Pressing plate state checking method, system and device and storage medium

Abstract

The invention discloses a pressing plate state checking method, system and device and a storage medium, and relates to the technical field of power equipment. Pressing plate names, standard switching states, pressing plate template frames and position information of the pressing plates in the pressing plate state disc diagram are extracted; and identifying the pressing plate name, the actual on-off state, the pressing plate detection frame and the position information of each pressing plate in the pressing plate shot image. And combining the position information of each pressing plate template frame and each pressing plate detection frame, and the text similarity between the pressing plate name corresponding to each pressing plate template frame and the pressing plate name corresponding to each pressing plate detection frame to process the pressing plate shot image, and matching the pressing plate template frames and the pressing plate detection frames in the same row and the same column. When a shot image of the pressing plate is blurred or shielded, the pressing plate template frame and the pressing plate detection frame can be accurately in one-to-one correspondence, and support is provided for state checking of the pressing plate. And automatically checking the state of the pressing plate according to whether the standard switching state of the pressing plate template frame corresponding to the pressing plate is consistent with the actual switching state of the pressing plate detection frame or not.

Description

Technical Field

[0001] The present invention relates to the technical field of electric power equipment, and in particular to a pressure plate status verification method, system, device and storage medium. Background Art

[0002] As the demand for electricity continues to rise, the importance of safe and stable operation of substations has become increasingly prominent. The secondary protection screen in the substation can quickly cut off the faulty equipment when a fault occurs in the power system to ensure the safe operation of the substation. The pressure plate is an important component in the secondary protection screen. Whether its activation and deactivation status is normal directly affects whether the function of the secondary protection screen is effective. Therefore, it is very necessary to check the status of the pressure plate.

[0003] Currently, manual verification is often used to verify the status of the pressure plates. However, there are many pressure plates in the substation, and the manual verification method has the defects of high labor intensity, low work efficiency, and easy omissions and misjudgments.

[0004] Another way is to use image recognition technology to identify the name and actual state of the pressure plate from the pressure plate image taken by the camera, and check the pressure plate state by judging whether the actual pressure plate state is consistent with the standard pressure plate state corresponding to the pressure plate of the name. However, due to the influence of many factors such as shooting angle, lighting, and obstacles, the recognition accuracy of the pressure plate name may be affected, and it is impossible to accurately find the standard pressure plate state corresponding to the pressure plate, resulting in errors in the pressure plate state verification, affecting the safe operation of the entire substation. Summary of the invention

[0005] The object of the present invention is to provide a platen status verification method, system, device and storage medium, which can automatically realize the platen status verification and ensure the accuracy of the platen status verification even when the image captured by the platen is blurred or blocked.

[0006] In order to solve the above technical problems, the present invention provides a method for checking the state of a pressing plate, comprising: Extract the pressure plate name, standard activation and deactivation status, pressure plate template frames corresponding to each pressure plate, and position information of each pressure plate template frame of each pressure plate of the substation from the obtained pressure plate status panel diagram; Identify the platen name, actual insertion and retraction status, platen detection frames corresponding to each platen, and position information of each platen detection frame contained in the platen shooting image; According to the position information of each platen template frame, the position information of each platen detection frame, and the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, the platen captured image is processed so that the number of rows and columns of the platen detection frame included in the platen captured image is consistent with the number of rows and columns of the platen template frame included in the platen status disk image, and the platen template frame and the platen detection frame corresponding to the same row and column are matched; For each pressing plate, whether the insertion and retraction state of the pressing plate is normal is determined based on whether the actual insertion and retraction state of the pressing plate detection frame corresponding to the pressing plate is consistent with the standard insertion and retraction state of the pressing plate template frame corresponding to the pressing plate.

[0007] Optionally, processing the platen shot image according to the position information of each platen template frame, the position information of each platen detection frame, and the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame includes: According to the position information of each platen template frame, each platen template frame located in the same row in the platen status disk diagram is divided into the same template row, and the total number of template rows of the platen status disk diagram is determined; According to the position information of each platen detection frame, each platen detection frame located in the same row in the platen captured image is divided into the same detection row, and the total number of detection rows of the platen captured image is determined; When the total number of template rows is less than the total number of detection rows, according to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, the detection rows that meet the row matching condition are determined from each detection row, and other detection rows that do not meet the row matching condition are eliminated, so that the total number of detection rows of the platen captured image is equal to the total number of template rows; For a template row and a detection row corresponding to the same row, when the number of platen template frames included in the template row is less than the number of platen detection frames included in the detection row, according to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, each platen detection frame that meets the detection frame matching condition is determined from the detection row, and other platen detection frames that do not meet the detection frame matching condition are eliminated, so that the number of platen detection frames included in each detection row of the platen captured image is equal to the number of platen template frames included in the template row corresponding to the same row; When the number of rows and columns of the platen template frame included in the platen status disk image is exactly the same as the number of rows and columns of the platen detection frame included in the platen captured image, the platen template frame and platen detection frame corresponding to the same row and column will be matched.

[0008] Optionally, determining the detection row satisfying the row matching condition from each detection row according to the text similarity between the pressure plate name corresponding to each pressure plate template frame and the pressure plate name corresponding to each pressure plate detection frame includes: According to the platen names of the platens corresponding to each platen template frame and the platen names of the platens corresponding to each platen detection frame, a CTC text similarity product matrix is ​​determined; wherein the number of rows of the CTC text similarity product matrix is ​​equal to the total number of template rows, the number of columns of the CTC text similarity product matrix is ​​equal to the total number of detection rows, and each element in the CTC text similarity product matrix represents the product of each CTC text similarity between the platen names corresponding to each platen template frame in the template row and the platen names corresponding to each platen detection frame in the detection row; A dynamic programming algorithm is used to select detection rows that meet the row matching condition from the detection rows included in the image captured by the pressure plate, wherein the number of selected detection rows is equal to the total number of template rows, and the row matching condition is that the total matching value between the selected detection rows and all template rows is the maximum, wherein the total matching value is the product of the detection rows selected by the total matching value and the corresponding elements of each template row in the CTC text similarity product matrix.

[0009] Optionally, determining each platen detection frame satisfying the detection frame matching condition from the detection row according to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame includes: For each platen template frame in the current template row, according to the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, determine the CTC text similarity between the platen template frame and each platen detection frame located in the same row, and obtain a CTC text similarity matrix; The number of rows of the CTC text similarity matrix is ​​the number of platen template frames included in the current template row, the number of columns of the CTC text similarity matrix is ​​the number of platen detection frames included in the detection row located in the same row as the current template row, and each element of the CTC text similarity matrix represents the CTC text similarity between the platen name corresponding to the platen template frame and the platen name corresponding to the platen detection frame; A dynamic programming algorithm is used to select a platen detection frame that meets the detection frame matching condition from each platen detection frame, wherein the number of selected platen detection frames is equal to the number of platen template frames included in the current template row, and the detection frame matching condition is that the product of each CTC text similarity between the platen name corresponding to the selected platen detection frame and the platen name corresponding to each platen template frame in the current template row is maximized.

[0010] Optionally, after determining the total number of detection rows and the total number of template rows, the following is further included: When the total number of template rows is greater than the total number of detection rows, or when the number of platen template frames included in the template row corresponding to the same row is greater than the number of platen detection frames included in the detection row, prompt information is generated to prompt the re-shooting of the platen image.

[0011] Optionally, extracting the pressure plate name, standard activation and deactivation status, pressure plate template frames corresponding to each pressure plate, and location information of each pressure plate template frame of each pressure plate of the substation from the obtained pressure plate status disk diagram includes: Extracting a platen template frame that meets a preset parsing condition from the platen status disk diagram, wherein each platen template frame must contain a set of symbols for indicating a standard insertion and withdrawal state of a platen and a text box for indicating a platen name; For each pressure plate template frame, the text in the text box within the pressure plate template frame is used as the pressure plate name of the pressure plate corresponding to the pressure plate in the pressure plate template frame, the standard investment and withdrawal status of the corresponding pressure plate is determined according to the symbol in the pressure plate template frame, and the coordinates of the preset points of the pressure plate template frame are used as the position information of the pressure plate template frame; wherein, the standard investment and withdrawal status includes an investment status and an exit status.

[0012] Optionally, the identification of the platen name, actual insertion and withdrawal status, platen detection frames corresponding to each platen, and position information of each platen detection frame contained in the platen shooting image includes: The target neural network is used to identify the image of the pressing plate to determine the actual insertion and withdrawal status of each pressing plate, the corresponding pressing plate nameplate, the corresponding pressing plate detection frame and the position information of each pressing plate detection frame; The actual insertion and withdrawal states include an insertion state and an exit state, and the pressure plate detection frame corresponding to each pressure plate includes an appearance image of the pressure plate and a pressure plate nameplate corresponding to the pressure plate, and the coordinates of the preset points of each pressure plate detection frame are used as the position information of the pressure plate detection frame; The text recognition neural network is used to identify the nameplates of each pressure plate and determine the pressure plate name of each pressure plate.

[0013] In order to solve the above technical problems, the present invention also provides a platen status verification system, comprising: A panel diagram information extraction unit is used to extract the pressure plate name, standard activation and deactivation status, pressure plate template frames corresponding to each pressure plate, and position information of each pressure plate template frame of each pressure plate of the substation from the obtained pressure plate status panel diagram; A platen image information extraction unit, used to identify the platen name, actual insertion and withdrawal status, platen detection frames corresponding to each platen, and position information of each platen detection frame contained in the platen shooting image; A matching unit, for processing the platen captured image so that the number of rows and columns of the platen detection frame included in the platen captured image is consistent with the number of rows and columns of the platen template frame included in the platen status disk image, and matching the platen template frame and the platen detection frame corresponding to the same row and column, based on the position information of each platen template frame, the position information of each platen detection frame, and the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame; The state detection unit is used to determine whether the insertion and retraction state of each pressure plate is normal based on whether the actual insertion and retraction state of the pressure plate detection frame corresponding to the pressure plate is consistent with the standard insertion and retraction state of the pressure plate template frame corresponding to the pressure plate.

[0014] In order to solve the above technical problems, the present invention also provides a platen status checking device, comprising: Memory for storing computer programs; A processor is used to implement the steps of any of the above-mentioned platen status verification methods when executing the computer program.

[0015] In order to solve the above technical problem, the present invention further provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned platen status verification methods are implemented.

[0016] The beneficial effects of the present invention are to provide a platen status verification method, system, device and storage medium, including extracting the platen name, standard insertion and withdrawal status, platen template frame and its position information of each platen in the platen status disk map. Then identify the platen name, actual insertion and withdrawal status, platen detection frame and its position information of each platen in the platen shooting image. Combined with the position information of each platen template frame and the platen detection frame, as well as the text similarity between the platen name corresponding to each platen template frame and the platen name corresponding to each platen detection frame, the platen shooting image is processed so that the number of rows and columns of the platen detection frame included in the platen shooting image is completely consistent with the number of rows and columns of the platen template frame included in the platen status disk map, and the platen template frame and the platen detection frame corresponding to the same row and column are matched. When the platen shooting image is blurred or blocked, the platen template frame and the platen detection frame can be accurately matched one by one, providing strong support for the subsequent platen status verification. Finally, the platen status is automatically checked based on whether the standard insertion and withdrawal status of the platen template frame corresponding to the platen is consistent with the actual insertion and withdrawal status of the platen detection frame. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the prior art and the drawings required for use in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 The present invention provides a flow chart of a method for checking the state of a pressure plate.

[0019] Figure 2 A plate status disk diagram provided by the present invention.

[0020] Figure 3 A pressing plate photographing image provided by the present invention.

[0021] Figure 4 The present invention provides a captured image of a pressing plate after removing erroneous detection rows and erroneous pressing plate detection frames.

[0022] Figure 5 Another flow chart of a method for checking the state of a pressure plate provided by the present invention.

[0023] Figure 6 A structural schematic diagram of a pressure plate status verification system provided by the present invention.

[0024] Figure 7 The present invention is a schematic structural diagram of a pressure plate status checking device provided by the present invention. DETAILED DESCRIPTION

[0025] The core of the present invention is to provide a platen status verification method, system, device and storage medium, which can automatically realize the platen status verification and ensure the accuracy of the platen status verification even when the image of the platen is blurred or blocked.

[0026] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] Please refer to Figure 1 , Figure 1 A flow chart of a method for checking the state of a pressing plate provided by the present invention, the method comprising: S101. Extract the pressure plate name, standard activation and deactivation status, pressure plate template frames corresponding to each pressure plate, and location information of each pressure plate template frame of each pressure plate of the substation from the obtained pressure plate status panel diagram.

[0028] Platen status panel Figure 1 It is generally provided by equipment manufacturers or power companies, etc., and records the names of the pressure plates in the substation, as well as the standard throw-in and throw-out states corresponding to the pressure plates when the substation is operating normally. In order to facilitate the subsequent check of the pressure plate state, the pressure plate name and the pressure plate position can be combined to determine whether the standard throw-in and throw-out states corresponding to the same pressure plate are consistent with the actual throw-in and throw-out states. The present invention also divides the pressure plate template frames corresponding to each pressure plate from the pressure plate state disk diagram, and reflects the position information of the pressure plate through the position information of the pressure plate template frame.

[0029] Please refer to Figure 2 , Figure 2 The present invention provides a platen status disk diagram, which records relevant information corresponding to the platens in two rows and eight columns, including the platen names of the platens and symbols used to indicate the standard insertion and withdrawal status of each platen. In the disk diagram, the platen is usually represented by two circles aligned vertically, and two vertical lines are combined to indicate the standard insertion and withdrawal status of the platen. When two parallel vertical lines are included, the standard insertion and withdrawal status of the platen is the insertion status; otherwise, the standard insertion and withdrawal status of the platen is the withdrawal status. Figure 2 Taking the disk diagram shown as an example, the pressure plate template frame corresponding to each pressure plate includes a pressure plate name text box of the pressure plate and a group of symbols used to represent the standard insertion and withdrawal state of the pressure plate. The position information of the pressure plate template frame can be the coordinates of its center point or other arbitrary points and the length and width information, etc., and the present invention does not make any special limitation on this.

[0030] Please refer to Figure 5 , Figure 5 Another flow chart of a platen status verification method provided by the present invention. The information extracted from the platen status disk map can also be stored in a database so as to be retrieved when the platen status verification is performed. Furthermore, unless the platen status disk map is updated, it is not necessary to re-extract the information in the platen status disk map each time the platen status verification method is executed.

[0031] As for how to extract the pressure plate name, standard insertion and withdrawal status, pressure plate template frame and its position information of each pressure plate from the pressure plate disk map, subsequent embodiments will explain it and will not be repeated here.

[0032] S102, identifying the platen name, actual insertion and withdrawal status, platen detection frames corresponding to each platen, and position information of each platen detection frame of each platen contained in the platen shooting image.

[0033] Please refer to Figure 3 , Figure 3 The present invention provides a press plate photographed image. The appearance of the press plate when its actual insertion and retraction state is the insertion state is obviously different from the appearance of the press plate when its actual insertion and retraction state is the retraction state. Therefore, the actual insertion and retraction state of each press plate can be determined by identifying the press plate photographed image. Figure 3 As shown, each pressure plate corresponds to a pressure plate nameplate, and the pressure plate name of each pressure plate can be determined by identifying the pressure plate nameplate.

[0034] Similarly, in order to facilitate the subsequent verification of the pressure plate status, the pressure plate name and pressure plate position can be combined to determine whether the standard insertion and withdrawal status and the actual insertion and withdrawal status corresponding to the same pressure plate are consistent. The present invention also divides the pressure plate captured image into pressure plate detection frames corresponding to each pressure plate, and the position information of the pressure plate is reflected by the position information of the pressure plate detection frame.

[0035] The specific method of identifying the pressing plate name, actual insertion and withdrawal status, and each pressing plate detection frame and its position information from the pressing plate photographed image will be described in subsequent embodiments and will not be elaborated here.

[0036] S103. Based on the position information of each platen template frame, the position information of each platen detection frame, and the text similarity between the platen name of each platen corresponding to each platen template frame and the platen name of each platen detection frame, the platen captured image is processed so that the number of rows and columns of the platen detection frame included in the platen captured image is consistent with the number of rows and columns of the platen template frame included in the platen status disk image, and the platen template frame and platen detection frame corresponding to the same row and column are matched.

[0037] According to the position information of the platen template frame, the row and column information of the platen template frame in the platen status disk image can be determined (that is, the row and column of the platen template frame in the platen status disk image), and the row and column information of the corresponding platen can be determined. According to the position information of the platen detection frame, the row and column information of the platen detection frame in the platen captured image can be determined, and the row and column information of the platen corresponding to the platen detection frame can be determined.

[0038] Theoretically, when the number of rows and columns of the platen detection rows included in the platen shooting image are exactly the same as the number of rows and columns of the platen template rows included in the platen status disk diagram, it can be considered that each platen detection frame in the platen shooting image corresponds one-to-one to each platen template frame in the platen status disk diagram, and the platen template frame and platen detection frame with consistent row information and column information (that is, the platen template frame and platen detection frame corresponding to the same row and column) correspond to the same platen, and the two match each other.

[0039] Therefore, if the number of rows and columns of the platen detection row included in the platen captured image is determined to be completely consistent with the number of rows and columns of the platen template row included in the platen status disk image based on the position information of each platen template frame and the position information of each platen detection frame, there is no need to process the platen captured image, and the platen template frame and platen detection frame corresponding to the same row and column are directly matched, and the subsequent platen status verification steps are performed. In addition, the method of matching the platen template frame and the platen detection frame according to the position information avoids the inaccurate problem of checking the platen status according to the platen name when the platen captured image is blurred or blocked, thereby providing strong support for the subsequent platen status verification.

[0040] Of course, in actual applications, the platen image may include other components besides the platen, resulting in an erroneous platen detection frame. In this case, the number of rows of the platen detection frame included in the platen image may be different from the number of rows of the platen template frame included in the platen status disk image, or the total number of platen detection frames corresponding to the same row may be different from the total number of platen template frames. In this case, the platen template frame and the platen detection frame cannot be accurately matched only by the position information of each platen template frame and the position information of each platen detection frame.

[0041] Therefore, the present invention further introduces the text similarity between the platen name corresponding to each platen template frame and the platen name corresponding to each platen detection frame, processes the platen shot image, removes the wrong platen detection frame, and makes the number of rows and columns of the platen detection row included in the processed platen shot image completely consistent with the number of rows and columns of the platen template row included in the platen status disk image, and then matches them according to the position information. It can also avoid the inaccurate problem of checking the platen status according to the platen name when the platen shot image is blurred or blocked.

[0042] It can be seen that when there are erroneous platen detection frames in the platen shooting image, based on the position information of each platen template frame, the position information of each platen detection frame and the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, this situation can be identified and the erroneous platen detection frame can be eliminated, providing strong support for subsequent platen status verification.

[0043] In addition, in actual applications, the platen image may miss some platens, in which case the user is prompted to retake the platen image, and after obtaining the new platen image, the steps of the platen status verification method provided by the present invention are executed again. This situation will be described in detail later and will not be described here.

[0044] S104. For each press plate, determine whether the insertion and retraction state of the press plate is normal based on whether the actual insertion and retraction state of the press plate detection frame corresponding to the press plate is consistent with the standard insertion and retraction state of the press plate template frame corresponding to the press plate.

[0045] After matching each platen template frame with each platen detection frame, the actual insertion and withdrawal state of the platen detection frame is consistent with the standard insertion and withdrawal state of the platen template frame corresponding to the platen, and whether the insertion and withdrawal state of the corresponding platen is normal is determined. Specifically, when the actual insertion and withdrawal state corresponding to the platen is consistent with the standard insertion and withdrawal state, the insertion and withdrawal state of the platen is determined to be normal; if not, the insertion and withdrawal state of the platen is determined to be abnormal.

[0046] In summary, the present invention provides a platen status verification method, first extracting the platen name, standard insertion and withdrawal status, platen template frame and its position information of each platen in the platen status disk diagram. Then identify the platen name, actual insertion and withdrawal status, platen detection frame and its position information of each platen in the platen shooting image. Combined with the position information of each platen template frame and the platen detection frame, as well as the text similarity between the platen name corresponding to each platen template frame and the platen name corresponding to each platen detection frame, the platen shooting image is processed so that the number of rows and columns of the platen detection frame included in the platen shooting image is completely consistent with the number of rows and columns of the platen template frame included in the platen status disk diagram, and the platen template frame and platen detection frame corresponding to the same row and column are matched. Even when the platen shooting image is blurred or blocked, the platen template frame and the platen detection frame can be accurately matched one by one, providing strong support for the subsequent platen status verification. Finally, the platen status is automatically checked based on whether the standard insertion and withdrawal status of the platen template frame corresponding to the platen is consistent with the actual insertion and withdrawal status of the platen detection frame.

[0047] In order to ensure that the number of rows and columns of the platen detection frame included in the platen shooting image is consistent with the number of rows and columns of the platen template frame included in the platen status disk image, it is necessary to eliminate the erroneous platen detection frames in the platen shooting image. The process is described below.

[0048] As an optional embodiment, processing the platen shot image according to the position information of each platen template frame, the position information of each platen detection frame, and the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame includes: According to the position information of each platen template frame, each platen template frame located in the same row in the platen status disk diagram is divided into the same template row, and the total number of template rows in the platen status disk diagram is determined; According to the position information of each platen detection frame, each platen detection frame located in the same row in the platen captured image is divided into the same detection row, and the total number of detection rows of the platen captured image is determined; When the total number of template rows is less than the total number of detection rows, according to the text similarity between the platen name corresponding to each platen template frame and the platen name corresponding to each platen detection frame, the detection rows that meet the row matching condition are determined from each detection row, and other detection rows that do not meet the row matching condition are eliminated, so that the total number of detection rows of the platen captured image is equal to the total number of template rows; For a template row and a detection row corresponding to the same row, when the number of platen template frames included in the template row is less than the number of platen detection frames included in the detection row, according to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, each platen detection frame that meets the detection frame matching condition is determined from the detection row, and other platen detection frames that do not meet the detection frame matching condition are eliminated, so that the number of platen detection frames included in each detection row of the platen captured image is equal to the number of platen template frames included in the template row corresponding to the same row; When the number of rows and columns of the platen template frame included in the platen status disk image is exactly the same as the number of rows and columns of the platen detection frame included in the platen captured image, the platen template frame and platen detection frame corresponding to the same row and column will be matched.

[0049] First, based on the position information of each platen template frame and the position information of each platen detection frame, it is determined whether there is an erroneous platen detection frame in the platen captured image. Specifically, it can be determined by determining whether the number of rows in the platen captured image is greater than the number of rows in the platen status disk image, and whether the number of columns in the platen captured image is greater than the number of columns in the platen status disk image.

[0050] Specifically, first, according to the position information of each platen template frame, the platen template frames located in the same row in the platen status disk map are divided into the same template row, and the total number of template rows in the platen status disk map is determined. The same is true for the platen shooting image, and each platen detection frame located in the same row is divided into the same detection row, and the total number of detection rows in the platen shooting image is determined.

[0051] As a specific implementation method, the position information of the platen template frame is the coordinates of the specified point of the platen template frame (for example, the coordinates of the center point of the platen template frame), so the platen template frames with the same vertical coordinates are divided into the same template row. The position information of the platen detection frame and the division method of the detection row are similar, which will not be repeated here.

[0052] In addition, in practical applications, the present invention can use the target neural network to identify each platen detection frame in the platen shooting image, but the platen detection frame output by the target neural network is disordered. The present invention uses the position information of each platen detection frame to arrange the disordered platen detection frames.

[0053] If the total number of template rows in the platen status disk image is less than the total number of detection rows in the platen captured image, it proves that there are redundant detection rows in the platen captured image, and each platen detection frame in the detection row is an incorrect platen detection frame. The text similarity between the incorrect platen detection frame and the correct platen detection frame is relatively low. Based on this, the present invention determines the detection rows that meet the row matching conditions from each detection row according to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, and eliminates other detection rows that do not meet the row matching conditions, so that the total number of detection rows in the platen captured image is equal to the total number of template rows. The row matching condition can be set according to actual needs, and the present invention does not make specific limitations. A setting method for the row matching condition is provided later, which is not repeated here.

[0054] If the total number of template rows in the platen status disk image is equal to the total number of detection rows in the platen captured image, but for the template row and detection row corresponding to the same row, the number of platen template frames included in the template row is less than the number of platen detection frames included in the detection row, then it is proved that there is an erroneous platen detection frame in the detection row. Similarly, the text similarity between the erroneous platen detection frame and the correct platen detection frame is relatively low. Based on this, the present invention determines each platen detection frame that meets the detection frame matching condition from the detection row according to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, and eliminates other platen detection frames that do not meet the detection frame matching condition, so that the number of platen detection frames included in each detection row of the platen captured image is equal to the number of platen template frames included in the template row corresponding to the same row. One of the setting methods of the detection frame matching condition is provided later, which will not be repeated here.

[0055] After the above operation, all the wrong platen detection frames in the platen shooting image are removed, so that the number of rows and columns of the platen template frame included in the platen status disk image is completely consistent with the number of rows and columns of the platen detection frame included in the platen shooting image. At this time, the platen template frame and the platen detection frame corresponding to the same row and column are matched to provide a basis for the subsequent platen status verification.

[0056] The following is a description of the process of removing erroneous detection lines in the image captured by the pressure plate.

[0057] As an optional embodiment, according to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, determining the detection row that meets the row matching condition from each detection row includes: According to the platen names of the platens corresponding to each platen template frame and the platen names of the platens corresponding to each platen detection frame, a CTC text similarity product matrix is ​​determined; wherein the number of rows of the CTC text similarity product matrix is ​​equal to the total number of template rows, the number of columns of the CTC text similarity product matrix is ​​equal to the total number of detection rows, and each element in the CTC text similarity product matrix represents the product of each CTC text similarity between the platen names corresponding to each platen template frame in the template row and the platen names corresponding to each platen detection frame in the detection row; A dynamic programming algorithm is used to select detection rows that meet the row matching conditions from the detection rows included in the platen image, where the number of selected detection rows is equal to the total number of template rows, and the row matching condition is that the total matching value between the selected detection rows and all template rows is the largest, where the total matching value is the product of the selected detection rows and the corresponding elements of each template row in the CTC text similarity product matrix.

[0058] Assume that the total number of template rows of the platen status disk image is A, the total number of detection rows of the unprocessed platen image is B, and B is greater than A. According to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, the CTC text similarity product value of each template row and each detection row is determined, that is, A*B CTC text similarity product values ​​are obtained, and they are combined into a CTC text similarity product matrix. The number of rows of the CTC text similarity product matrix is ​​equal to the total number of template rows, and the number of columns of the CTC text similarity product matrix is ​​equal to the total number of detection rows.

[0059] Assume that Score (An*Bn) represents the CTC text similarity product value between the nth template line and the nth detection line, and Score (An*Bn) is equal to the product of the CTC text similarities between each platen template frame in the nth template line and each platen detection frame in the nth detection line. The CTC text similarity between each platen template frame and each platen detection frame is determined based on the platen name corresponding to the platen template frame and the platen name corresponding to the platen detection frame.

[0060] After determining the CTC text similarity product matrix, a dynamic programming algorithm is used to select A detection rows from each detection row so that the total matching value between the A detection rows and the A template rows is maximized. In this embodiment, the row matching condition is that the total matching value between the selected A detection rows and the A template rows is maximized. The total matching value is the product of the corresponding elements in the CTC text similarity product matrix corresponding to the selected detection row and each template row.

[0061] If the number of platen inspection frames included in the inspection row corresponding to the same row is equal to the number of platen verification frames included in the template row, the platen inspection frame and the platen template frame can be matched directly; if they are not equal, continue to execute the subsequent steps of eliminating redundant platen inspection frames.

[0062] In summary, the text similarity between the pressure plate name corresponding to each pressure plate template frame and the pressure plate name corresponding to each pressure plate detection frame is used to eliminate the erroneous detection rows in the pressure plate shooting image. This method has high reliability and lays the basis for subsequent pressure plate status verification.

[0063] The following is a description of the process of removing erroneous pressure plate detection frames in the pressure plate captured image.

[0064] As an optional embodiment, according to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, each platen detection frame satisfying the detection frame matching condition is determined from the detection row, including: For each platen template frame in the current template row, according to the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, determine the CTC text similarity between the platen template frame and each platen detection frame located in the same row, and obtain a CTC text similarity matrix; Among them, the number of rows of the CTC text similarity matrix is ​​the number of pressure plate template frames included in the current template row, the number of columns of the CTC text similarity matrix is ​​the number of pressure plate detection frames included in the detection row located in the same row as the current template row, and each element of the CTC text similarity matrix represents the CTC text similarity between the pressure plate name corresponding to the pressure plate template frame and the pressure plate name corresponding to the pressure plate detection frame; A dynamic programming algorithm is used to select a platen detection frame that meets the detection frame matching condition from each platen detection frame, wherein the number of selected platen detection frames is equal to the number of platen template frames included in the current template row, and the detection frame matching condition is that the product of each CTC text similarity between the platen name corresponding to the selected platen detection frame and the platen name corresponding to each platen template frame in the current template row is maximized.

[0065] For the template row and the detection row corresponding to the same row, it is assumed that the template row includes A platen template frames, and the detection row includes B platen detection frames, and B is greater than A. For each platen template frame in the current template row, according to the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, the CTC text similarity between the platen template frame and each platen detection frame located in the same row is determined, that is, A*B CTC text similarities are obtained, and they are composed into a CTC text similarity matrix. The number of rows of the CTC text similarity matrix is ​​A, and the number of columns is B. Each element of the CTC text similarity matrix represents the CTC text similarity between the platen name corresponding to the platen template frame and the platen name corresponding to the platen detection frame.

[0066] After determining the CTC text similarity matrix, a dynamic programming algorithm is used to select A pressure plate detection frames from each pressure plate detection frame so that the product of the CTC text similarities between these A pressure plate detection frames and the A pressure plate template frames is maximized. These A pressure plate detection frames are the selected pressure plate detection frames that meet the detection frame matching conditions.

[0067] Please refer to Figure 4 and Figure 5 , Figure 4 The image of the press plate after removing the wrong detection line and the wrong press plate detection frame provided by the present invention, Figure 5 Another flow chart of a platen status verification method provided by the present invention. After eliminating the wrong detection rows and wrong platen detection frames, the total number of detection rows of the platen captured image is equal to the total number of template rows of the platen status disk image, and the number of platen detection frames included in the detection rows in the same row is equal to the number of platen template frames included in the template row, and then the platen detection frames and platen template frames in the same row and column can be matched.

[0068] In summary, the text similarity between the pressure plate name corresponding to each pressure plate template frame and the pressure plate name corresponding to each pressure plate detection frame is used to eliminate the erroneous pressure plate detection frames in the pressure plate shooting image. This method has high reliability and lays the basis for subsequent pressure plate status verification.

[0069] As an optional embodiment, after determining the total number of detection rows and the total number of template rows, the method further includes: When the total number of template rows is greater than the total number of detection rows, or when the number of platen template frames included in the template row corresponding to the same row is greater than the number of platen detection frames included in the detection row, prompt information is generated to prompt the re-shooting of the platen image.

[0070] As mentioned above, in actual applications, the platen image may miss some platens. In this case, the total number of template rows of the platen status disk image is greater than the total number of detection rows of the platen image, and / or the number of platen template frames included in the template row corresponding to the same row is greater than the number of platen detection frames included in the detection row. In this case, a prompt message is generated to prompt the user to retake the platen image, so as to prompt the user to retake the platen image, and after obtaining the new platen image, the steps of the platen status verification method provided by the present invention are executed again.

[0071] The specific process of extracting relevant information of each pressure plate from the pressure plate status disk diagram is described below.

[0072] As an optional embodiment, the pressure plate name, standard activation and deactivation status, pressure plate template frames corresponding to each pressure plate, and location information of each pressure plate template frame of each pressure plate of the substation are extracted from the obtained pressure plate status disk map, including: Extracting a platen template frame that meets preset parsing conditions from the platen status disk diagram, wherein the preset parsing conditions include that each platen template frame must contain a set of symbols for indicating the standard insertion and withdrawal states of the platen and a text box for indicating the name of the platen; For each pressure plate template frame, the text in the text box within the pressure plate template frame is used as the pressure plate name of the pressure plate corresponding to the pressure plate in the pressure plate template frame, the standard investment and withdrawal status of the corresponding pressure plate is determined according to the symbol in the pressure plate template frame, and the coordinates of the preset points of the pressure plate template frame are used as the position information of the pressure plate template frame; wherein, the standard investment and withdrawal status includes an investment status and an exit status.

[0073] like Figure 2 In the platen status panel diagram shown in the figure, each platen is represented by two circles aligned vertically, and two vertical lines are combined to represent the standard insertion and withdrawal state of the platen. Each platen also corresponds to a text box. Based on this, in this embodiment, each platen template box must contain a set of symbols for representing the standard insertion and withdrawal state of the platen (for example, Figure 2 The two circles and two vertical lines aligned up and down in the figure) and a text box for indicating the name of the pressure plate are used. This spatial relationship is used as the preset parsing condition to extract the pressure plate template frame corresponding to each pressure plate from the pressure plate status disk diagram, and the interference of other components in the pressure plate status disk diagram is eliminated.

[0074] As a specific implementation method, VBA code can be used to extract the coordinates of the circles corresponding to each pressure plate, the coordinates of each vertical line, and the content and coordinates of each text box from the pressure plate status disk diagram in Excel format, and the pressure plate template frame corresponding to each pressure plate can be extracted according to the above-mentioned preset parsing conditions.

[0075] The following describes a process of identifying relevant information of each press plate from the captured image of the press plate.

[0076] As an optional embodiment, identifying the platen name, actual insertion and retraction status, platen detection frames corresponding to each platen, and position information of each platen detection frame contained in the platen shooting image includes: The target neural network is used to identify the images of the pressing plates to determine the actual insertion and retraction status of each pressing plate, the corresponding pressing plate nameplate, the corresponding pressing plate detection frame and the position information of each pressing plate detection frame; The actual insertion and withdrawal states include the insertion state and the withdrawal state. The pressure plate detection frame corresponding to each pressure plate includes the appearance image of the pressure plate and the pressure plate nameplate corresponding to the pressure plate. The coordinates of the preset points of each pressure plate detection frame are used as the position information of the pressure plate detection frame. The text recognition neural network is used to identify the nameplates of each pressure plate and determine the pressure plate name of each pressure plate.

[0077] Please refer to Figure 3 , Figure 3 A press plate photographed image provided by the present invention. The appearance of the press plate when its actual insertion and withdrawal state is the insertion state is obviously different from its appearance when its actual insertion and withdrawal state is the withdrawal state. Therefore, the present embodiment uses the target neural network to recognize the press plate photographed image to determine the actual insertion and withdrawal state of each press plate, the corresponding press plate nameplate, the corresponding press plate detection frame and the position information of each press plate detection frame. The appearance image of each press plate and the press plate nameplate corresponding to the press plate are respectively divided into different press plate detection frames. The text recognition neural network is used to recognize each press plate nameplate to determine the press plate name of each press plate.

[0078] Please refer to Figure 6 , Figure 6 The present invention provides a structural schematic diagram of a platen status verification system, the platen status verification system comprising: The panel diagram information extraction unit 601 is used to extract the pressure plate name, standard start-stop status, pressure plate template frame corresponding to each pressure plate and position information of each pressure plate template frame of each pressure plate of the substation from the obtained pressure plate status panel diagram; The platen image information extraction unit 602 is used to identify the platen name, actual insertion and withdrawal status, platen detection frames corresponding to each platen, and position information of each platen detection frame contained in the platen shooting image; A matching unit 603 is used to process the platen captured image so that the number of rows and columns of the platen detection frame included in the platen captured image is consistent with the number of rows and columns of the platen template frame included in the platen status disk image, and match the platen template frame and the platen detection frame corresponding to the same row and column according to the position information of each platen template frame, the position information of each platen detection frame, and the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame; The state detection unit 604 is used to determine whether the insertion and retraction state of each pressure plate is normal based on whether the actual insertion and retraction state of the pressure plate detection frame corresponding to the pressure plate is consistent with the standard insertion and retraction state of the pressure plate template frame corresponding to the pressure plate.

[0079] For a detailed introduction to the platen status verification system provided by the present invention, please refer to the above-mentioned embodiment of the platen status verification method, and the present invention will not be elaborated here.

[0080] Please refer to Figure 7 , Figure 7 The present invention provides a structural schematic diagram of a platen status verification device, the platen status verification device comprising: Memory 701, used for storing computer programs; The processor 702 is used to implement any of the steps of the above-mentioned platen status verification method when executing a computer program.

[0081] For a detailed introduction to the pressure plate state checking device provided by the present invention, please refer to the above-mentioned embodiment of the pressure plate state checking method, and the present invention will not be elaborated here.

[0082] The present invention also provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned platen status verification methods are implemented.

[0083] For a detailed introduction to the storage medium provided by the present invention, please refer to the embodiment of the above-mentioned pressure plate status verification method, and the present invention will not be elaborated here.

[0084] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. It should also be noted that in this specification, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, article or device. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the presence of other identical elements in the process, article or device including the elements.

[0085] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be 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 present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for checking the state of a pressure plate, characterized in that: include: Extract the pressure plate name, standard activation and deactivation status, pressure plate template frames corresponding to each pressure plate, and position information of each pressure plate template frame of each pressure plate of the substation from the obtained pressure plate status panel diagram; Identify the platen name, actual insertion and retraction status, platen detection frames corresponding to each platen, and position information of each platen detection frame contained in the platen shooting image; According to the position information of each platen template frame, the position information of each platen detection frame, and the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, the platen captured image is processed so that the number of rows and columns of the platen detection frame included in the platen captured image is consistent with the number of rows and columns of the platen template frame included in the platen status disk image, and the platen template frame and the platen detection frame corresponding to the same row and column are matched; For each pressing plate, whether the insertion and retraction state of the pressing plate is normal is determined based on whether the actual insertion and retraction state of the pressing plate detection frame corresponding to the pressing plate is consistent with the standard insertion and retraction state of the pressing plate template frame corresponding to the pressing plate.

2. The method for checking the state of a pressure plate according to claim 1, characterized in that: The processing of the platen shot image according to the position information of each platen template frame, the position information of each platen detection frame, and the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame comprises: According to the position information of each platen template frame, each platen template frame located in the same row in the platen status disk diagram is divided into the same template row, and the total number of template rows of the platen status disk diagram is determined; According to the position information of each platen detection frame, each platen detection frame located in the same row in the platen captured image is divided into the same detection row, and the total number of detection rows of the platen captured image is determined; When the total number of template rows is less than the total number of detection rows, according to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, the detection rows that meet the row matching condition are determined from each detection row, and other detection rows that do not meet the row matching condition are eliminated, so that the total number of detection rows of the platen captured image is equal to the total number of template rows; For a template row and a detection row corresponding to the same row, when the number of platen template frames included in the template row is less than the number of platen detection frames included in the detection row, according to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, each platen detection frame that meets the detection frame matching condition is determined from the detection row, and other platen detection frames that do not meet the detection frame matching condition are eliminated, so that the number of platen detection frames included in each detection row of the platen captured image is equal to the number of platen template frames included in the template row corresponding to the same row; When the number of rows and columns of the platen template frame included in the platen status disk image is exactly the same as the number of rows and columns of the platen detection frame included in the platen captured image, the platen template frame and platen detection frame corresponding to the same row and column will be matched.

3. The method for checking the state of a pressure plate according to claim 2, characterized in that: The step of determining the detection row satisfying the row matching condition from each detection row according to the text similarity between the platen name corresponding to each platen template frame and the platen name corresponding to each platen detection frame includes: According to the platen names of the platens corresponding to each platen template frame and the platen names of the platens corresponding to each platen detection frame, a CTC text similarity product matrix is ​​determined; wherein the number of rows of the CTC text similarity product matrix is ​​equal to the total number of template rows, the number of columns of the CTC text similarity product matrix is ​​equal to the total number of detection rows, and each element in the CTC text similarity product matrix represents the product of each CTC text similarity between the platen names corresponding to each platen template frame in the template row and the platen names corresponding to each platen detection frame in the detection row; A dynamic programming algorithm is used to select detection rows that meet the row matching condition from the detection rows included in the image captured by the pressure plate, wherein the number of selected detection rows is equal to the total number of template rows, and the row matching condition is that the total matching value between the selected detection rows and all template rows is the maximum, wherein the total matching value is the product of the selected detection rows and the corresponding elements of each template row in the CTC text similarity product matrix.

4. The method for checking the state of a pressure plate according to claim 2, characterized in that: The step of determining each platen detection frame satisfying the detection frame matching condition from the detection row according to the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame comprises: For each platen template frame in the current template row, according to the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame, determine the CTC text similarity between the platen template frame and each platen detection frame located in the same row, and obtain a CTC text similarity matrix; The number of rows of the CTC text similarity matrix is ​​the number of platen template frames included in the current template row, the number of columns of the CTC text similarity matrix is ​​the number of platen detection frames included in the detection row located in the same row as the current template row, and each element of the CTC text similarity matrix represents the CTC text similarity between the platen name corresponding to the platen template frame and the platen name corresponding to the platen detection frame; A dynamic programming algorithm is used to select a platen detection frame that meets the detection frame matching condition from each platen detection frame, wherein the number of selected platen detection frames is equal to the number of platen template frames included in the current template row, and the detection frame matching condition is that the product of each CTC text similarity between the platen name corresponding to the selected platen detection frame and the platen name corresponding to each platen template frame in the current template row is maximized.

5. The method for checking the state of a pressure plate according to claim 2, characterized in that: After determining the total number of detection rows and the total number of template rows, it also includes: When the total number of template rows is greater than the total number of detection rows, or when the number of platen template frames included in the template row corresponding to the same row is greater than the number of platen detection frames included in the detection row, prompt information is generated to prompt the re-shooting of the platen image.

6. The method for checking the state of a pressure plate according to claim 1, characterized in that: The method of extracting the pressure plate name, standard activation and deactivation status, pressure plate template frames corresponding to each pressure plate and position information of each pressure plate template frame of each pressure plate of the substation from the obtained pressure plate status disk diagram includes: Extracting a platen template frame that meets a preset parsing condition from the platen status disk diagram, wherein each platen template frame must contain a set of symbols for indicating a standard insertion and withdrawal state of a platen and a text box for indicating a platen name; For each pressure plate template frame, the text in the text box within the pressure plate template frame is used as the pressure plate name of the pressure plate corresponding to the pressure plate in the pressure plate template frame, the standard investment and withdrawal status of the corresponding pressure plate is determined according to the symbol in the pressure plate template frame, and the coordinates of the preset points of the pressure plate template frame are used as the position information of the pressure plate template frame; wherein, the standard investment and withdrawal status includes an investment status and an exit status.

7. The method for checking the state of a pressure plate according to claim 1, characterized in that: The identification of the platen name, actual insertion and withdrawal status, platen detection frames corresponding to each platen, and position information of each platen detection frame contained in the platen shooting image includes: The target neural network is used to identify the image of the pressing plate to determine the actual insertion and withdrawal status of each pressing plate, the corresponding pressing plate nameplate, the corresponding pressing plate detection frame and the position information of each pressing plate detection frame; The actual insertion and withdrawal states include an insertion state and an exit state, and the pressure plate detection frame corresponding to each pressure plate includes an appearance image of the pressure plate and a pressure plate nameplate corresponding to the pressure plate, and the coordinates of the preset points of each pressure plate detection frame are used as the position information of the pressure plate detection frame; The text recognition neural network is used to identify the nameplates of each pressure plate and determine the pressure plate name of each pressure plate.

8. A platen status verification system, characterized in that: include: A panel diagram information extraction unit is used to extract the pressure plate name, standard activation and deactivation status, pressure plate template frames corresponding to each pressure plate, and position information of each pressure plate template frame of each pressure plate of the substation from the obtained pressure plate status panel diagram; A platen image information extraction unit, used to identify the platen name, actual insertion and withdrawal status, platen detection frames corresponding to each platen, and position information of each platen detection frame contained in the platen shooting image; A matching unit, for processing the platen captured image so that the number of rows and columns of the platen detection frame included in the platen captured image is consistent with the number of rows and columns of the platen template frame included in the platen status disk image, and matching the platen template frame and the platen detection frame corresponding to the same row and column, based on the position information of each platen template frame, the position information of each platen detection frame, and the text similarity between the platen name of the platen corresponding to each platen template frame and the platen name of the platen corresponding to each platen detection frame; The state detection unit is used to determine whether the insertion and retraction state of each pressure plate is normal based on whether the actual insertion and retraction state of the pressure plate detection frame corresponding to the pressure plate is consistent with the standard insertion and retraction state of the pressure plate template frame corresponding to the pressure plate.

9. A device for checking the state of a pressure plate, characterized in that: include: Memory for storing computer programs; A processor, configured to implement the steps of the pressure plate status verification method as described in any one of claims 1 to 7 when executing the computer program.

10. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the pressure plate status verification method according to any one of claims 1 to 7 are implemented.

Images