Method and System for Identifying and Extracting Nuclear Power Plant Drawing Information

Abstract

The present invention relates to a method and system for identifying and extracting drawing information of a nuclear power plant. The method includes: S10, table recognition: identifying the graphics containing table information from the drawings; S20, text recognition and extraction: recognizing and extracting text information from each cell, where the text information includes text style, text content, and text type; S30, text division: during the text extraction process, establishing a rectangular coordinate system, and according to the cell positions, assigning independent coordinate numbers to the text information in the corresponding cells; according to the corresponding coordinate numbers, dividing the text information with the same ordinate into the same information type, and dividing the text information with the same abscissa into different information types of the same spare part; S40, text sorting: establishing a general table, and arranging the extracted text information in order in the general table according to the information type and the corresponding coordinate numbers; which can reduce labor consumption and has high data accuracy.

Description

Technical Field

[0001] The present invention relates to the field of nuclear power, and particularly to a method for identifying and extracting drawing information of a nuclear power plant. Background Art

[0002] In a nuclear power plant, there are approximately 100,000 drawings corresponding to spare parts in the first-phase project. If the information required in the drawings needs to be extracted, the workload is huge. If it is manually extracted item by item from the drawings, a large number of personnel are required to spend nearly 4 years, with extremely high intensity, and there are very high requirements for data accuracy. Otherwise, it may cause suppliers to be unable to identify or spare parts to be mismatched during on-site maintenance. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a method for identifying and extracting drawing information of a nuclear power plant.

[0004] The technical solution adopted by the present invention to solve its technical problem is: constructing a method for identifying and extracting drawing information of a nuclear power plant, including:

[0005] S10. Table recognition: Recognize the graphics containing table information from the drawings. The table information includes table lines, table positions, cells formed by the table lines, cell positions, and text information in each cell.

[0006] S20. Text recognition and extraction: Recognize and extract the text information from each cell. The text information includes text styles, text contents, and text types.

[0007] S30. Text division: During the text extraction process, establish a rectangular coordinate system, and according to the cell positions, assign independent coordinate numbers to the text information in the corresponding cells.

[0008] According to the corresponding coordinate numbers, divide the text information with the same ordinate into the same information type, and divide the text information with the same abscissa into different information types of the same spare part.

[0009] S40. Text sorting: Establish a general table, and sort the extracted text information in order in the general table according to the information type and the corresponding coordinate numbers.

[0010] Preferably, in step S20, it includes the following sub-steps:

[0011] S21: Recognize the text information in each cell of the table through an OCR recognition tool.

[0012] S22: Correct the recognized text information according to general text and / or preset text.

[0013] S23: Extract the text information in each cell after text correction.

[0014] Preferably, in step S30, the following sub-steps are further included:

[0015] S31: Determine whether the text information at the same ordinate is of the same information type; if so, proceed to the next step; if not, perform text content analysis, and classify the text information corresponding to different information types into the correct information type.

[0016] S32: Determine whether the text information at the same abscissa is of different information types of the same spare part; if so, proceed to the next step; if not, perform text content analysis, and classify the text information corresponding to information types that are not of the same spare part into the correct information type.

[0017] Preferably, in step S30, the following sub-steps are further included:

[0018] S33: Determine whether the text information in the same cell contains multiple information types; if so, perform text classification, reclassify the text information in the same cell into the corresponding information types, and assign different vertical coordinates.

[0019] Preferably, in step S30, the following sub-steps are further included:

[0020] S34: The text information in the same cell includes N information types. Among them, if the value of the vertical coordinate corresponding to the first information type is M, then the value of the vertical coordinate corresponding to the nth information type is M + n / N; N and n are positive integers, and n < N.

[0021] Preferably, the value of the coordinate number corresponding to the text information is the coordinate value at the upper right corner of the corresponding cell minus the coordinate value at the upper left corner of it.

[0022] Preferably, after step S40, step S50 may further be included:

[0023] S50. Coding conversion: Compile different information types of the same spare part into a string-type code according to the preset coding rules.

[0024] Preferably, before step S10, step S01 is further included:

[0025] S01. Drawing interception: Intercept a partial position of the drawing at the first position according to the preset table in the drawing, and then perform table recognition.

[0026] Preferably, in step S40, the following sub-steps are further included:

[0027] S41: Select at least two text messages in the same ordinate, and determine whether the selected text messages are of the same message type; if so, determine that the same ordinate is of the same message type; if not, re-identify the text messages and correct the corresponding correct message types.

[0028] The present invention also constructs a nuclear power plant drawing information recognition and extraction system adopting the above nuclear power plant drawing information recognition and extraction method, including:

[0029] Table recognition module: used to recognize the graph containing table information from the drawing;

[0030] Text recognition and extraction module: used to recognize and extract text information from each cell

[0031] Text division module: used to establish a rectangular coordinate system, and assign independent coordinate numbers to the text information in the corresponding cell according to the cell position; and according to the corresponding coordinate numbers, divide the text information in the same ordinate into the same message type, and divide the text information in the same abscissa into different message types of the same spare part;

[0032] Text sorting module: used to recognize the message type of the text information in the same ordinate; and establish a general table, and arrange the extracted text information in order in the general table according to the message type and the corresponding coordinate numbers.

[0033] Implementing the present invention has the following beneficial effects: recognizing and extracting the required information content in the nuclear power plant drawing can reduce the manpower consumption, and the data has high accuracy, reducing the probability of problems such as the supplier being unable to recognize or the spare parts not matching during on-site maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The following will further illustrate the present invention in conjunction with the drawings and embodiments. In the drawings:

[0035] Figure 1 is the program flow chart of the nuclear power plant drawing information recognition and extraction method in the nuclear power plant drawing information recognition and extraction method and system of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0036] For a clearer understanding of the technical features, objectives, and effects of the present invention, the specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings. In the following description, it should be understood that the orientation or positional relationships indicated by "front", "rear", "upper", "lower", "left", "right", "longitudinal", "transverse", "vertical", "horizontal", "top", "bottom", "inner", "outer", "head", "tail", etc. are based on the orientation or positional relationships shown in the accompanying drawings, and are constructed and operated in a specific orientation, solely for the convenience of describing the present technical solution, rather than indicating that the indicated device or element must have a specific orientation. Therefore, it should not be construed as a limitation to the present invention.

[0037] It should also be noted that, unless otherwise clearly defined and limited, terms such as "install", "connect", "link", "fix", "set", etc. should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or integrated; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements or the interaction relationship between two elements. When an element is referred to as being "on" or "under" another element, the element can be "directly" or "indirectly" located above the other element, or there may also be one or more intermediate elements. The terms "first", "second", "third", etc. are only for the convenience of describing the present technical solution, and should not be construed as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first", "second", "third", etc. may explicitly or implicitly include one or more of such features. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0038] In the following description, for the purpose of illustration rather than limitation, specific details such as specific system structures and technologies are proposed to thoroughly understand the embodiments of the present invention. However, those skilled in the art should clearly understand that the present invention can also be implemented in other embodiments without these specific details. In other cases, the detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present invention.

[0039] The present invention provides a method for identifying and extracting nuclear power plant drawing information. By identifying and extracting useful information from the content on the drawing and sorting the extracted information regularly, it can be applied to the production of a spare part coding import form.

[0040] Among them, as Figure 1 shown, the method for identifying and extracting nuclear power plant drawing information includes:

[0041] S10. Table Recognition: Recognize the graphics containing table information from the drawings. The table information includes table lines, table positions, cells formed by the table lines, cell positions, and text information within each cell.

[0042] Understandably, the content of nuclear power plant drawings includes spare part structure diagrams, performance tables, BOM tables, EOMM manual numbers, power plant drawing numbers, manufacturer drawing numbers, manufacturer names, supplier names, and other residual information. The purpose of the present invention is to recognize and extract the BOM tables, EOMM manual numbers, power plant drawing numbers, manufacturer drawing numbers, manufacturer names, and supplier names among them. And this part of the content is usually listed in the form of a table in the drawings. Therefore, through table recognition, the present invention plans the specific positions and performs text recognition at these specific positions to avoid the problem of text confusion.

[0043] Further, nuclear power plant drawings are generally saved in data in file formats such as dwg file format, pdf file format, and picture file format. In some embodiments of the present invention, horizontal and vertical lines with multiple intersection points are recognized and searched in the drawings; or, square graphics are searched in the drawings to determine as a table. It should be noted that if no text information is retrieved in the table, it can be considered an empty table or part of the spare part structure diagram, and it can be ignored at this time.

[0044] Furthermore, the table information includes table lines, table positions, cell information formed by the table lines, cell positions, and text information within each cell. Understandably, a table includes several cells arranged in sequence. After recognizing it as a table, recognize the cells and the text information contained in the corresponding cells, and then extract the text information.

[0045] Optionally, before step S10, there is also step S01 for accelerating the speed of table recognition.

[0046] Specifically, step S01. Drawing Interception: According to the preset first position of the table in the drawing, intercept the partial position of the drawing at the first position, and then perform table recognition.

[0047] Understandably, in the prior art, table information is generally concentrated in the lower right corner of the drawing. By directly cropping part of the drawing, the speed of table recognition can be accelerated. The specific length and width of the first position can be determined according to the actual situation and can be the table format given by the nuclear power plant. In addition, the range of the first position is larger than the size of the entire table to ensure that there is no omission of information.

[0048] S20. Text Recognition and Extraction: Recognize and extract text information from each cell. The text information includes text style and text content.

[0049] Understandably, text information is provided in each cell, and the text information needs to be recognized and extracted for step S30;

[0050] The text information includes text style, text content, and text type; in some embodiments of the present invention, the text information may be a combination of one or more of letters, numbers, symbols, or Chinese characters.

[0051] Specifically, in step S20, the following sub-steps are included:

[0052] S21: Use an OCR recognition tool to recognize the text information in each cell of the table;

[0053] S22: Correct the recognized text information according to general text and / or preset text;

[0054] S23: Extract the text information in each cell after text cleaning;

[0055] Understandably, OCR (optical character recognition) text recognition is the process of translating the shape of the text to be recognized into computer text using character recognition methods; that is, scanning the text materials and then analyzing and processing the image files to obtain the text and layout information; the OCR recognition tool can refer to the prior art and will not be described here.

[0056] The OCR recognition tool can be implemented based on a computer. Before step S21, the OCR recognition tool can be started by an RPA robot to achieve automation.

[0057] Use an OCR recognition tool to recognize the text information in each cell, and correct the recognized text information according to general text and / or preset text; general text refers to text that is prone to misjudgment in general, such as the number character 1 and the letter character l, the number character 0 and the letter character O, which have similar shapes; the preset text can refer to professional vocabulary in the technical field of nuclear power plants or symbols with special shapes, such as ⊙, Δ such unit symbols.

[0058] S30, Text sorting: During the text extraction process, establish a rectangular coordinate system, and assign independent coordinate numbers to the text information in the corresponding cells according to the cell positions;

[0059] According to the corresponding coordinate numbers, divide the text information with the same ordinate into the same information type, and divide the text information with the same abscissa into different information types of the same spare part;

[0060] Understandably, the information type can be standards, manufacturer's item numbers, manufacturer's reference numbers, Chinese descriptions of spare parts, spare part model columns, number of sub-spare parts, materials, manufacturer codes, supplier codes, EOMM manual codes, power plant drawing numbers, power plant drawing version numbers, etc. Specifically, different information types of the same spare part are contained in cells in the same horizontal row in the table; for example, the first horizontal cell contains the number "1" representing the serial number, the second horizontal cell contains the Chinese character "bearing" representing the Chinese description of the spare part, the third horizontal cell contains the string of letters, symbols, and numbers "GB / T292" representing the standard in use, and the fourth horizontal cell contains the number "1" representing the quantity of the spare part; while in the table, the text information of the same information type is contained in cells in the same vertical column; for example, the text content in the vertical cells corresponding to the first horizontal cell all represents the serial number, and so on.

[0061] Establish a rectangular coordinate system, and assign independent coordinate numbers to the text information in each cell of the table. Since in the drawings of nuclear power plants, the lengths of multiple cells in the horizontal direction may be different, which may easily cause problems when setting coordinate numbers, therefore, it can be set that the upper left corner of the cell is used as the coordinate number to correspond to the text information in that cell; of course, using the upper left corner of the cell as the coordinate point is only an embodiment of the present invention, and other ways of setting the cell as the coordinate point should also be included in the protection scope of the present invention. Preferably, since the positions of the cells in the table may be intricate and the cells have different lengths; in order to reduce the error rate of recognition, the value of the coordinate number corresponding to the text information is the coordinate value of the upper right corner of the corresponding cell minus the coordinate value of its upper left corner;

[0062] Extract the text information in each cell after text correction and proceed to the next step.

[0063] Furthermore, in step S30, the following sub-steps are further included:

[0064] S31: Determine whether the text information in the same vertical coordinate is of the same information type; if so, proceed to the next step; if not, perform text content analysis and divide the text information corresponding to the information type different from the same category into the correct information type;

[0065] S32: Determine whether the text information in the same horizontal coordinate is different information types of the same spare part; if so, proceed to the next step; if not, perform text content analysis and divide the text information corresponding to the information type that is not of the same spare part into the correct information type;

[0066] Understandably, in the nuclear power plant drawings, the BOM table can be set above the box containing text content such as the manufacturer's name, or on the horizontal side of the box. At this time, the text information with the same ordinate may not be of the same information type, and the text information with the same abscissa may not be of different information types of the same spare part. Therefore, the text content can be analyzed and judged. If it is found through analysis that the text content contains specific words, the text information corresponding to different information types can be classified into the correct information type. For example, the manufacturer's name generally contains specific words such as "company" or "factory", and at this time, this type of text information can be classified into the column of supplier or manufacturer.

[0067] Furthermore, in step S30, the following sub-steps are also included:

[0068] S33: Determine whether the text information in the same cell contains multiple information types. If so, perform text classification, reclassify the text information in the same cell into the corresponding information types, and assign different vertical coordinates;

[0069] Understandably, in any cell, the text information may contain multiple information types. For example, a cell contains the text information "plug M20x1.25", where "plug" represents the information type of the Chinese description of the spare part, and "M20x1.25" represents the model and size of the plug. At this time, the text information in the cell is reclassified into the corresponding information types and different vertical coordinates are assigned.

[0070] S34: The text information in the same cell includes N information types. Among them, if the value of the vertical coordinate corresponding to the first information type is M, then the value of the vertical coordinate corresponding to the nth information type is M + n / N; N and n are positive integers, and n < N.

[0071] Understandably, if it is set with the upper left corner of the cell as the coordinate point, and the text information in the same cell contains multiple information types, and the vertical coordinate corresponding to the first information type is M, then the vertical coordinate corresponding to the nth information type is M + n / N; N and n are positive integers, and n < N.

[0072] S40, Text sorting: Establish a general table, and arrange the extracted text information in order in the general table according to the information type and the corresponding coordinate number.

[0073] Understandably, after assigning independent coordinate numbers to the text information in the corresponding cell, since the text information with the same vertical coordinate is of the same information type, as long as the information type of any text information in the same vertical coordinate is identified, the information types of all the text information in this vertical coordinate can be judged.

[0074] Preferably, in step S40, the following sub-steps are further included:

[0075] S41: Select at least two text messages in the same ordinate, and determine whether the selected text messages are of the same information type; if so, determine that the same ordinate is of the same information type; if not, re-identify the text messages and correct the corresponding correct information type.

[0076] Furthermore, the information type can be identified by analyzing the text content; for example, in the drawings of nuclear power plants, the text content generally with GB, JB / T, NB / T, etc. is the usage standard. If the text message containing this type of text content is analyzed, it can be identified as the information type of the usage standard. In addition, a preset relationship table can also be created. When the corresponding text or English appears in the text message, it is judged as the corresponding information type; for example, when judging the information type as manufacturer and supplier, since the identified text messages are all text contents composed of characters, at this time, by setting the relationship between the respective corresponding text contents and information types in the preset relationship table, the corresponding information type can be identified when the content in the text message is analyzed.

[0077] Optionally, after step S40, step S50 is further included, specifically:

[0078] S50. Coding conversion: Compile different information types of the same spare part into a string type code according to the preset coding rules.

[0079] It can be understood that in a nuclear power plant, there is a large amount of relevant information about spare parts. If it is compiled into a code, it is convenient for the storage and use of spare part data. Specifically, after identifying the Chinese description of the spare part in the text message, the text messages of other information types in the spare part are converted according to the preset coding rules to obtain the corresponding string type code.

[0080] Based on the same general inventive concept, the present invention also constructs a nuclear power plant drawing information recognition and extraction system adopting the above-mentioned nuclear power plant drawing information recognition and extraction method, including:

[0081] Table recognition module: used to recognize the graphics containing table information from the drawings;

[0082] Text recognition and extraction module: used to recognize and extract text information from each cell

[0083] Text division module: used to establish a rectangular coordinate system, and assign independent coordinate numbers to the text information in the corresponding cell according to the cell position; and according to the corresponding coordinate numbers, divide the text information in the same ordinate into the same information type, and divide the text information in the same abscissa into different information types of the same spare part;

[0084] Text sorting module: used to identify the information types of text information with the same ordinate; and establish a general table, and arrange the extracted text information in order in the general table according to the information type and the corresponding coordinate number.

[0085] It can be understood that the above embodiments only express the preferred embodiments of the present invention, and the description thereof is relatively specific and detailed, but it cannot be construed as a limitation on the scope of the patent of the present invention; it should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, the above technical features can be freely combined, and several deformations and improvements can also be made, all of which belong to the protection scope of the present invention; therefore, all equivalent transformations and modifications made to the scope of the claims of the present invention shall fall within the scope covered by the claims of the present invention.

Claims

1. A method for identifying and extracting drawing information of a nuclear power plant, characterized in that, Including: S10. Table recognition: Recognize the graphics containing table information from the drawings. The table information includes table lines, table positions, cells formed by the table lines, cell positions, and text information within each cell. The content of the drawings includes spare part structure diagrams, performance tables, BOM tables, EOMM manual numbers, power station drawing numbers, manufacturer drawing numbers, manufacturer names, and supplier names. S20. Text recognition and extraction: Recognize and extract text information from each cell. The text information includes text style, text content, and text type. S30. Text division: During the text extraction process, establish a rectangular coordinate system. According to the cell positions, assign independent coordinate numbers to the text information within the corresponding cells. The numerical value of the coordinate number corresponding to the text information is the coordinate value of the upper right corner of the corresponding cell minus the coordinate value of its upper left corner. According to the corresponding coordinate numbers, divide the text information with the same ordinate into the same information type, and divide the text information with the same abscissa into different information types of the same spare part. And analyze whether the text information within the cell contains specific words to divide the corresponding text information to the correct information type. Among them, in step S30, the following sub-steps are further included: S31: Judge whether the text information with the same ordinate is of the same information type. If so, execute the next step; if not, perform text content analysis and divide the text information corresponding to the different information types into the correct information types. S32: Judge whether the text information with the same abscissa is of different information types of the same spare part. If so, execute the next step; if not, perform text content analysis and divide the text information corresponding to the information types that are not of the same spare part into the correct information types. S33: Judge whether the text information within the same cell contains multiple information types. If so, perform text classification, re-divide the text information within the same cell into the corresponding information types, and assign different vertical coordinates. The text information within the same cell includes N information types. Among them, the numerical value of the vertical coordinate corresponding to the first information type is M, then the numerical value of the vertical coordinate corresponding to the nth information type is M + n / N; N and n are positive integers, and n < N. S40. Text sorting: Establish a total table, and arrange the extracted text information in order in the total table according to the information type and the corresponding coordinate numbers.

2. The method for recognizing and extracting nuclear power plant drawing information according to claim 1, characterized in that in step S20, the following sub-steps are included: S21: Recognize the text information within each cell in the table through an OCR recognition tool. S22: Correct the recognized text information according to general text and / or preset text. S23: Extract the text information within each cell after text correction.

3. The method for recognizing and extracting nuclear power plant drawing information according to claim 1, characterized in that after step S40, step S50 may further be included: S50. Coding conversion: Compile different information types of the same spare part into string-type codes according to the preset coding rules.

4. The method for identifying and extracting nuclear power plant drawing information according to claim 1, characterized in that, Before step S10, it further includes step S01: S01. Drawing Interception: According to the first position of the preset table in the drawing, intercept a partial position of the drawing at the first position, and then perform table recognition.

5. The method for identifying and extracting nuclear power plant drawing information according to claim 1, characterized in that, In step S40, it further includes the following sub-steps: S41: Select at least two text information in the same ordinate, and determine whether the selected text information is of the same information type; if so, determine that the same ordinate is of the same information type; if not, re-identify the text information and correct the corresponding correct information type.

6. A nuclear power plant drawing information recognition and extraction system adopting the nuclear power plant drawing information recognition and extraction method described in any one of claims 1-5, characterized in that, It includes: Table Recognition Module: Used to recognize the graphics containing table information from the drawing; Text Recognition and Extraction Module: Used to recognize and extract text information from each cell Text Division Module: Used to establish a rectangular coordinate system, and according to the cell position, assign independent coordinate numbers to the text information in the corresponding cell; and according to the corresponding coordinate numbers, divide the text information in the same ordinate into the same information type, and divide the text information in the same abscissa into different information types of the same spare part; Text Sorting Module: Used to recognize the information type of the text information in the same ordinate; and establish a total table, and arrange the extracted text information in order in the total table according to the information type and the corresponding coordinate numbers.

Images