Method for extracting production data from plate pattern

Abstract

The invention provides a method for extracting production data from a board graph, which comprises the following steps of: obtaining an object set comprising geometric elements, text labels and symbol labels in a design drawing, and comparing object attributes by traversing drawing data to finish preliminary screening to obtain an expanded object set; screening specific objects in the expanded object set according to a preset condition, and identifying geometric elements, text labels and symbol marks which meet a preset type requirement to form a screened object subset; and extracting detailed information of a target object based on the grouped object set, analyzing a geometrical relationship between the text annotation content and the object to complete extraction, and binding the extracted information with a grouping result to form a structured object description.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, and in particular to a plate pattern extraction production data method. BACKGROUND

[0002] In processing the production drawing data of the steel plate of the iron tower, there is a unique and complex technical problem, that is, how to accurately process the correlation between the geometric relationship of the steel plate of the iron tower and the text annotation on the basis of multi-level object screening and grouping, while coping with the standardization problem caused by the inconsistency of external configuration data. The specific scenario is that the production drawing of the steel plate of the iron tower contains a large number of heterogeneous objects, such as geometric elements such as steel plate contour, hole symbol, and weld mark, text annotations such as thickness and material, and special connection hole symbol marks. These objects may have slight deviations in spatial distribution, and the setting of the preset tolerance is difficult to adapt to the diversified structure types of the steel plate of the iron tower, resulting in that part of the logically related objects cannot be correctly grouped. In addition, the text annotation often contains non-standardized production process description of the steel plate of the iron tower, and the external configuration data may have version differences or missing iron tower industry special key fields, which causes the key word matching to appear missing or mismatching, thereby affecting the accuracy of the standardized replacement. More troublesome is that in calculating the physical parameters of the steel plate of the iron tower, the extraction of geometric information may produce errors due to the non-uniformity of the drawing data format, such as incomplete hole symbol boundary definition and missing weld height annotation of some steel plates of the iron tower, resulting in that the calculation result based on the preset formula deviates from the actual value. In the data output stage, if there are abnormal data in the final file that have not been processed, the existing threshold judgment mechanism may be too broad or strict, which cannot effectively distinguish the severity of the abnormality, thereby affecting the priority judgment and data updating efficiency when supplementing the geometric relationship. This problem not only involves the multi-link data processing logic, but also involves the particularity of the drawing data of the steel plate of the iron tower and the dynamics of the external configuration, and a fine processing mechanism is needed to balance the precision and efficiency of screening, grouping, standardization, and abnormality processing, to ensure the integrity and consistency of the final data, and to meet the needs of the production, processing, quality inspection, and installation adaptation of the iron tower. SUMMARY

[0003] The present application provides a plate pattern extraction production data method, mainly comprising: Obtaining an object set including geometric elements, text annotations, and symbol marks in a design drawing, completing preliminary screening by comparing object attributes through traversing the drawing data, and obtaining an expanded object set; Screening specific objects in the expanded object set according to a preset condition, identifying geometric elements, text annotations, and symbol marks that meet the preset type requirements, and forming an object subset after screening; The screened object subset is associated and grouped through a collision detection algorithm, spatial intersection detection is performed based on object boundary ranges, scattered objects are integrated into a logically associated set in combination with a preset tolerance parameter, and a grouped object set is obtained; Detailed information of a target object is extracted based on the grouped object set, extraction is completed by analyzing the geometric relationship between text label content and objects, the extracted information is bound with the grouping result to form a structured object description; A preset external configuration data is read, standardized replacement is performed on a process description field of the structured object description, non-standard expressions are replaced into unified standard terms by matching keywords, and a standardized object description is obtained; Physical parameters are calculated according to geometric information recorded in the standardized object description, and the calculation is completed based on a preset calculation formula and a preset material attribute parameter, the calculation result is integrated with the standardized object description, and an integrated data set is formed; The integrated data set is output to an external storage file in a preset structured format, the output style is adjusted through a preset formatting rule, data verification is performed synchronously during output, abnormal data is marked or corrected, and a final output file is obtained; If there is still unprocessed abnormal data in the final output file, the severity of the abnormality is determined according to a preset abnormality judgment threshold, the missing geometric relationship data is supplemented and the complete information of the related object is re-extracted, and the integrated data set is updated.

[0004] Further, the object set including geometric elements, text labels and symbol marks in the design drawing is obtained by traversing the drawing data and comparing object attributes to complete preliminary screening, and an expanded object set is obtained, including: An object set is obtained from the drawing data by traversing the drawing data, the object set includes geometric elements and text labels and symbol marks, and a preliminary object list is obtained; For the preliminary object list, object attributes are compared, a preset threshold is used to judge attribute matching degree, and a screened object subset is determined through geometric element recognition and attribute matching verification; According to the screened object subset, element position coordinates and label content are analyzed, a subset integrity index is obtained by text label extraction and label content analysis, and the subset integrity is judged; If the subset integrity reaches a preset threshold, symbol mark classification and mark type differentiation are integrated, and an expanded object set is obtained by symbol mark classification and set integration output.

[0005] Further, the specific object in the expanded object set is screened according to a preset condition, geometric elements, text labels and symbol marks meeting the preset type requirements are identified, and a screened object subset is formed, including: acquire a specific object from the expanded object set, compare object attributes through a preset condition to obtain a filtered object list; For the filtered object list, identify geometric elements, and determine an element subset that meets the requirements of a preset type using attribute matching; According to the element subset, extract text labels, judge the matching degree of label content and preset type requirements, and obtain label verification results; Integrate symbolic markers through the label verification results, identify marker types, and form a filtered object subset.

[0006] Further, the filtered object subset is associated and grouped by the collision detection algorithm, spatial intersection detection is performed based on object boundary ranges, and dispersed objects are integrated into logically associated sets based on a preset tolerance parameter to obtain a grouped object set, including: Get boundary range data from the filtered object subset, use spatial intersection detection to judge the overlapping area between objects, and determine the preliminary association pair based on the preset tolerance parameter to obtain an association pair list; Integrate dispersed objects for the association pair list, perform collision detection algorithm comparison of boundary ranges, and obtain logically associated groups; Determine the grouping integrity by extracting geometric features from the logically associated groups and judging feature matching degrees; According to the grouping integrity, fuse the symbolic markers to obtain the final grouped object set.

[0007] Further, based on the grouped object set, detailed information of the target object is extracted, the text label content and the geometric relationship between objects are analyzed to complete the extraction, and the extracted information is bound to the grouping result to form a structured object description, including: Get the boundary coordinates of the target object from the grouped object set, use spatial geometry calculation to determine the distance relationship between objects through coordinate difference, and obtain a distance association matrix; For the distance association matrix, analyze the text label content, judge the matching degree of the label and the object based on a preset tolerance threshold, and determine a matching label list; According to the matching label list, extract detailed information fields, and through a field fusion process, bind the fields to the grouping result to obtain a bound object dataset; Integrate and expand attribute labels based on the bound object dataset, and if the expanded attribute labels exceed a preset threshold, adjust the association weight based on the functional association between the labels and the target object to obtain the final structured object description.

[0008] Furthermore, the step of reading preset external configuration data and performing standardized replacement on the process description field of the structured object description, matching keywords to replace non-standard expressions with unified standard terms, and obtaining a standardized object description includes: Obtain a keyword list from preset external configuration data, and perform a matching operation on the process description field in the structured object description to obtain a set of matching keywords; The non-standard expressions are replaced according to the set of matching keywords, and the process description field is updated with unified standard terminology to obtain a standardized object description.

[0009] Furthermore, the step of calculating physical parameters based on the geometric information recorded in the standardized object description, completing the calculation based on a preset calculation formula and preset material property parameters, and integrating the calculation results with the standardized object description to form an integrated data set, including: The geometric dimension extraction results are obtained from the standardized object description. Material property parameters are queried based on the results. A preset calculation formula is used to perform parameter calculation operation by combining the material property parameters and the geometric dimension extraction results through product operation to obtain the physical parameter calculation results. Based on the physical parameter calculation results, match the material type matching information, update the standardized object description content through result data integration, and determine the calculation accuracy control index, wherein the calculation accuracy control index is the error rate between the physical parameter calculation results and the standard reference value; The calculation accuracy control index is obtained, and physical formulas are used in the relevant processing based on standardized information. If the index exceeds the preset threshold, the geometric feature recognition parameters are adjusted, and the parameter calculation operation is performed by re-querying the material property parameters to obtain the adjusted physical parameter calculation results. The calculated results of the adjusted physical parameters are integrated into a data set to form a structure, thus forming an integrated data set.

[0010] Furthermore, the step of outputting the integrated data set to an external storage file according to a preset structured format, adjusting the output style according to preset formatting rules, and simultaneously performing data verification, marking or correcting abnormal data to obtain the final output file includes: The structured format parameters are obtained from the integrated dataset. The parameter details are determined by querying a preset format library. The output data style is adjusted using preset formatting rules. Data verification is performed synchronously on the adjusted data and the operation of adjusting the output data style to obtain the verification result. If the verification result shows abnormal data, the abnormal data is marked, the data content is updated through correction operations, and the data is output to an external storage file to determine the final output file.

[0011] Furthermore, if the final output file still contains unprocessed abnormal data, the severity of the abnormality is determined according to a preset abnormality judgment threshold. The integrated data set is then updated by supplementing missing geometric relationship data and re-extracting complete information about the relevant objects, including: Unprocessed abnormal data is obtained from the final output file, and the severity of the abnormality is determined by a preset abnormality judgment threshold to obtain the severity result; If the severity result exceeds the preset anomaly judgment threshold, the missing geometric relationship data is supplemented by querying the original layer data of the design drawings or calling the architectural standard geometric relationship library to obtain the supplemented geometric relationship data. Based on the supplemented geometric relationship data, the complete information of the relevant objects is extracted again, and the extracted complete object information is determined. The integrated data set is updated using the extracted complete object information to obtain the updated data set.

[0012] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses a method for extracting production data from steel plate drawings. It addresses a combination of problems in traditional steel tower drawing processing, including inaccurate object identification, difficulties in grouping related data, large errors in physical parameter calculations due to non-standardized process descriptions, and untimely handling of abnormal data. These problems stem from the complexity of drawing data and reliance on manual intervention, leading to low efficiency and frequent errors. This invention traverses the steel tower drawing data, filtering geometric elements, text annotations, and symbol markers to form object subsets. It utilizes a collision detection algorithm based on boundary ranges and preset tolerances to achieve grouping related data, extracting detailed information and binding it to structured descriptions. Subsequently, it reads external configurations for standardized replacement using keyword matching, calculates physical parameters to integrate the data, and outputs the data in structured form. Simultaneously, it performs data verification and anomaly marking; if anomalies exceed a threshold, it supplements geometric relationships and re-extracts and updates the set, thereby solving the aforementioned problems. This significantly improves the automation level and accuracy of steel tower drawing processing, reduces human intervention, and achieves efficient and reliable engineering data output. Attached Figure Description

[0013] Fig. 1 This is a flowchart of a method for extracting production data from sheet metal graphics according to the present invention.

[0014] Fig. 2 This is a schematic diagram of a method for extracting production data from sheet metal graphics according to the present invention.

[0015] Fig. 3 This is another schematic diagram of a method for extracting production data from sheet metal graphics according to the present invention. Detailed Implementation

[0016] To further understand the content of this invention, a detailed description of the invention is provided in conjunction with the accompanying drawings and embodiments. The specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0017] like Figs. 1-3 This embodiment of a method for extracting production data from sheet metal graphics may specifically include: S101. Obtain the set of objects in the design drawings, including geometric elements, text annotations, and symbol markings. Perform preliminary screening by traversing the drawing data and comparing object attributes to obtain an extended set of objects.

[0018] By traversing the drawing data, an object set is obtained from the drawing data. The object set includes geometric elements, text annotations, and symbol markers, resulting in a preliminary object list. For this preliminary object list, object attributes are compared, and attribute matching is determined using a preset threshold. A filtered subset of objects is determined through geometric element identification and attribute matching verification. Based on the filtered subset of objects, element position coordinates and annotation content are parsed. A subset completeness index is obtained through text annotation extraction and annotation content parsing to determine subset completeness. If the subset completeness reaches a preset threshold, symbol marker classification and marker type differentiation are integrated. An expanded object set is obtained through symbol marker classification and set integration output.

[0019] For example, when processing tower steel plate design drawing data, the first step is to obtain a set of objects by traversing the drawing data. This traversal process is similar to a system scanning all elements in a CAD file, such as the outline segments of the main tower leg steel plates, the circular holes of the diagonal brace steel plates, the weld marking lines of the base connection plates, and other geometric elements, as well as textual annotations such as the steel plate thickness marking 14mm, the material marking Q355B, and the specification marking L600×300×16, and markings such as the tower bolt connection hole symbol Φ26 and the fillet weld marking △5, thus forming a preliminary object list. This list will list all detected objects. For example, a tower base steel plate plan view may include hundreds of line segments as steel plate outline geometric elements, dozens of text strings representing steel plate parameters, and various arrows or circles as connection or process symbols. In this way, it is ensured that all relevant components of the tower steel plates are comprehensively extracted from the original drawing data, forming a basic and operable list, laying the foundation for subsequent analysis.

[0020] In one possible implementation, the initial object list is compared against its attributes. This involves comparing the attributes of each object, such as whether the line segment length of the tower steel plate geometric elements conforms to the designed 600mm, whether the hole diameter is a standard Φ26mm bolt hole, and whether the weld marking color is red to indicate a stressed weld, against preset standard attributes. A preset threshold is used to determine the attribute matching degree; for example, if the length of a tower steel plate outline line segment differs from the standard design length by less than 5%, it is considered a match. Geometric element identification, such as using algorithms to detect whether line segments form closed rectangles to identify the tower base plate outline, is combined with attribute matching verification, such as checking whether the color is blue to indicate a non-stressed auxiliary line, to determine the filtered object subset. This subset excludes irrelevant or noisy objects, such as stray short line segments at the edge of the drawing or irrelevant annotation text, ultimately resulting in a refined set, such as retaining only the outline, bolt hole, and weld elements related to the tower's load-bearing steel plate.

[0021] It should be noted that, based on the filtered subset of objects, the element position coordinates and annotation content are further analyzed. This includes extracting the coordinates of each geometric element, such as the starting point (x=200, y=500) and ending point (x=800, y=500) coordinates of the outline segment of a steel plate for a tower brace, as well as analyzing adjacent text annotation content, such as "bolt hole spacing: 100mm". Text annotation extraction, for example, using optical character recognition technology to read strings such as "Q355B" and "16mm" from the drawings, and parsing the annotation content, such as analyzing whether "100mm" conforms to the standard bolt hole spacing of 100mm ± 0.5mm for towers, obtains a subset integrity index. This index may be a score, calculating whether all elements in the subset have matching annotations. If more than 90% of the tower steel plate elements have complete annotations, the subset integrity is considered high. This integrity assessment helps ensure data reliability and avoids errors in subsequent tower steel plate processing dimensions due to missing information.

[0022] In one possible implementation, if the subset completeness reaches a preset threshold, such as exceeding 85%, then the symbol classification and label type differentiation are integrated. For example, the symbols are classified into tower connection symbols and process symbols, grouped by type, and then aggregated and output to obtain an expanded object set. This expanded set combines the initial subset with the classified symbols to form a more comprehensive description of the tower steel plate model, such as integrating the outline of the tower leg base plate, bolt hole annotations, and fillet weld symbols, ultimately outputting basic data for CNC cutting and welding process planning of tower steel plates. This integration improves the usability of drawing data, realizing the transformation from two-dimensional drawings to intelligent production data.

[0023] S102. Filter specific objects in the extended object set according to preset conditions, identify geometric elements, text annotations and symbol marks that meet preset type requirements, and form a filtered object subset.

[0024] Specific objects are obtained from the extended object set, and their attributes are compared according to preset conditions to obtain a filtered object list. For the filtered object list, geometric elements are identified, and an attribute matching method is used to determine a subset of elements that meet preset type requirements. Based on the element subset, text annotations are extracted, and the degree of matching between the annotation content and the preset type requirements is judged to obtain annotation verification results. Based on the annotation verification results, symbol markers are integrated, marker types are identified, and a filtered object subset is formed.

[0025] Specifically, this involves retrieving specific objects from an extended set of objects. When processing tower steel plate design drawing data, the process begins by selecting objects related to specific tower components from the previously formed extended set (e.g., a set containing tower web steel plate outlines, bolt hole symbols, and thickness text annotations). For example, it focuses on elements related to the tower crossarm connection plate or tower leg base plate. Object attributes are compared using preset conditions, such as checking if the object's "component type" attribute is "crossarm connection plate" or if its "thickness attribute" is within the 5-20mm range (common thickness range for tower steel plates). This filters out objects that meet the criteria, creating a list of filtered objects. This list may include a dozen line segments representing the crossarm connection plate outline, and associated "Φ22 bolt hole" symbols representing connection holes, ensuring the list focuses on specific tower steel plate components rather than the entire drawing.

[0026] In one possible implementation, geometric elements are identified for this list of objects to be filtered. Specifically, an algorithm scans each object in the list, distinguishing geometric elements such as rectangular outline segments of steel tower plates, circular bolt holes, and straight weld seams. Then, attribute matching is used, for example, comparing whether the length of the line segments falls within the common length range of 300-800mm for crossarm connecting plates, and whether the bolt hole diameter falls within the standard bolt hole range of 18-26mm for steel towers, to determine whether it meets the preset type requirements of the crossarm connecting plate. In this way, a subset of elements can be obtained, such as retaining only those line segment combinations that form rectangular outlines and have standard bolt holes, excluding irrelevant curves or out-of-specification hole diameters, thereby refining the geometric parts related to the crossarm connecting plate structure.

[0027] It should be noted that the process of further extracting text annotations based on this subset of elements involves locating adjacent text objects. For example, strings such as "Connecting plate specifications: L800×300×12" and "Material: Q355B" are extracted from the vicinity of the crossarm connecting plate outline in the subset. Then, the matching degree between the annotation content and the preset type requirements is determined. For example, it is checked whether "Q355B" is a commonly used material for steel tower plates and whether "12mm" is within the preset thickness range. If the matching degree is higher than 80%, the annotation verification result is passed; otherwise, it is marked as needing review, thus obtaining a verification result list. This result can guide the subsequent confirmation of steel tower plate processing parameters.

[0028] In one possible implementation, the annotation verification results are used to integrate symbol markers and identify marker types. For example, the verified "Φ22" annotation is matched with a circular symbol to identify it as a "standard bolt hole for iron towers" type; the "△6" annotation is matched with a triangle symbol to identify it as a "fillet weld" type, ultimately forming a filtered subset of objects. This subset integrates the geometric contour of the crossarm connection plate, verified specification text, and categorized connection or process symbols, such as a complete crossarm connection plate object group, including a rectangular contour of L800×300×12, 6 Φ22 bolt holes, and 2 △6 fillet welds, for subsequent generation of CNC machining data for iron tower steel plates.

[0029] For example, in the actual processing of steel plate drawings for iron towers, this integration can be extended to other components, such as the screening of the base plate of the tower legs. Flange-related objects are obtained from the extended set, and attributes such as "square attribute" are compared to see if they are common squares of 400-800mm. A list is obtained, and then geometric elements such as square outlines and evenly distributed bolt holes are identified. The matching type requirements are used to form a subset, and annotations such as "base plate thickness: 40mm" and "number of bolt holes: 4" are extracted to determine the matching degree. After obtaining the results, symbols such as "T-joint weld mark" are identified as "base welding" type to form the final subset. This ensures that the whole process is logically coherent and builds reliable basic data for the production of iron tower steel plates step by step from the initial screening to the final integration.

[0030] S103. The filtered subset of objects is associated and grouped using a collision detection algorithm. Spatial cross detection is performed based on the object boundary range. The scattered objects are integrated into a logical association set by combining a preset tolerance parameter to obtain the grouped object set.

[0031] Boundary range data is obtained from the filtered subset of objects. Spatial cross-validation is used to determine overlapping areas between objects, and preliminary association pairs are determined based on preset tolerance parameters, resulting in an association pair list. Scattered objects are integrated into the association pair list, and a collision detection algorithm is executed to compare boundary ranges, obtaining logical association groups. Geometric features are extracted from these logical association groups, feature matching degrees are assessed, and group integrity is determined. Based on the group integrity, symbolic markers are fused to obtain the final grouped object set.

[0032] For example, when processing tower steel plate design drawing data, boundary range data is obtained from a filtered subset of objects, such as the outline segments related to the tower crossarm connecting plate, Φ22 bolt hole markings, and △6 weld annotations. This data typically includes the bounding box coordinates of each object; for example, the boundary of a bolt hole circle ranges from x=300 to x=320 and y=400 to y=420. Spatial cross-detection is used to determine overlapping areas between objects. This is a geometric algorithm that checks for intersections by comparing the coordinate ranges of the bounding boxes. For example, if the x-axis range of the crossarm connecting plate outline segment overlaps with the bolt hole marking and the y-axis also partially coincides, it is considered a potential overlap. Preliminary association pairs are determined by combining a preset tolerance parameter of 3-5 pixels to adapt to the tower steel plate production and processing error range, avoiding misjudgments caused by minor deviations. This results in a list of association pairs; for example, the outline segments of the crossarm connecting plate are paired with adjacent bolt holes, and the outline is paired with weld annotations to form more than a dozen association pairs, ensuring that these pairs reflect the actual connection relationships of the tower steel plate components.

[0033] In one possible implementation, scattered objects are integrated for this association list. For example, isolated weld seam annotation segments are merged with the outlines of nearby connecting plates. A collision detection algorithm is executed to compare the boundary range. This algorithm is similar to the graphic collision check in CNC cutting software. The degree of overlap is confirmed by calculating the intersection area of ​​bounding boxes. For example, if the intersection exceeds a preset 20% threshold, the association is confirmed. Finally, logical association groups are obtained, such as grouping the outlines of all crossarm connecting plates, 6 bolt holes, and 2 welds into one crossarm connecting plate unit.

[0034] For example, geometric features are extracted through these logical association groups. For instance, it is calculated whether the distribution spacing of bolt holes in the group is uniformly distributed at 100mm, and whether the length of the weld is consistent with the side length of the outline. The feature matching degree is judged, which involves comparing the similarity between the extracted features and the design template of the tower crossarm connection plate. For example, if the bolt hole spacing matching degree reaches 90%, it is considered a high match, and then the integrity of the group is determined, such as confirming whether the group contains the complete three elements of "outline + bolt hole + weld" rather than missing bolt holes.

[0035] In one possible implementation, the symbol markings are fused based on the grouping integrity. For example, the complete "crossarm connecting plate unit" is combined with the "hot-dip galvanizing mark Zn-85μm" symbol to obtain the final grouped object set. This set can be used for the subsequent generation of tower steel plate processing process cards, which can bring more accurate tower component identification results in business and improve the planning efficiency of CNC cutting and welding processes.

[0036] S104. Based on the grouped object set, extract detailed information of the target object, parse the text annotation content and the geometric relationship between the object to complete the extraction, and bind the extracted information with the grouping results to form a structured object description.

[0037] The boundary coordinates of the target objects are obtained from the grouped object set. Spatial geometric calculations are used to determine the distance relationship between objects through coordinate differences, resulting in a distance association matrix. The text annotation content of the distance association matrix is ​​parsed, and the matching degree between the annotations and objects is determined based on a preset tolerance threshold, thus establishing a matching annotation list. Detailed information fields are extracted from the matching annotation list, and the fields are bound to the grouping results through a field fusion process, resulting in a bound object dataset. Extended attribute labels are integrated into the bound object dataset. If the extended attribute labels exceed a preset threshold, the association weights are adjusted to obtain the final structured object description.

[0038] For example, when processing the production drawing data of steel plates for tower connecting plates, the boundary coordinates of the target objects are obtained from the grouped object set. For instance, the grouped set of a tower connecting plate unit includes a circular outline, 12 bolt holes, and T-shaped welds. The boundary coordinates of these objects may be defined as the minimum x value of the connecting plate outline being 100, the maximum x value being 900, the minimum y value being 100, and the maximum y value being 900; the boundary of a single bolt hole is x=200-220, y=200-220, etc. These coordinates are used to represent the spatial position of each object. Spatial geometric calculations are used to determine the distance relationship between objects by using coordinate differences. Here, spatial geometric calculations refer to using the Euclidean distance formula to calculate whether the distance from the center of the bolt hole to the center of the hanging hole is the design value of 400mm, and whether the distance from the weld to the bolt hole is the safety distance of 50mm. For example, the difference in x-coordinate and y-coordinate of the center points of two adjacent bolt holes is calculated, and then the square root of the sum of the squares is taken to construct a distance correlation matrix. The elements of the matrix represent the distance value between each pair of objects. For example, the value of 100 in row 1 and column 2 of the matrix means that the distance between the first bolt hole and the second bolt hole is 100mm. Finally, this distance correlation matrix is ​​obtained. It is a two-dimensional array and is used for subsequent verification of the connection accuracy of the tower hanging plate.

[0039] In one possible implementation, the text annotation content is parsed based on this distance correlation matrix. For example, if the matrix shows that the flange outline is close to the annotation text "Φ800", the text such as "flange diameter: Φ800mm" and "number of bolt holes: 12" will be extracted. Then, the matching degree is judged by combining the preset 10-pixel tolerance threshold. Specifically, if the distance is less than the threshold, the matching score is calculated. By comparing whether the position coordinates of the text and the boundary of the object overlap within the threshold, the matching annotation list is determined. For example, the annotations "Φ800" and "12 bolt holes" are matched with the flange outline to form a list containing 3 matching items.

[0040] For example, detailed information fields are extracted from this matching label list. These fields may include attributes such as flange diameter Φ800mm, thickness 20mm, bolt hole specification Φ24, and material Q355B. Through the field fusion process, these fields are bound to the corresponding grouping results. For example, the "Φ800mm" field is bound to the group "flange connection plate unit". This fusion process involves a mapping algorithm that maps the key-value pairs of fields to the identifiers of the groups one by one, and finally obtains the bound object dataset. This dataset is a structured table, where each row represents a steel tower plate grouping unit, and the columns include the bound field information.

[0041] In one possible implementation, extended attribute tags are integrated on top of this bound object dataset. For example, tags such as "Surface treatment: hot-dip galvanized Zn-85μm" and "Processing technology: CNC cutting + milling" are added. If the number of extended attribute tags exceeds a preset threshold of 3, the association weight is adjusted. Specifically, the influence of too many tags is weakened by reducing the weight coefficient, so as to avoid non-critical tags from interfering with the core processing parameters. This results in the final structured object description, which is a complete JSON format data containing all integrated attributes for further processing of architectural drawings.

[0042] S105. Read preset external configuration data, perform standardized replacement on the process description field of the structured object description, match keywords to replace non-standard expressions with unified standard terms, and obtain the standardized object description.

[0043] A keyword list is obtained from preset external configuration data. A matching operation is performed on the process description field in the structured object description to obtain a set of matching keywords. Non-standard expressions are replaced according to the set of matching keywords, and the process description field is updated with unified standard terminology to obtain a standardized object description.

[0044] It should be noted that this acquisition process involves reading external files or API interfaces to ensure that the list is updated in real time to adapt to changes in different building standards, such as extracting a keyword array from a JSON-formatted configuration file, where each element is a string.

[0045] In one possible implementation, a matching operation is performed on the process description field in a structured object description. This structured object description may be JSON data generated in previous steps, containing multiple fields. The process description field records text content such as "cutting before bending" or "fire-curling treatment." The matching operation is based on a string similarity algorithm, such as using Levenshtein distance to compare the field text with items in the keyword list, calculating the edit distance. If the distance is less than a preset threshold of 2, it is considered a match, thus obtaining a set of matching keywords. Specifically, this process iterates through each sentence in the process description field. For example, if the field content is "cutting before bending, then welding and ensuring quality secondary welds," it matches it with the keyword list to find keywords such as "cutting," "bending," and "welding," forming a set such as {"cutting," "bending," "welding"}. This set is used to identify non-standard expressions.

[0046] For example, after obtaining the set of matching keywords, the non-standard expressions are replaced based on this set. Here, non-standard expressions refer to inconsistent expressions in the field. For example, "rust removal" may be considered non-standard, while the industry standard term for steel tower plates is "shot blasting (Sa2.5 grade)"; "welding" is non-standard, and the standard term is "gas shielded welding (ER50-6 welding wire)". The replacement operation uses a mapping table to match keywords with standard terms. For example, a dictionary structure {"rust removal":"shot blasting (Sa2.5 grade)","galvanizing":"hot-dip galvanizing (Zn-85μm)","welding":"arc welding (ER50-6 welding wire)","cutting":"CNC plasma cutting (accuracy ±0.5mm)"} is used to update the process description field and obtain a standardized object description. For example, the original field "using gas shielded welding for T-joint welding of the base plate, followed by galvanizing" is updated to "using gas shielded welding (ER50-6 welding wire) for T-joint welding of the base plate, followed by hot-dip galvanizing (Zn-85μm) treatment".

[0047] In one possible implementation, this standardization process not only improves the consistency of steel plate production data for power transmission towers but also facilitates seamless integration with subsequent tower manufacturing processes. For example, in CNC welding equipment for power transmission towers, the standardized description of "gas shielded welding (ER50-6 welding wire)" can be directly imported into the equipment parameter library, automatically matching parameters such as welding current, welding voltage, and welding speed. This ensures consistent process descriptions across different drawings, thereby reducing welding quality issues caused by differences in description. In actual business operations, this replacement operation considers the context. For instance, if "bolt hole machining" appears in the matching keyword set, it is preferentially replaced with "CNC drilling" to match the requirements of the "Technical Conditions for Manufacturing Transmission Line Towers." Thus, the final object description becomes a standardized JSON structure containing more precise process fields, used for compiling steel plate processing process cards and establishing quality inspection standards for power transmission towers.

[0048] For example, extending to more complex scenarios, if the object description involves multiple process fields, such as "edge processing" and "hole machining" for the steel plate of the tower web, the matching operation will generate a set {"edge processing", "hole machining"}, which will then be replaced with the standard terms "milling (roughness Ra12.5)" and "CNC drilling (Φ20H12)". The updated description is "The steel plate of the tower web is formed by CNC plasma cutting (accuracy ±0.5mm), followed by milling (roughness Ra12.5) processing, and the bolt holes are machined by CNC drilling (Φ20H12)", which makes it easier for cross-departmental collaboration to achieve technical goals such as improving the efficiency of data exchange in the production of tower steel plates and avoiding processing delays caused by differences in description.

[0049] S106. Calculate physical parameters based on the geometric information recorded in the standardized object description, complete the calculation based on the preset calculation formula and preset material property parameters, and integrate the calculation results with the standardized object description to form an integrated data set.

[0050] Geometric dimension extraction results are obtained from the standardized object description. Material property parameters are queried based on these results. A preset calculation formula is used to perform parameter calculations by combining the material property parameters and the geometric dimension extraction results through multiplication, resulting in physical parameter calculations. Material type matching information is matched based on the physical parameter calculation results. The standardized object description content is updated through result data integration to determine calculation accuracy control indicators. These accuracy control indicators are obtained. Based on the standardized information, physical formulas are applied in related processing. If the indicator exceeds a preset threshold, the geometric feature recognition parameters are adjusted. The parameter calculations are then performed again by re-querying material property parameters to obtain adjusted physical parameter calculation results. The adjusted physical parameter calculation results are then integrated into a structured dataset, forming an integrated dataset.

[0051] For example, when processing tower steel plate design drawing data, the process of obtaining geometric dimension extraction results from the standardized object description typically involves parsing the geometric fields in the JSON structure. For instance, the standardized tower web steel plate process description in the object description includes specific values ​​for length, width, thickness, and bolt hole diameter, such as a length of 500mm, a width of 250mm, a thickness of 10mm, and a bolt hole diameter of 20mm. These values ​​are obtained through drawing parsing or image recognition in previous steps, such as CAD file coordinate extraction or scanned drawing dimension measurement, ensuring the accuracy of the dimensional data to support subsequent calculation of the physical parameters of the tower steel plate.

[0052] In one possible implementation, the material property parameters of the steel plate for the tower are queried based on these geometric dimension extraction results. This step connects to an external tower material standard database, such as querying the density parameters (7.85 g / cm³), elastic modulus (206 GPa), and tensile strength (≥470 MPa) of Q355B type tower-specific steel plate. These parameters are obtained in real time from a pre-set standard library of "Low Alloy High Strength Structural Steel". This query process matches the corresponding material properties based on keywords in the dimensional results, thereby forming a parameter set for calculation preparation.

[0053] Specifically, a preset calculation formula is used to perform parameter calculation operations by combining the material property parameters and geometric dimension extraction results through multiplication, resulting in physical parameter calculation results. For example, for the calculation of the unit weight of the steel plate of the iron tower web, the formula is: Unit weight = Length × Width × Thickness × Density × 1000, Substituting the value = 0.5m × 0.25m × 0.01m × 7.85g / cm × 1000 = 9.8125kg; Another example is the formula for calculating the machining allowance of bolt holes: Machining allowance = Design diameter + 2 × Drilling tolerance, Substituting the value = 20mm + 2 × 0.1mm = 20.2mm; This physical parameter calculation result represents the core production parameters of the steel plate of the iron tower web, helping to evaluate the overall weight of the iron tower, the amount of material purchased, and the adjustment of CNC machining tools.

[0054] For example, the material type matching information of the steel plate of the tower is matched according to the calculation results of the physical parameters. The standardized object description content is updated by integrating the result data to determine the calculation accuracy control index. For example, if the calculated single weight of the web is 9.8125kg, it is matched with the "Q355B steel plate single weight calculation standard", and "single weight: 9.8125kg" and "machining allowance: 20.2mm" are integrated into the "physical parameters" field of the object description. At the same time, the calculation accuracy index is set to ≤1%. This index is obtained by comparing the difference between the calculated value of 9.8125kg and the standard reference value to ensure that the description after data integration is more complete and facilitates the verification of the strength of the tower structure and the calculation of production costs.

[0055] In one possible implementation, the calculation accuracy control index is obtained, and physical formulas are applied in the relevant processing based on standardized information. If the index exceeds a preset 1% threshold, the geometric feature identification parameters of the steel plate of the tower are adjusted. The parameter calculation operation is performed by re-querying the material property parameters to obtain the adjusted physical parameter calculation results. For example, if the original length measurement deviation is 0.505m, it is adjusted to 0.5m, and the recalculated unit weight is 9.8125kg, reducing the error rate to 0.02%. This process improves the reliability of the calculation and reduces the risk of material waste in the tower production business.

[0056] Specifically, the data set calculated based on the adjusted physical parameters is integrated into a structure to form an integrated data set. For example, the adjusted "single weight 9.8125kg", "machining allowance 20.2mm" is integrated with "material Q355B" and "specification 500×250×10" into a nested JSON structure, which is convenient for importing into the iron tower production ERP system and CNC cutting equipment for simulation analysis, thereby achieving precise management of architectural drawings and improving the efficiency of construction planning.

[0057] S107. The integrated data set is output to an external storage file according to a preset structured format. The output style is adjusted by a preset formatting rule. Data verification is performed synchronously during output, and abnormal data is marked or corrected to obtain the final output file.

[0058] Structured format parameters are obtained from the integrated dataset. Parameter details are determined by querying a preset format library. The output data style is adjusted using preset formatting rules. A synchronous verification operation is performed on the adjusted data to obtain the verification result. If the verification result shows abnormal data, the abnormal data is marked, the data content is updated through a correction operation, and the data is output to an external storage file to determine the final output file.

[0059] For example, when processing steel plate design drawing data for iron towers, the process of obtaining structured format parameters from the integrated dataset typically involves parsing nested JSON structures. For instance, the dataset might contain the geometric dimensions of the iron tower web steel plate (500×250×10mm), material properties (Q355B), and physical parameters (unit weight 9.8125kg). These parameters are derived from previous calculations, ensuring the integrity of the format parameters to support subsequent production file output. In this approach, structured format parameters can be understood as pre-organized key-value pairs within the dataset, such as "steel plate specifications: 500×250×10mm" and "unit weight: 9.8125kg." The acquisition of these parameters is based on standardized fields in the dataset, helping the system identify the details required for output.

[0060] In one possible implementation, parameter details are determined by querying a pre-defined format library. This format library is a pre-defined database that stores commonly used document templates for tower production. For example, there are Excel templates for ERP systems that include fields such as "material ID, component name, specifications, material, unit weight, and processing technology," and XML templates for CNC equipment that include fields such as "coordinates, cutting path, and tolerance." For instance, for data on the steel plate of the tower web, querying the matching "CNC cutting document template" in the library will return specific requirements such as "unit: mm," "coordinate accuracy: ±0.01 mm," and "cutting speed: 800 mm / min." This query process relies on keywords in the parameters to ensure that the details are consistent with the standards of tower manufacturing equipment, thus providing a basis for adjusting the data style.

[0061] Specifically, preset formatting rules are used to adjust the output data style. For example, the rules include unit unification, field alignment, and process description standardization. For instance, the web plate length is changed from 500mm to the formatted "500.00mm", and the bolt hole machining allowance is changed from 20.2mm to "20.20mm". This adjustment is achieved by applying rule scripts, which involves checking data types and applying templates to ensure that the output style conforms to the tower production document specifications.

[0062] In one possible implementation, a synchronous verification operation is performed on the adjusted data to obtain the verification result. This synchronous verification can be understood as a mechanism for real-time comparison with the tower manufacturing standards. For example, it verifies whether the thickness of the web steel plate of 10mm is within the commonly used thickness range of 5-20mm for towers; whether the machining allowance of bolt holes of 20.2mm meets the requirement of drilling tolerance ≤20.3mm; and whether the material Q355B is the recommended material for towers. By comparing the thickness deviation of ±0.3mm and other thresholds, the result is obtained as "normal" or "abnormal". This operation ensures the consistency of the tower steel plate production data.

[0063] For example, if the verification result shows abnormal data, the abnormal data is marked. For instance, if the web plate thickness is 10.5mm, which exceeds the design by 10mm ± 0.5mm, it is marked as "Abnormal: Thickness exceeds the limit, 10.5mm > 10.3mm". This processing is achieved by marking the abnormal cells in red font and adding comments explaining the reasons in the Excel file, which facilitates subsequent correction.

[0064] In one possible implementation, the data content is updated through correction operations, such as correcting the marked "thickness 10.5mm" to "10.0mm". The update process involves manual adjustment after confirmation by technicians, or automatic correction by the system by calling the original data from the design drawings to ensure data accuracy.

[0065] Specifically, output to an external storage file and determine the final output file. For example, save the updated data as a CSV file, such as "Iron Tower Steel Plate Production Parameter Report.csv", which contains all the processed iron tower steel plate production parameters.

[0066] S108. If there is still unprocessed abnormal data in the final output file, the severity of the abnormality is determined according to the preset abnormality judgment threshold. The integrated data set is updated by supplementing the missing geometric relationship data and re-extracting the complete information of the relevant objects.

[0067] Unprocessed abnormal data is obtained from the final output file. The severity of the abnormality is determined by a preset abnormality judgment threshold, resulting in a severity result. If the severity result exceeds the preset abnormality judgment threshold, missing geometric relationship data is supplemented, resulting in supplemented geometric relationship data. Based on the supplemented geometric relationship data, complete information of relevant objects is re-extracted to determine the extracted complete object information. The integrated dataset is updated using the extracted complete object information, resulting in an updated dataset.

[0068] For example, when processing tower steel plate design drawing data, the process of obtaining unprocessed abnormal data from the final output file typically involves scanning specific fields in the CSV file, such as the spacing data of bolt holes in the tower bracing steel plates marked as "Abnormal: Exceeding Limits" in the file "Tower Steel Plate Production Parameter Report.csv". These data may be uncorrected items left over from previous verification. For example, if the center distance of bolt holes in a bracing steel plate is displayed as 101mm (design requirement 100mm±0.5mm), it will be marked as abnormal. This acquisition is done by reading file lines through a script and filtering entries where the "Abnormal Mark" column is not empty, ensuring that only tower steel plate data with abnormal labels are extracted, thereby providing basic data for subsequent judgment.

[0069] In one possible implementation, the severity of the anomaly is determined by a preset anomaly assessment threshold, and the severity result is obtained. Here, the anomaly assessment threshold is set based on the safety level of the tower structure. Severe anomalies: Critical parameters deviate from tolerance, potentially affecting structural safety; General anomalies: Non-critical parameters deviate from tolerance, not affecting core performance. For example, if the center distance of bolt holes in a diagonal brace is 101mm, exceeding the tolerance by 1.0mm (>0.5mm), it is classified as "severe"; if the diameter of an auxiliary hole is 20.35mm, exceeding the tolerance by 0.35mm (>0.3mm), it is classified as "minor". This process involves comparing data values ​​with thresholds. In principle, the thresholds are set based on the safety margins of the "Technical Conditions for Manufacturing Transmission Line Towers" to ensure that the severity reflects potential risks, thus obtaining results such as "severe" or "minor". This classification helps to prioritize the handling of high-risk anomalies.

[0070] Specifically, regarding the severity of the results, if they exceed the preset anomaly judgment threshold, the missing geometric relationship data of the tower steel plates is supplemented to obtain the supplemented geometric relationship data. For example, for severely abnormal bracing steel plates, if the data on the "distance between bolt holes and the edge of the steel plate" is missing, it is supplemented from the tower industry standard geometric relationship library. The standard for the edge distance of the bolt holes of the bracing steel plates is queried to obtain the requirement of "edge distance ≥ 25mm". Combined with the original layer data of the design drawings, the distance from the center of the bolt hole to the edge is calculated to be 28mm and added to the data. This supplementation is based on the geometric association principle of tower steel plates. The process includes querying the coordinates of the relevant objects, such as the steel plate edge x=100 and the bolt hole center x=128, and calculating the distance relationship to ensure that the supplemented data is consistent with the overall structure. The supplemented data obtained in this way can completely describe the rationality of the position of the bolt holes of the bracing steel plates, thereby avoiding processing deviations caused by incomplete data. In business, this can improve the accuracy of data processing and reduce subsequent assembly errors.

[0071] For example, based on the supplemented geometric relationship data, the complete information of the relevant objects is re-extracted to determine the extracted complete object information. This extraction process can be understood as pulling all related attributes from the updated data. For instance, for a diagonal bracing steel plate with "edge distance 28mm" added, its complete information includes: specifications: 800×300×8mm; bolt holes: Φ18, center distance 100mm (corrected), edge distance 28mm; material: Q355B; unit weight: 800×300×8×7.85×10 -6 ≈15.07kg; In principle, SQL query statements are used to traverse the steel plate database of iron towers, aggregate relevant fields, and ensure the integrity of information. This determination helps to form a reliable description of the production object of iron tower steel plates.

[0072] In one possible implementation, the integrated dataset is updated by extracting complete object information to obtain an updated dataset. For example, the "bolt hole center distance 105mm" of the diagonal brace steel plate is replaced with the corrected "100mm", the "edge distance 28mm" field is added, and the unit weight is updated to 33.91kg. This update ensures the consistency of the dataset and can bring more accurate drawing analysis results in the tower production business.

[0073] Specifically, the connection of the above process lies in the logical chain from acquisition to update of abnormal data, ensuring the continuity and reliability of tower steel plate data processing. For example, information is extracted after supplementing geometric relationships, which directly supports the optimization of the data set and avoids inconsistencies caused by isolated processing.

[0074] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from the spirit or essential characteristics of this application. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of this application is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within this application. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for extracting production data from sheet metal graphics, characterized in that, The method includes: Obtain the set of objects in the design drawings, including geometric elements, text annotations, and symbol markers. Perform preliminary filtering by traversing the drawing data and comparing object attributes to obtain an extended set of objects. Based on preset conditions, specific objects in the extended object set are filtered, and geometric elements, text annotations, and symbol markers that meet preset type requirements are identified to form a filtered subset of objects. The filtered subset of objects is associated and grouped using a collision detection algorithm. Spatial cross-detection is performed based on the object boundary range. The scattered objects are integrated into a logically associated set by combining a preset tolerance parameter, resulting in a grouped set of objects. Based on the grouped object set, extract detailed information of the target object, parse the text annotation content and the geometric relationship between the object to complete the extraction, and bind the extracted information with the grouping results to form a structured object description; Read preset external configuration data, perform standardized replacement on the process description field of the structured object description, match keywords to replace non-standard expressions with unified standard terms, and obtain the standardized object description; The physical parameters are calculated based on the geometric information recorded in the standardized object description. The calculation is completed based on the preset calculation formula and preset material property parameters. The calculation results are integrated with the standardized object description to form an integrated data set. The integrated data set is output to an external storage file according to a preset structured format. The output style is adjusted according to preset formatting rules. Data verification is performed synchronously during output, and abnormal data is marked or corrected to obtain the final output file. If there is still unprocessed abnormal data in the final output file, the severity of the abnormality is determined according to the preset abnormality judgment threshold. The integrated data set is then updated by supplementing the missing geometric relationship data and re-extracting the complete information of the relevant objects.

2. The method for extracting production data from sheet metal graphics according to claim 1, characterized in that, The process involves acquiring a set of objects from the design drawings, including geometric elements, text annotations, and symbols. Preliminary filtering is performed by traversing the drawing data and comparing object attributes to obtain an expanded set of objects, including: By traversing the drawing data, an object set is obtained from the drawing data. The object set includes geometric elements, text annotations, and symbol markers, resulting in a preliminary object list. For the initial object list, the object attributes are compared, and the attribute matching degree is judged using a preset threshold. The filtered object subset is determined through geometric element recognition and attribute matching verification. Based on the filtered subset of objects, the element position coordinates and annotation content are parsed, and the subset integrity index is obtained through text annotation extraction and annotation content parsing to determine the subset integrity. If the completeness of the subset reaches a preset threshold, the symbol tag classification and tag type distinction are integrated, and the extended object set is obtained by integrating the symbol tag classification and set integration output.

3. The method for extracting production data from sheet metal graphics according to claim 1, characterized in that, The step of filtering specific objects in the expanded object set according to preset conditions, identifying geometric elements, text annotations, and symbol markers that meet preset type requirements, and forming a filtered object subset includes: Specific objects are obtained from the extended object set, and object attributes are compared according to preset conditions to obtain a filtered object list; For the list of objects to be filtered, geometric elements are identified, and a subset of elements that meet the preset type requirements is determined by attribute matching. Based on the element subset, extract text annotations, determine the matching degree between the annotation content and the preset type requirements, and obtain the annotation verification result; Based on the annotation verification results, the symbol tags are integrated, the tag types are identified, and a filtered subset of objects is formed.

4. The method for extracting production data from sheet metal graphics according to claim 1, characterized in that, The process involves associating and grouping the filtered subset of objects using a collision detection algorithm, performing spatial cross-checking based on object boundary ranges, and integrating the scattered objects into a logically associated set using a preset tolerance parameter, resulting in a grouped object set, including: Boundary range data is obtained from the filtered subset of objects, spatial cross detection is used to determine the overlapping areas between objects, and preliminary association pairs are determined by combining preset tolerance parameters to obtain a list of association pairs; For the list of associated pairs, the scattered objects are integrated, and a collision detection algorithm is executed to compare the boundary ranges to obtain logically associated groups; Geometric features are extracted through the logical association grouping, the feature matching degree is judged, and the grouping integrity is determined; Based on the grouping integrity fusion symbol marker, obtain the final grouped object set.

5. The method for extracting production data from sheet metal graphics according to claim 1, characterized in that, The process of extracting detailed information about target objects based on the grouped object set, parsing the text annotation content and the geometric relationship between objects to complete the extraction, and binding the extracted information with the grouping results to form a structured object description includes: Obtain the boundary coordinates of the target object from the grouped object set, and use spatial geometric calculations to determine the distance relationship between objects through coordinate differences to obtain the distance correlation matrix; The text annotation content is parsed based on the distance correlation matrix, and the matching degree between the annotation and the object is determined by combining the preset tolerance threshold to determine the matching annotation list; Extract detailed information fields from the matching label list, and bind the fields to the grouping results through a field fusion process to obtain the bound object dataset. Extended attribute labels are integrated on the bound object dataset. If the number of extended attribute labels exceeds a preset threshold, the association weight is adjusted based on the functional relevance between the label and the target object to obtain the final structured object description.

6. The method for extracting production data from sheet metal graphics according to claim 1, characterized in that, The process of reading preset external configuration data and performing standardized replacement on the process description field of the structured object description, matching keywords to replace non-standard expressions with unified standard terms, yields a standardized object description, including: Obtain a keyword list from preset external configuration data, and perform a matching operation on the process description field in the structured object description to obtain a set of matching keywords; The non-standard expressions are replaced according to the set of matching keywords, and the process description field is updated with unified standard terminology to obtain a standardized object description.

7. The method for extracting production data from sheet metal graphics according to claim 1, characterized in that, The physical parameters are calculated based on the geometric information recorded in the standardized object description. The calculation is performed based on a preset calculation formula and preset material property parameters. The calculation results are then integrated with the standardized object description to form an integrated data set, including: The geometric dimension extraction results are obtained from the standardized object description. Material property parameters are queried based on the results. A preset calculation formula is used to perform parameter calculation operation by combining the material property parameters and the geometric dimension extraction results through product operation to obtain the physical parameter calculation results. Based on the physical parameter calculation results, match the material type matching information, update the standardized object description content through result data integration, and determine the calculation accuracy control index, wherein the calculation accuracy control index is the error rate between the physical parameter calculation results and the standard reference value; The calculation accuracy control index is obtained, and physical formulas are used in the relevant processing based on standardized information. If the index exceeds the preset threshold, the geometric feature recognition parameters are adjusted, and the parameter calculation operation is performed by re-querying the material property parameters to obtain the adjusted physical parameter calculation results. The calculated results of the adjusted physical parameters are integrated into a data set to form a structure, thus forming an integrated data set.

8. The method for extracting production data from sheet metal graphics according to claim 1, characterized in that, The process of outputting the integrated data set to an external storage file according to a preset structured format, adjusting the output style according to preset formatting rules, and simultaneously performing data verification, marking or correcting abnormal data to obtain the final output file includes: The structured format parameters are obtained from the integrated dataset. The parameter details are determined by querying a preset format library. The output data style is adjusted using preset formatting rules. Data verification is performed synchronously on the adjusted data and the operation of adjusting the output data style to obtain the verification result. If the verification result shows abnormal data, the abnormal data is marked, the data content is updated through correction operations, and the data is output to an external storage file to determine the final output file.

9. The method for extracting production data from sheet metal graphics according to claim 1, characterized in that, If the final output file still contains unprocessed abnormal data, the severity of the abnormality is determined according to a preset abnormality judgment threshold. The integrated data set is then updated by supplementing missing geometric relationship data and re-extracting complete information about the relevant objects, including: Unprocessed abnormal data is obtained from the final output file, and the severity of the abnormality is determined by a preset abnormality judgment threshold to obtain the severity result; If the severity result exceeds the preset anomaly judgment threshold, the missing geometric relationship data is supplemented by querying the original layer data of the design drawings or calling the architectural standard geometric relationship library to obtain the supplemented geometric relationship data. Based on the supplemented geometric relationship data, the complete information of the relevant objects is extracted again, and the extracted complete object information is determined. The integrated data set is updated using the extracted complete object information to obtain the updated data set.

Images