pixel after the binarization process, s and t are preset constants, and f s (n) is the sum of the first s pixels at point n.

[0023] Optionally, the discontinuous line parts in the black-and-white image are reconnected to generate a second processed image, and the expression is as follows:

[0024]

[0025] Among them, X is the object to be processed, x is the point in X, and B is the structural element.

[0026] Optionally, obtaining four corner points of the outline of the table in the first processed image by using a Harris corner detection algorithm includes:

[0027] Scanning the first processed image using a window function to obtain a grayscale change of each pixel;

[0028] When the grayscale change exceeds a preset threshold, the current pixel is determined to be a corner point.

[0029] Optionally, based on the edge and the corner point, the first processed image is tilt-corrected by perspective transformation to obtain a second processed image, which is expressed as follows:

[0030]

[0031] Wherein, [uvw] is a pixel point on the first processed image, [x1 y1 z1] is a pixel point on the first processed image, is the transformation matrix.

[0032] Optionally, segmenting cells containing a fixed value in the second image includes:

[0033] Obtain the horizontal and vertical lines in the fixed value table, and then perform "AND" processing and "overlay" processing to obtain the table intersection diagram and the cell outline diagram;

[0034] Scan each row and column of the intersection diagram of the table to obtain a set of intersection coordinates;

[0035] Correcting the intersection coordinate set so that the vertical coordinates of the intersections in the same row are the same and the horizontal coordinates of the intersections in the same column are the same, thereby obtaining a reconstructed cell;

[0036] According to the reconstructed cell, cell segmentation is performed from top to bottom and from left to right to obtain the cell.

[0037] The present application also provides a relay protection device setting automatic checking device, comprising:

[0038] An acquisition module, used for acquiring a setting table image of a relay protection device;

[0039] A grayscale module, used for converting the fixed value table image into a grayscale image;

[0040] A black and white module, used for converting the grayscale image into a black and white image by using an adaptive threshold binarization method based on the Wellner algorithm;

[0041] A connection module, used for reconnecting the discontinuous line parts in the black-and-white image to generate a first processed image;

[0042] An edge module, used for detecting the edge of the first processed image using a Canny operator;

[0043] A foot point module, used for obtaining four corner points of the table outline in the first processed image by using a Harris corner point detection algorithm;

[0044] A correction module, configured to perform tilt correction on the first processed image through perspective transformation based on the edge and the corner point to obtain a second processed image;

[0045] A segmentation module, used for segmenting cells containing a fixed value in the second image;

[0046] A recognition module, used for recognizing text information in the cell;

[0047] The judgment module is used to compare the text information with the fixed value information in the setting fixed value list issued by the power grid dispatching organization to determine whether the fixed values ​​are consistent: if the fixed values ​​are consistent, the text information is stored; if the fixed values ​​are inconsistent, it is marked and the user is prompted to perform manual review.

[0048] The present application also provides a relay protection device setting automatic checking device, comprising:

[0049] A memory for storing a computer executable program of the above-mentioned method for automatically checking the setting value of the relay protection device;

[0050] A processor is used to retrieve the computer executable program from the memory and execute: obtaining a set value table image of a relay protection device; converting the set value table image into a grayscale image; using an adaptive threshold binarization method based on the Wellner algorithm to convert the grayscale image into a black and white image; reconnecting the discontinuous line parts in the black and white image to generate a first processed image; using a Canny operator to detect the edge of the first processed image; obtaining four corner points of the table outline in the first processed image by a Harris corner point detection algorithm; based on the edges and corner points, performing tilt correction on the first processed image by perspective transformation to obtain a second processed image; segmenting cells containing set values ​​in the second image; identifying text information in the cells; comparing the text information with the set value information in the set value sheet issued by the power grid dispatching agency to determine whether the set values ​​are consistent: if the set values ​​are consistent, storing the text information; if the set values ​​are inconsistent, marking and prompting the user to perform manual review.

[0051] The present application also provides a storage medium, comprising a computer executable program stored therein, wherein the computer executable program is used to be called by a processor to execute the steps of the above-mentioned method for automatically checking the constant values ​​of a relay protection device.

[0052] The beneficial effects of this application are:

[0053] The present application provides a method for automatically checking the fixed values ​​of a relay protection device, including: obtaining a fixed value table image of the relay protection device; converting the fixed value table image into a grayscale image; using an adaptive threshold binarization method based on the Wellner algorithm to convert the grayscale image into a black and white image; reconnecting the discontinuous line parts in the black and white image to generate a first processed image; using the Canny operator to detect the edge of the first processed image; obtaining the four corner points of the table outline in the first processed image by the Harris corner point detection algorithm; based on the edges and corner points, performing tilt correction on the first processed image by perspective transformation to obtain a second processed image; segmenting the cells containing fixed values ​​in the second image; identifying the text information in the cells; comparing the text information with the fixed value information in the setting fixed value sheet issued by the power grid dispatching agency to determine whether the fixed values ​​are consistent: if the fixed values ​​are consistent, the text information is stored; if the fixed values ​​are inconsistent, the user is marked and prompted to perform manual review. The present application automatically checks the fixed values ​​of relay protection devices based on fixed value table recognition, which improves the accuracy of the verification results compared to traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a schematic diagram of the automatic verification process of the relay protection device settings in this application.

[0055] Figure 2 This is a schematic diagram of a single image for setting a fixed value in this application;

[0056] Figure 3 This is a schematic diagram of the effects of image grayscale enhancement, binarization and morphological processing in this application;

[0057] Figure 4 This is a schematic diagram of the table contour extraction results in this application;

[0058] Figure 5 For this application Figure 5 This is a schematic diagram of an image after tilt correction processing;

[0059] Figure 6 This is a schematic diagram of the table cell segmentation results in this application;

[0060] Figure 7 This is a schematic diagram of the image segmentation results in row 30 of the table in this application;

[0061] Figure 8 This is a schematic diagram of the character recognition results in this application;

[0062] Fig. 9 This is a schematic diagram of the storage results of the fixed value information database in this application;

[0063] Fig.10 It is a schematic diagram of the setting list of two different types of protection devices in this application. DETAILED DESCRIPTION

[0064] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0065] Please refer to Figure 1 As shown, a method for automatically checking the setting value of a relay protection device comprises the following steps:

[0066] S101, obtaining a setting table image of a relay protection device;

[0067] S102, converting the fixed value table image into a grayscale image;

[0068] S103, converting the grayscale image into a black-and-white image by using an adaptive threshold binarization method based on the Wellner algorithm;

[0069] S104, reconnecting the discontinuous line parts in the black-and-white image to generate a first processed image;

[0070] S105, using a Canny operator to detect the edge of the first processed image;

[0071] S106, obtaining four corner points of the outline of the table in the first processed image by using the Harris corner detection algorithm;

[0072] S107, performing tilt correction on the first processed image through perspective transformation based on the edge and the corner point to obtain a second processed image;

[0073] S108, segmenting cells containing constant values ​​in the second image;

[0074] S109, identifying text information in the cell;

[0075] S110. Compare the text information with the set value information in the set value list issued by the power grid dispatching organization to determine whether the set values ​​are consistent: if the set values ​​are consistent, store the text information; if the set values ​​are inconsistent, mark and prompt the user to perform manual review.

[0076] like Figure 2 As shown in (a), the setting value list of the relay protection device obtained by directly photographing it with a mobile phone is a color picture.

[0077] In order to enhance the table information in the image, the photo is first grayscaled, and then the obtained grayscale image is subjected to histogram normalization image enhancement processing, thereby expanding the grayscale range and making the contrast between the foreground content and the background more obvious.

[0078] The mapping relationship expression of the grayscale enhancement processing is:

[0079]

[0080] Wherein, in the mapping relationship expression, H and W are the height and width of the input image I; I(r, c) represents the grayscale value of the rth row and cth column of the image; I min is the minimum grayscale in I; I max is the maximum grayscale in I; min , O max ] is the grayscale range of the image to be output.

[0081] In order to further enhance the foreground image and weaken the background image, the image is subjected to the binarization process again to convert it into a black and white image.

[0082] The technical solution in this embodiment takes into account the limitations of interference factors such as lighting conditions during shooting, and adopts a binarization processing method based on an adaptive threshold of the Wellner algorithm to eliminate the influence of uneven brightness of the photo.

[0083] The binary processing first assumes that the image is a single row connected in sequence by all rows. For the nth point, the pixel is P n After the binarization process, whether the pixel T(n) is 1 (black) or 0 (white) depends on whether it is t percent darker than the average value of the previous s pixels. The corresponding mapping relationship expression is:

[0084]

[0085] T(n)=0Other cases

[0086] The f in the mapping relationship expression s (n) is the sum of the first s pixels at point n, and its expression is:

[0087]

[0088] The setting value sheet of the relay protection device is a printed file. When printing the setting value sheet, the printer often has uneven or detached ink dots, resulting in discontinuity of connected lines and strokes in the image. By performing morphological processing on the image, that is, multiple closing operations of the expansion and corrosion processing, the discontinuous parts that were originally connected to each other are reconnected.

[0089] The expansion corrosion process is characterized in that its mapping relationship expression is:

[0090]

[0091] In the mapping relationship expression, X is the object to be processed, x is a point in X, and B is a structural element.

[0092] Depend on Figure 2 As can be seen in (a), the fixed value list is presented in a table form, and it is inevitable that it will be tilted when photographed by a mobile phone.

[0093] Therefore, it is necessary to first delineate the outline of the table and perform overall tilt correction on the table area.

[0094] In order to obtain the outline of the table in the fixed value list, the Canny operator is selected to perform edge detection on the table, and then the Harris corner point detection principle is used to obtain the four corner points of the table outline.

[0095] The Harris corner detection algorithm uses the window function w(x, y) to scan the image I and obtain the grayscale change E(u, v). If there is a point where the window movement changes greatly in both the horizontal and vertical directions, it is considered to be a corner point.

[0096] The grayscale variation E(u, v) can be expressed as:

[0097]

[0098] Among them, the vector [u, v] represents the window displacement.

[0099] To maximize the value of E(u, v), the second term [I(x+u, y+v)-I(x, y)] 2 The value of should be the largest, and the I(x+u, y+v) is expanded according to Taylor's formula to obtain:

[0100] I(x+u,y+v)≈I(x,y)+uI x (x,y)+vI y (x, y)

[0101] Then the second term [I(x+u, y+v)-I(x, y)] 2 It can be expressed as:

[0102]

[0103] Furthermore, the E(u, v) can be expressed as:

[0104]

[0105] Among them, the eigenvalue of the real symmetric matrix M is solved, and the eigenvectors corresponding to the eigenvalues ​​λ1 and λ2 are the grayscale change directions. If the two eigenvalues ​​are both large, it means that there are two directions where the grayscale changes faster, and it is considered that there are corner points in the window.

[0106] by Figure 3 Take (b) as an example, use the above algorithm to detect the corner points of the table in the figure, connect the four corner points to draw the outline of the table, as shown in Figure 4 shown.

[0107] After the contour detection, the image is tilted and corrected by using perspective transformation. The perspective transformation is an algorithm that projects the image onto a new viewing plane by using the condition that the perspective center, the image point and the target point are collinear.

[0108] Let (u, v) be a point on the original image, corresponding to the point (x, y) on the transformed image, then the expression of the perspective transformation, that is, the perspective transformation coefficient matrix is:

[0109]

[0110] In the perspective transformation coefficient matrix, represents a linear transformation; [a 31 a 32 ] indicates translation; Indicates the perspective transformation is generated, then the perspective transformation coefficient matrix is ​​solved as follows:

[0111] x=x1 / w1=(a 11 u+a 21 v+a 31 ) / (a 13 u+a 23 v+a 33 )

[0112] y=y1 / w1=(a 12 u+a 22 v+a 32 ) / (a 13 u+a 23 v+a 33 )

[0113] Therefore, after determining four groups of one-to-one corresponding points on the original image and the transformed image, namely the four corner points of the table contour through the contour detection, the perspective transformation coefficient matrix can be solved, thereby performing the perspective transformation to achieve the tilt correction.

[0114] right Figure 4 The fixed value single image after tilt correction is as follows Figure 5 shown.

[0115] Cell segmentation and recognition include three parts: cell segmentation, cell recognition and recognition result association correction.

[0116] The cell segmentation:

[0117] After completing the preprocessing of the fixed-value single photo, the obtained table is segmented to segment each cell image. The key is to obtain the horizontal and vertical lines in the table, and then obtain the table intersection map and cell outline map.

[0118] In the present application, the image is subjected to the expansion and corrosion processing based on the morphological processing method, thereby eliminating small element blocks in the image (ie, the text portion in the cell) and retaining the horizontal or vertical line portion.

[0119] The expansion and corrosion processing of the image is based on the principle formula mentioned above. The maximum width of the size of the structure element B must be greater than the maximum width of the text in the cell to obtain the horizontal line in the figure; the maximum height of the size of the structure element B must be greater than the maximum height of the text to obtain the vertical line in the figure.

[0120] The processed horizontal and vertical lines are further subjected to “AND” processing and “overlay” processing to obtain the table intersection diagram and the cell outline diagram.

[0121] Local image Figure 5 For example, Figure 6 Part (a) and Figure 6 The (b) part is the processed horizontal and vertical line graphs of the table, and then the "AND" and "overlay" processing is performed to obtain the table intersection graph and cell outline graph, as shown in the figure. Figure 6 Part (c) and Figure 6 As shown in part (d) of .

[0122] Based on the morphological processing method, the horizontal and vertical lines obtained are not completely horizontal or vertical, and the table intersection diagram needs to be scanned. Figure 6 After obtaining the intersection coordinate set of each row and column of (c), first correct it so that the vertical coordinates of the intersections in the same row are the same, and the horizontal coordinates of the intersections in the same column are the same.

[0123] According to the intersection coordinate set Figure 6 (d) Reconstructed cell, cell segmentation is performed from top to bottom and from left to right.

[0124] By processing the image according to the above steps, accurate segmentation of complex cells can be completed.

[0125] Figure 7This is the segmentation result of the image of the 30th row of the entire fixed value table. There is a situation where the upper and lower cells are merged, but it is still correctly segmented into 6 cell images from No. 131 to No. 136. The two sets of coordinates below the cell are the coordinates of the upper left corner and lower right corner of the cell. It can be seen that the vertical coordinates of the vertices in the same row are exactly the same.

[0126] The cell identification:

[0127] After performing the cell segmentation on the fixed value single image, the Tesseract-OCR method is called to recognize the text in each cell image. The recognition process and principle of the Tesseract-OCR method mainly include:

[0128] Perform image layout analysis on character images to separate the text, background and other contents of the image, and then extract the character area;

[0129] Conduct connected domain analysis on the extracted character regions, roughly divide the characters into block regions, and then locate the text lines through character contour detection;

[0130] Find the baseline based on the text line and cut the text into characters by detecting the fixed spacing;

[0131] By training the adaptive classifier character by character and comparing it with the character library based on continuous learning of character features, more accurate judgment results can be obtained.

[0132] Since the Tesseract-OCR method currently does not have an official character library specifically for protecting the fixed value list, the recognition effect for some characters in the fixed value list is not ideal. At the same time, for the recognition of the fixed value list, the most important thing is to ensure the recognition accuracy of the fixed value value. However, due to the limitations of the printing and photographing conditions of the fixed value list, if the universal character library in the Tesseract-OCR method is used to recognize the Arabic numerals in the fixed value list, there are often recognition errors or omissions of decimal points. Therefore, this application uses the character increment calling tool jTessBoxEditor, based on the fixed value list used in the past in this area, to establish a local character set for autonomous training, thereby significantly improving the recognition effect.

[0133] Recognition result association correction:

[0134] After performing the cell recognition, in order to further improve the recognition accuracy, it is necessary to perform correlation correction on the recognition result.

[0135] For the identification result of the set value, firstly, check whether the value of the item is within the corresponding reasonable range; secondly, compare it with the previous set value to check whether there is a large degree of deviation in the identification result. In this way, it is judged whether the value is credible, and a "warning" signal is given for the set value in question.

[0136] For the text characters, considering the professionalism of relay protection, specific keywords often appear in the setting, such as "phase distance" and "grounding distance" in line protection, "differential protection" and "failure protection" in busbar protection, and "voltage" and "current" in various protections; at the same time, for a certain type of setting, the corresponding unit of the setting value is also fixed, for example, the unit of voltage setting is "V", and the unit of current setting is "A". Based on these characteristics, a keyword library is established, and the recognition results are further corrected by associating keywords.

[0137] Please refer to Figure 7 As shown, text recognition is performed on the images after the cell segmentation is completed. Figure 8 Part (a) is the recognition result using the default font library. Figure 8 Part (b) shows the recognition result after local character training and result association correction. It can be seen that the recognition accuracy is significantly improved.

[0138] In order to retain the recognition results, the fixed value results obtained by recognition are written into the database for storage.

[0139] Please refer to Figure 7 As shown in the figure, for the setting value list, a row of the setting value contains information such as serial number, name, unit, original value, new value, etc. Therefore, based on these characteristics, the recognition results are scanned line by line to determine the location (row and column) of the setting value information and store it in the database. Figure 7 The storage result of the setting value of this row in the database is as follows Fig. 9 shown.

[0140] The fixed value database storage is characterized by Fig.10 Part (a) and Fig.10 As can be seen in part (b), although they are all busbar protection devices, the layout of their device setting sheets is not exactly the same due to different manufacturers and models. Therefore, it is necessary to first extract the protection model according to the setting sheet, and then customize the processing flow of the device setting sheet image according to different models.

[0141] for Fig.10 (a), since the fixed value name, fixed value amount and unit characters are connected together, the characters need to be separated before writing into the database. Fig.10In part (b), we can see that the layout of the table frame lines is incomplete, so we need to fill in the table frame lines based on the brightness contrast between different rows. The other image processing and database storage processes are consistent with the above-mentioned setting value single processing process.

[0142] The fixed value verification is characterized in that after the information on the setting fixed value sheet and the device fixed value sheet are stored in the database, the fixed value data can be read to perform automatic verification.

[0143] In the above-mentioned fixed value verification, it should be noted that: for the setting value list, the setting value includes the original setting value and the new setting value; for the device setting value list, the setting value usually also includes different groups of setting values ​​such as Group 1 and Group 2. In addition, even if they point to the same item setting value, the setting value names on the two setting value lists are not exactly the same. For example, the "ratio differential high threshold" in the setting value list is described as "differential current threshold high value" in the device setting value list.

[0144] Therefore, the fixed value verification includes the following steps:

[0145] For the setting value list, the setting value shall give priority to the new setting value. If no new setting value is given, the original setting value shall be selected.

[0146] For the device setting list, select the group setting currently in use according to the control word.

[0147] The set value names in the two set value lists are matched according to the established keyword library.

[0148] Compare the setting values ​​of the successfully paired ones. If the setting values ​​are the same, it will be displayed as "normal", if they are different, it will be reported as "warning"; other situations are considered to be unrecognizable and an "abnormal" signal will be reported.

[0149] Although the present invention is disclosed as above by the embodiments, it is not intended to limit the protection scope of the present invention. Any changes and modifications made by any technician familiar with the technology without departing from the concept and scope of the present invention should fall within the protection scope of the present invention.

Claims

1. A method for automatically checking the setting value of a relay protection device. It is characterized in that include: Get the setting table image of the relay protection device; Converting the fixed value table image into a grayscale image; The grayscale image is converted into a black-and-white image using an adaptive threshold binarization method based on the Wellner algorithm; reconnecting discontinuous line portions in the black-and-white image to generate a first processed image; Detecting the edge of the first processed image using a Canny operator; Obtaining four corner points of the table outline in the first processed image by using the Harris corner detection algorithm; Based on the edges and corner points, performing tilt correction on the first processed image by perspective transformation to obtain a second processed image; segmenting cells containing a constant value in the second image; Identify text information in the cell; Compare the text information with the fixed value information in the setting fixed value list issued by the power grid dispatching organization to determine whether the fixed values ​​are consistent: if the fixed values ​​are consistent, store the text information; If the set values ​​are inconsistent, mark and prompt the user to perform manual review.

2. According to the automatic checking method of relay protection device setting value according to claim 1, It is characterized in that The image is converted into a grayscale image, and the conversion expression is as follows: including: Where H and W are the height and width of the input image I; I(r, c) represents the grayscale value of the rth row and cth column of the image; I min is the minimum grayscale in I; I max is the maximum grayscale in I; min , O max ] is the grayscale range of the image to be output.

3. According to the automatic checking method of relay protection device setting value as described in claim 1, It is characterized in that The grayscale image is converted into a black-and-white image using an adaptive threshold binarization method based on the Wellner algorithm. The conversion expression is as follows: Where n is the number of pixels, P n is a pixel, T(n) is a pixel after the binarization process, s and t are preset constants, and f s (n) is the sum of the first s pixels at point n.

4. According to the automatic checking method of relay protection device setting value according to claim 1, It is characterized in that The discontinuous line parts in the black and white image are reconnected to generate a second processed image, which is expressed as follows: Among them, X is the object to be processed, x is the point in X, and B is the structural element.

5. According to the automatic checking method of relay protection device setting value according to claim 1, It is characterized in that The four corner points of the table outline in the first processed image are obtained by using the Harris corner detection algorithm, including: Scanning the first processed image using a window function to obtain a grayscale change of each pixel; When the grayscale variation exceeds a preset threshold, the current pixel is determined to be a corner point.

6. According to the automatic checking method of relay protection device setting value as described in claim 1, It is characterized in that Based on the edge and the corner point, the first processed image is tilt-corrected by perspective transformation to obtain a second processed image, which is expressed as follows: Wherein, [uvw] is the pixel point on the first processed image, [x 1 y 1 z 1 ] is a pixel point on the first processed image, is the transformation matrix.

7. According to the automatic checking method of relay protection device setting value as described in claim 1, It is characterized in that Segmenting cells containing a fixed value in the second image includes: Obtain the horizontal and vertical lines in the fixed value table, and then perform "AND" processing and "overlay" processing to obtain the table intersection diagram and the cell outline diagram; Scan each row and column of the intersection diagram of the table to obtain a set of intersection coordinates; The intersection coordinate set is corrected so that the vertical coordinates of the intersections in the same row are the same and the horizontal coordinates of the intersections in the same column are the same, thereby obtaining a reconstructed cell; According to the reconstructed cell, cell segmentation is performed from top to bottom and from left to right to obtain the cell.

8. An automatic checking device for setting value of relay protection device, It is characterized in that include: An acquisition module, used for acquiring a setting table image of a relay protection device; A grayscale module, used for converting the fixed value table image into a grayscale image; A black and white module, used for converting the grayscale image into a black and white image by using an adaptive threshold binarization method based on the Wellner algorithm; A connection module, used for reconnecting the discontinuous line parts in the black-and-white image to generate a first processed image; An edge module, used for detecting the edge of the first processed image using a Canny operator; A foot point module, used for obtaining four corner points of the table outline in the first processed image by using a Harris corner point detection algorithm; A correction module, configured to perform tilt correction on the first processed image through perspective transformation based on the edge and the corner point to obtain a second processed image; A segmentation module, used for segmenting cells containing a fixed value in the second image; A recognition module, used for recognizing text information in the cell; A judgment module, used for comparing the text information with the fixed value information in the setting fixed value list issued by the power grid dispatching organization, and judging whether the fixed values ​​are consistent: if the fixed values ​​are consistent, storing the text information; If the set values ​​are inconsistent, mark and prompt the user to perform manual review.

9. An automatic checking device for setting value of relay protection device, It is characterized in that include: A memory for storing a computer executable program of the method for automatically checking the setting value of a relay protection device according to any one of claims 1 to 7; The processor is used to retrieve the computer executable program from the memory and execute: obtaining a setting table image of a relay protection device; converting the setting table image into a grayscale image; converting the grayscale image into a black-and-white image using an adaptive threshold binarization method based on the Wellner algorithm; and reconnecting the discontinuous line parts in the black-and-white image to generate a first processed image; Using the Canny operator to detect the edge of the first processed image; using the Harris corner point detection algorithm to obtain four corner points of the table outline in the first processed image; based on the edge and the corner points, performing tilt correction on the first processed image through perspective transformation to obtain a second processed image; Segment the cells containing fixed values ​​in the second image; identify text information in the cells; compare the text information with the fixed value information in the setting fixed value list issued by the power grid dispatching agency to determine whether the fixed values ​​are consistent: if the fixed values ​​are consistent, store the text information; If the set values ​​are inconsistent, mark and prompt the user to perform manual review.

10. A storage medium, It is characterized in that The invention comprises a computer executable program stored therein, wherein the computer executable program is used to be called by a processor to execute the steps of the method for automatically checking the setting values ​​of a relay protection device as described in any one of claims 1 to 7.

Citation Information

Patent Citations

  • Image processing method and device, electronic equipment and medium

    CN111444770A

  • Urban road signal lamp pole inclination fault detection method

    CN112686956A

  • Rapid digitization method and system for meteorological paper form document

    CN112818785A

  • Power grid monitoring system detection method and device, computer equipment and medium

    CN113411573A

  • Relay protection setting value automatic checking and correcting device and method based on intelligent identification technology

    CN114172112A