AI-based CAD pipe network data vector conversion method, medium and equipment

Abstract

The invention provides a CAD pipe network data vector conversion method based on AI, a medium and equipment, and relates to the technical field of CAD data processing. Converting the CAD data into vector data of original points, lines and planes, and performing data extraction on the vector data of the original points, lines and planes to obtain pipe network spatial data; identifying and extracting fields of the CAD data to obtain attribute information, hooking the attribute information with the pipe network spatial data, performing classified identification and extraction on various annotation information of the CAD data by utilizing a trained AI algorithm, and hooking the annotation information with the pipe network spatial data; and outputting the pipe network spatial data hooked with the attribute and annotation information to obtain vector conversion data. According to the method, the AI algorithm can be used for automatically matching and hooking the extracted attribute information, spatial information and annotations, so that quick and accurate vectorization processing of the CAD pipe network data is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of CAD data processing, and in particular to an AI-based CAD pipe network data vector conversion method, medium and equipment. Background Art

[0002] In pipeline engineering, CAD (computer-aided design) technology is used to draw and design pipelines, equipment and other infrastructure. When building a GIS (geographic information system) system, these CAD data usually need to be converted into vector data formats (such as Shapefile, GeoJSON) for further analysis and visualization in the GIS system. However, this conversion process usually requires manual operation, which is time-consuming and prone to errors; at the same time, pipeline data involves complex network structures and detailed attribute information, so there are high requirements for the accuracy of data conversion to ensure the accuracy and consistency of the data.

[0003] Existing data conversion methods mainly rely on manual operations, requiring technicians to extract and match spatial data, attribute data and annotation data one by one, which is cumbersome and inefficient. Manual operations can easily lead to data inconsistencies and errors, especially when processing large amounts of complex data, making it difficult to ensure data accuracy and consistency. Existing technologies lack automated and intelligent processing methods and cannot efficiently handle complex CAD data conversion needs. Summary of the invention

[0004] The purpose of the present invention is to provide a CAD pipe network data vector conversion method, medium and device based on AI (artificial intelligence), aiming to efficiently and accurately convert CAD pipe network data into vector format. The specific technical solution is as follows:

[0005] A CAD pipe network data vector conversion method based on AI, the method comprising the following steps:

[0006] S100, data preparation: obtaining CAD data of the pipe network;

[0007] S200, pipe network spatial data extraction: converting CAD data into vector data of original points, lines, and surfaces, and extracting data from the vector data of original points, lines, and surfaces to obtain pipe network spatial data;

[0008] S300, extracting and linking attribute and annotation information: identifying and extracting the fields of the CAD data to obtain attribute information and linking it with the pipe network space data, using the trained AI algorithm to classify, identify and extract various annotation information of the CAD data and link it with the pipe network space data;

[0009] S400, data output: outputting the pipe network spatial data linked with the attribute and annotation information to obtain vector conversion data.

[0010] Furthermore, in step S200, ARCGIS software is used to convert the CAD data into vector data of original points, lines and surfaces.

[0011] Further, in the step S200, data extraction of the vector data of the original points, lines, and surfaces to obtain the pipeline network spatial data specifically includes: layer judgment on the original point data, if it belongs to the pipeline point layer, coordinate matching is performed as pipeline point pre-data, and if they are consistent, pipeline point type attachment is performed; line type judgment is performed on the original line data, if the line type is a closed line, the symbol composed of the closed line is matched with the symbol library, if it is consistent, it is used as pipeline point pre-data for coordinate matching, and if it is consistent, pipeline point type attachment is performed; if the line type is a short broken line, it is identified as a note flag line, and if the line type is a normal line, layer judgment is performed, and if there is a corresponding value, pipeline type attachment is performed; layer judgment is performed on the original surface data, if it belongs to the pipeline point layer, spatial judgment is performed, if it is a small regular surface, the symbol composed of the small regular surface is matched with the symbol library, if it is consistent, it is used as pipeline point pre-data for coordinate matching, and if it is consistent, pipeline point type attachment is performed.

[0012] Furthermore, the point data is intelligently parsed, the block information in the original CAD data is extracted, a point symbol library is constructed, and it is compared with the national standard point symbols, and the point symbol library is modified and improved.

[0013] Furthermore, in the step S300, the fields of the CAD data are identified and extracted to obtain attribute information and linked with the pipe network space data, specifically including: having each field of the pipe network data specified in the built-in national standard for storing the extracted attribute information, after identifying and extracting the fields of the CAD data, matching them to the built-in national standard fields, and linking the corresponding attribute information with the pipe network space data.

[0014] Furthermore, in step S300, the types of annotation information include: ordinary text annotation, flow direction annotation, flag annotation, line annotation, and length annotation; using the trained AI algorithm to classify, identify, extract, and connect various annotation information of the CAD data with the pipeline network spatial data, specifically including: using the trained YOLOv5 target detection model to identify the flow direction arrows of the CAD data, using spatial proximity judgment according to the start and end coordinates of the arrow returned by the model, matching the pipelines within the range, and matching the pipelines parallel to the arrows according to the arrow angle, connecting the flow direction to the pipeline, and assigning a value of 0 or 1, 0 for downstream and 1 for upstream; using the trained EAST text detection model to identify the flow direction arrows of the CAD data, and matching the start and end coordinates of the arrows according to the start and end coordinates of the arrows ... matching the pipelines within the range, and matching the pipelines parallel to the arrows, connecting the flow direction to the pipeline, and assigning a value of 0 or 1, 0 for downstream and 1 for upstream; using the trained EAST text detection model to identify the flow direction arrows of the CAD data, and matching the start and end coordinates of the arrows according to the start and end coordinates of the arrows returned by the model, matching the pipelines within the range, matching the pipelines within the range, and matching the pipelines along the arrows, matching the flow direction to the pipeline, and connecting the flow direction to the pipeline, The three types of text annotations, namely, length annotations, ordinary text annotations, and notes along the line, are recognized. The spatial proximity algorithm is used to match the recognized text with the pipeline. During the matching process, the recognized text is split into different fields, and the split fields are matched to the built-in national standard fields, and the corresponding attribute information is attached to the pipeline network spatial data; the trained YOLOv5 model is used to recognize the flag-pulling line of the flag-pulling annotation of the CAD data, and the trained EAST model is used to recognize the flag-pulling text of the flag-pulling annotation of the CAD data, and the recognized flag-pulling line coordinates are spatially judged, and the coordinate position of the flag-pulling root node is matched with the pipe point data, and then the recognized flag-pulling text is attached to the pipe point.

[0015] Furthermore, CAD data is collected to organize the image data set, and the LabelImg tool is used to annotate the flow direction indication symbols, text annotations, flag lines, and flag texts in the image, and the images are saved in YOLO format. After configuring the training path and parameters, the YOLOv5 target detection model, EAST text detection model, YOLOv5 model, and EAST model are trained respectively to obtain the trained models.

[0016] Furthermore, the step S400 specifically includes: outputting the pipe network spatial data linked with the attributes and annotation information into three vector files of points, lines and surfaces, and generating corresponding illustration files according to the original CAD style and the extracted symbol library and line type information.

[0017] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the AI-based CAD pipe network data vector conversion method as described above.

[0018] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the above-mentioned AI-based CAD pipe network data vector conversion method are implemented.

[0019] The present invention provides an AI-based CAD pipe network data vector conversion method, medium and device, which have the following beneficial effects:

[0020] The present invention converts CAD data into vector data of original points, lines and surfaces, extracts data from the vector data of original points, lines and surfaces to obtain pipe network space data, identifies and extracts fields of CAD data to obtain attribute information and connects it with pipe network space data, uses a trained AI algorithm to classify, identify and extract various annotation information of CAD data and connects it with pipe network space data, outputs the pipe network space data connected with the attribute and annotation information to obtain vector conversion data, and can replace the traditional steps of manual identification and screening of CAD pipe network data with AI algorithm, automatically match and connect the extracted attribute information, spatial information and annotations, thereby realizing fast and accurate vectorization processing of CAD pipe network data. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a flow chart of a method for vector conversion of CAD pipe network data based on AI provided by the present invention;

[0022] Figure 2 is an overall flow chart of an embodiment of the present invention;

[0023] Figure 3 This is a schematic diagram of the recognition results of the original point data after conversion by ARCGIS software;

[0024] Figure 4 This is a general flow chart of extracting pipe network spatial data according to an embodiment of the present invention;

[0025] Figure 5 A schematic diagram of multi-field text splitting and attribute extraction according to an embodiment of the present invention;

[0026] Figure 6 4 is a structural block diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The following will be combined with the accompanying drawings provided by the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. According to the following description, the advantages and features of the present invention will be more clear. It should be noted that the accompanying drawings are all in a very simplified form and are not in precise proportions, which are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention.

[0028] Example 1

[0029] This embodiment provides a CAD pipe network data vector conversion method based on AI, see Figure 1 , 2 As shown, the method comprises the following steps:

[0030] S100, data preparation: obtaining CAD data of the pipe network.

[0031] S200, pipe network spatial data extraction: converting CAD data into vector data of original points, lines, and surfaces, and extracting data from the vector data of original points, lines, and surfaces to obtain pipe network spatial data.

[0032] In one embodiment, ARCGIS software is used to convert CAD data into vector data of original points, lines, and surfaces.

[0033] Traditional CAD conversion tools often lose CAD block information when performing vector conversion. For example, ARCGISDESTOP software will convert the original geometric blocks into closed lines when vectorizing CAD data. Figure 3 As shown in the figure, after conversion using ARCGIS, the original pipe point icon has no change in display effect, but the vector type has become a polyline. In addition, in the process of outputting CAD pipe network data, lines are used to construct point icons, which will often lead to meaningless closed lines being extracted from line data. In addition, pipe network data generally only needs to identify point and line data, but point data may also have CAD blocks mistakenly converted into surface data. Therefore, it is also necessary to identify and extract icons from surface data and compare them with the point symbol library.

[0034] In one embodiment, further referring to Figure 4 As shown, data extraction of the vector data of the original points, lines, and surfaces to obtain the spatial data of the pipeline network specifically includes: layer judgment on the original point data. If it belongs to the pipeline point layer, it is used as the pipeline point pre-data for coordinate matching. If they are consistent, the pipeline point type is attached; line type judgment is performed on the original line data. If the line type is a closed line, the symbol composed of the closed line is matched with the symbol library. If it is consistent, the pipeline point type is attached. If the line type is a short broken line, it is identified as an annotation flag line. If the line type is a normal line, a layer judgment is performed. If a corresponding value exists, the pipeline type is attached; layer judgment is performed on the original surface data. If it belongs to the pipeline point layer, a spatial judgment is performed. If it is a small regular surface, the symbol composed of the small regular surface is matched with the symbol library. If it is consistent, it is used as the pipeline point pre-data for coordinate matching. If it is consistent, the pipeline point type is attached.

[0035] In a preferred embodiment, the point data is intelligently parsed, the block information in the original CAD data is extracted, a point symbol library is constructed, and the point symbol library is compared with the national standard point symbols, and the point symbol library is modified and improved.

[0036] S300, attribute and annotation information extraction and linking: identify and extract the fields of CAD data to obtain attribute information and link it with the pipe network space data, use the trained AI algorithm to classify, identify and extract various annotation information of CAD data and link it with the pipe network space data.

[0037] In one embodiment, identifying and extracting the fields of CAD data to obtain attribute information and linking it with the pipe network space data specifically includes: having each field of the pipe network data specified in the built-in national standard for storing the extracted attribute information, after identifying and extracting the fields of CAD data, matching them to the built-in national standard fields, and linking the corresponding attribute information with the pipe network space data.

[0038] The keyword recognition method is used here. For example, if material information such as "concrete", "iron", and "PE" is recognized, it is attached to the material field; if a single or a pair of integer numbers is recognized, it is attached to the pipe diameter field. Some of the rules are as follows:

[0039] Keywords / Format Match Fields Integer Pipe diameter Integer x Integer Pipe diameter YS / WS / HS / PS (not case sensitive) Pipeline network type Rainwater / sewage / combined / drainage Pipeline network type Concrete / Iron / PE / HDPE / PVC Pipe network material With decimal numbers / With decimal numbers (smaller difference) Ground elevation / bottom elevation With decimal numbers / With decimal numbers (large difference) Buried Depth

[0040] Most CAD data have limited attribute data, and the information extracted from the attribute data is relatively small, so the main information still needs to be extracted through annotation extraction.

[0041] In one embodiment, the types of annotation information include: ordinary text annotations, flow direction annotations, flag annotations, line annotations, and length annotations; using a trained AI algorithm to classify, identify, extract, and link various types of annotation information of CAD data with the spatial data of the pipe network, specifically including: using a trained YOLOv5 target detection model to identify the flow direction arrows of the CAD data, using spatial proximity judgment to match the pipelines within the range according to the starting and ending coordinates of the arrows returned by the model, and matching the pipelines parallel to the arrows according to the arrow angles, linking the flow direction to the pipelines, and assigning a value of 0 or 1, 0 for downstream and 1 for upstream; using a trained EAST text detection model to identify three types of text annotations, including length annotations, ordinary text annotations, and line annotations of CAD data, using a spatial proximity algorithm to match the identified text with the pipelines, and during the matching process, splitting the identified text into different fields, such as Figure 5 As shown, the split fields are matched to the built-in national standard fields, and the corresponding attribute information is attached to the pipe network spatial data; the trained YOLOv5 model is used to identify the flag lines of the flag annotations in the CAD data, and the trained EAST model is used to identify the flag texts of the flag annotations in the CAD data, and the identified flag line coordinates are spatially judged, and the coordinate position of the flag root node is matched with the pipe point data, and then the identified flag text is attached to the pipe point.

[0042] In a preferred embodiment, CAD data is collected to organize the image data set, and the LabelImg tool is used to respectively annotate the flow direction indicator symbols, text annotations, flag lines, and flag texts in the image, and the images are saved in YOLO format. After the training path and parameters are configured, the YOLOv5 target detection model, EAST text detection model, YOLOv5 model, and EAST model are trained respectively to obtain trained models.

[0043] Since the patterns of flow direction and flag-pull lines are relatively simple, after 20 epochs of training, the model can accurately identify the flow direction arrows and flag-pull lines in the image. Text detection is more complicated than arrow and flag-pull line recognition, and it is necessary to set the training for 100 epochs.

[0044] S400, data output: outputting the pipe network spatial data linked with the attribute and annotation information to obtain vector conversion data.

[0045] In one embodiment, it specifically includes: outputting the pipe network spatial data after connecting with the attributes and annotation information into three vector files of points, lines and surfaces, and generating corresponding illustration files (SLD format) according to the original CAD style and the extracted symbol library and line type information, so as to support the style that can be directly applied when publishing the vector data as a map service.

[0046] The AI-based CAD pipe network data vector conversion method provided by the present invention converts CAD data into vector data of original points, lines and surfaces, extracts data from the vector data of original points, lines and surfaces to obtain pipe network space data, identifies and extracts fields of CAD data to obtain attribute information and connects it with the pipe network space data, uses a trained AI algorithm to classify, identify and extract various annotation information of CAD data and connects it with the pipe network space data, outputs the pipe network space data connected with the attribute and annotation information to obtain vector conversion data, and can replace the traditional steps of manual identification and screening of CAD pipe network data with the AI ​​algorithm, automatically matches and connects the extracted attribute information, spatial information and annotations, thereby realizing fast and accurate vectorization processing of CAD pipe network data.

[0047] Example 2

[0048] This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the AI-based CAD pipe network data vector conversion method described above are implemented.

[0049] The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk (HDD) or a solid-state drive (SSD), etc.; the storage medium may also include a combination of the above-mentioned types of memory.

[0050] Example 3

[0051] This embodiment provides a computer device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the AI-based CAD pipe network data vector conversion method described above when executing the program.

[0052] like Figure 6 As shown, the computer device may include: at least one processor 71, such as a CPU (Central Processing Unit), at least one communication interface 73, a memory 74, and at least one communication bus 72. The communication bus 72 is used to realize the connection and communication between these components. The communication interface 73 may include a display screen (Display) and a keyboard (Keyboard), and the optional communication interface 73 may also include a standard wired interface and a wireless interface. The memory 74 may be a high-speed RAM memory (Random Access Memory, volatile random access memory) or a non-volatile memory (non-volatile memory), such as at least one disk storage. The memory 74 may optionally be at least one storage device located away from the aforementioned processor 71. The memory 74 stores application programs, and the processor 71 calls the program code stored in the memory 74 to execute any of the above method steps.

[0053] The communication bus 72 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The communication bus 72 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0054] Among them, the memory 74 may include a volatile memory (English: volatile memory), such as a random access memory (English: random-access memory, abbreviated: RAM); the memory may also include a non-volatile memory (English: non-volatile memory), such as a flash memory (English: flash memory), a hard disk (English: hard disk drive, abbreviated: HDD) or a solid-state drive (English: solid-state drive, abbreviated: SSD); the memory 74 may also include a combination of the above types of memory.

[0055] The processor 71 may be a central processing unit (CPU), a network processor (NP), or a combination of a CPU and a NP.

[0056] The processor 71 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0057] Optionally, the memory 74 is also used to store program instructions. The processor 71 can call the program instructions to implement the AI-based CAD pipe network data vector conversion method of the present invention.

[0058] Those skilled in the art should understand that the present invention can be implemented in many other specific forms without departing from the spirit and scope of the present invention. Based on the embodiments of the present invention, any changes and modifications made by ordinary technicians in the field of the present invention according to the above disclosure are within the scope of protection of the claims.

Claims

1. A CAD pipe network data vector conversion method based on AI, characterized in that: The method comprises the following steps: S100, data preparation: obtaining CAD data of the pipe network; S200, pipe network spatial data extraction: converting CAD data into vector data of original points, lines, and surfaces, and extracting data from the vector data of original points, lines, and surfaces to obtain pipe network spatial data; S300, extracting and linking attribute and annotation information: identifying and extracting the fields of the CAD data to obtain attribute information and linking it with the pipe network space data, using the trained AI algorithm to classify, identify and extract various annotation information of the CAD data and link it with the pipe network space data; S400, data output: outputting the pipe network spatial data linked with the attribute and annotation information to obtain vector conversion data.

2. The AI-based CAD pipe network data vector conversion method according to claim 1 is characterized in that: In step S200, ARCGIS software is used to convert the CAD data into vector data of original points, lines and surfaces.

3. The AI-based CAD pipe network data vector conversion method according to claim 1 is characterized in that: In the step S200, data extraction of the vector data of the original points, lines and surfaces to obtain the spatial data of the pipe network specifically includes: layer judgment on the original point data, if it belongs to the pipe point layer, coordinate matching is performed as pipe point pre-data, and if they are consistent, pipe point type attachment is performed; line type judgment is performed on the original line data, if the line type is a closed line, the symbol composed of the closed line is matched with the symbol library, if it is consistent, it is used as pipe point pre-data for coordinate matching, and if it is consistent, pipe point type attachment is performed; if the line type is a short broken line, it is identified as an annotation flag line, and if the line type is a normal line, layer judgment is performed, and if there is a corresponding value, pipeline type attachment is performed; layer judgment is performed on the original surface data, if it belongs to the pipe point layer, space judgment is performed, if it is a small regular surface, the symbol composed of the small regular surface is matched with the symbol library, if it is consistent, it is used as pipe point pre-data for coordinate matching, and if it is consistent, pipe point type attachment is performed.

4. The AI-based CAD pipe network data vector conversion method according to claim 3 is characterized in that: Intelligently analyze point data, extract block information from original CAD data, build a point symbol library, compare it with national standard point symbols, and modify and improve the point symbol library.

5. The AI-based CAD pipe network data vector conversion method according to claim 1 is characterized in that: In the step S300, the fields of the CAD data are identified and extracted to obtain attribute information and linked with the pipe network space data, specifically including: each field of the pipe network data specified in the built-in national standard is used to store the extracted attribute information, after the fields of the CAD data are identified and extracted, they are matched to the built-in national standard fields, and the corresponding attribute information is linked with the pipe network space data.

6. The AI-based CAD pipe network data vector conversion method according to claim 1 is characterized in that: In the step S300, the types of annotation information include: ordinary text annotation, flow direction annotation, flag annotation, line annotation, and length annotation; using the trained AI algorithm to classify, identify, extract, and link the various annotation information of the CAD data with the pipeline network spatial data, specifically including: using the trained YOLOv5 target detection model to identify the flow direction arrows of the CAD data, using spatial proximity judgment according to the start and end coordinates of the arrow returned by the model, matching the pipelines within the range, and matching the pipelines parallel to the arrows according to the arrow angle, linking the flow direction to the pipeline, and assigning a value of 0 or 1, 0 for downstream and 1 for upstream; using the trained EAST text detection model to identify the length of the CAD data Three types of text annotations, including degree annotations, ordinary text annotations, and line annotations, are recognized. The spatial proximity algorithm is used to match the recognized text with the pipeline. During the matching process, the recognized text is split into different fields, and the split fields are matched to the built-in national standard fields, and the corresponding attribute information is attached to the pipeline network spatial data; the trained YOLOv5 model is used to recognize the flag lines of the flag annotations of the CAD data, and the trained EAST model is used to recognize the flag texts of the flag annotations of the CAD data, and the recognized flag line coordinates are spatially judged, and the coordinate position of the flag root node is matched with the pipe point data, and then the recognized flag text is attached to the pipe point.

7. The AI-based CAD pipe network data vector conversion method according to claim 6 is characterized in that: Collect CAD data and organize image data sets. Use the LabelImg tool to annotate the flow direction indication symbols, text annotations, flag lines, and flag texts in the images and save them in YOLO format. After configuring the training path and parameters, train the YOLOv5 target detection model, EAST text detection model, YOLOv5 model, and EAST model to obtain the trained models.

8. The AI-based CAD pipe network data vector conversion method according to claim 1 is characterized in that: The step S400 specifically includes: outputting the pipe network spatial data linked with the attributes and annotation information into three vector files of points, lines and surfaces, and generating corresponding illustration files according to the original CAD style and the extracted symbol library and line type information.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the AI-based CAD pipe network data vector conversion method as described in any one of claims 1 to 8 are implemented.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the AI-based CAD pipe network data vector conversion method as described in any one of claims 1-8 are implemented.

Images