Underground pipe network construction drawing generation and verification method and system based on image recognition

Abstract

The invention discloses an underground pipe network construction drawing generation and verification method and system based on image recognition, relates to the technical field of construction drawing verification analysis, and solves the technical problem that in the verification process, detection of pipe network element integrity and space conflicts is not comprehensive and efficient enough. According to the method, extracted pipe network element data is compared with original design data element by element, verification is carried out from element types, number, attributes and other dimensions, whether a construction drawing is complete or not can be rapidly and accurately judged, missing or wrong elements can be found in time, and a minimum rectangular bounding box and a three-dimensional projection overlapping judgment algorithm are utilized to judge whether the construction drawing is complete or not. According to the method, the space conflicts among the pipe network elements can be efficiently detected, the conflict detection efficiency and accuracy are improved, the overlapped elements are classified according to the elastic modulus of the element material, the hard and non-hard overlapped elements are subjected to targeted analysis, whether design conflicts exist or not can be judged, and a more scientific basis is provided for design optimization.

Description

Technical Field

[0001] The present invention relates to the technical field of construction drawing verification and analysis, and in particular to an underground pipe network construction drawing generation and verification method and system based on image recognition. Background Art

[0002] In urban infrastructure construction, the underground pipeline system is complex and huge, covering various types of pipelines such as water supply and drainage, gas, electricity, and communications.

[0003] Traditional underground pipeline network construction drawing design and verification rely primarily on manual drawing and review, which is inefficient and prone to errors. With the advancement of computer technology, some designs use CAD software for two-dimensional drawing. While this improves drawing efficiency, manual verification of drawings still requires checking each drawing against design specifications and original data, making it difficult to detect complex spatial conflicts.

[0004] Although image recognition and 3D modeling technologies have begun to be applied in engineering fields in recent years, existing technologies still have limitations in generating and verifying underground pipeline network construction drawings. For example, image recognition accuracy is insufficient, making it difficult to accurately extract pipeline network element features; the 3D modeling process is complex, and there is a lack of effective means to automatically generate construction drawings; and during the verification process, the detection of pipeline network element integrity and spatial conflicts is not comprehensive and efficient enough, failing to meet the high-precision and high-efficiency requirements of modern urban underground pipeline network construction. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides a method and system for generating and verifying underground pipeline network construction drawings based on image recognition, which solves the problem of insufficient comprehensiveness and efficiency in detecting the integrity of pipeline network elements and spatial conflicts during the verification process.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a method for generating and verifying underground pipe network construction drawings based on image recognition, the method specifically comprising the following steps: Collect image data of the underground pipe network area and preprocess it to obtain preprocessed images. At the same time, extract image features and build a three-dimensional model to generate construction drawings; Verify and analyze the construction drawings, compare the pipe network elements in the construction drawings with the original data, and generate conflict verification analysis signals if the two are the same. If the two are different, obtain the different elements to generate missing information; Process the conflict check analysis signal, label each pipe network element and determine the corresponding adjacent elements in order, then use the minimum bounding box algorithm to determine the overlap between the pipe network element and its adjacent elements, and generate normal design information or overlap analysis signal; Analyze the acquired overlap analysis signal, obtain overlapping adjacent elements and classify them into hard overlapping elements and non-hard overlapping elements according to their design properties, analyze the hard overlapping elements, calculate their corresponding overlap distances and compare them with preset distance values, generate conflict design information or impact analysis signals, and generate overlap level information based on the overlap distances; Analyze non-rigid overlapping elements, determine their impact on pipe network elements, and generate normal design information or conflicting design information.

[0007] As a further solution of the present invention, the specific method of generating the construction drawings is: A high-definition camera equipped with a drone is used to collect image data of the underground pipeline network area, and noise reduction, grayscale, and binarization preprocessing operations are performed to obtain preprocessed images. Image recognition algorithms are used to extract image features. Combined with geographic information system technology, a three-dimensional model of the underground pipeline network is constructed, and construction drawings are automatically generated in accordance with the drawing specifications and requirements of construction drawings.

[0008] As a further solution of the present invention, the specific method of verifying and analyzing the construction drawings is: Obtain all pipe network elements in the construction drawing and compare them with the original data. If the pipe network elements are the same as the original data, the construction drawing is complete and a conflict detection analysis signal is generated. Conversely, if the pipe network elements are different from the original data, the construction drawing is incomplete and an incomplete analysis signal is generated. For incomplete analysis signal analysis, the corresponding incomplete elements are obtained based on the original data, and the missing information is generated based on it.

[0009] As a further solution of the present invention, the specific method of processing the conflict check analysis signal is: Get all the pipe network elements and label them as i, and i=1, 2, ..., j, where j represents the type of pipe network element. Then, according to the order of pipe network element i, get the adjacent elements whose distance is less than the threshold and label them as i. a , and a=1, 2, …, b, where b represents the type of adjacent elements; Then, the pipe network element i is taken as the target element, and its corresponding minimum rectangular bounding box is established as the target bounding box. The three-dimensional coordinate system is established with the center point of the bounding box as the origin, and the maximum vertex coordinate max (x i ,y i , z i ) and the minimum vertex coordinate min(x i ,y i , z i), similarly, establish the minimum rectangular bounding box corresponding to the adjacent elements and record it as the bounding box to be analyzed, and at the same time obtain the maximum fixed point coordinates and minimum vertex coordinates min (x ia ,y ia , z ia ), and analyze based on the established minimum rectangular bounding box.

[0010] As a further solution of the present invention, the specific method of performing analysis based on the established minimum rectangular bounding box is: According to the adjacent element number i a Analyze the size relationship between its projection on the x, y and z axes and the projection of the target bounding box in turn. The specific judgment relationship is as follows: x-axis projection overlap: min i .x≤max ia .x and max i .x≥min ia .x; y-axis projection overlap: min i .y≤ma ia .y and max i .y≥min ia .y; z-axis projection overlap: min i .z≤ma ia .z and max i .z≥min ia .z; If any set of projections on the x, y, and z axes do not overlap, it means that there is no overlap between the two, and normal design information is generated. Conversely, if the projections on the x, y, and z axes all overlap, it means that there is overlap between the two, and an overlap analysis signal is generated.

[0011] As a further solution of the present invention, the specific method of analyzing the acquired overlapping analysis signals is: Adjacent elements that overlap with the target element are obtained and recorded as elements to be analyzed. At the same time, the design properties of the elements to be analyzed are analyzed, and they are classified according to the design properties to obtain hard overlapping elements and non-hard overlapping elements. Specifically, elements with an elastic modulus E greater than 5 GPa are classified as hard elements, and elements with an elastic modulus E less than 5 GPa are classified as non-hard elements. The classified hard overlapping elements and non-hard overlapping elements are analyzed separately.

[0012] As a further solution of the present invention, the specific method of separately analyzing the hard overlapping elements and non-hard overlapping elements obtained by the classification is: Analyze the obtained hard overlapping elements, take the maximum straight-line distance between the overlapping parts of the two elements and compare it with the preset distance value. If the overlapping distance is greater than the preset distance value, generate conflict design information; if the overlapping distance is less than the preset distance, generate an impact analysis signal; Then, the generated impact analysis signal is analyzed, and the overlap level is determined according to the obtained overlap distance, and overlap level information is generated at the same time.

[0013] The obtained non-rigid overlapping elements are analyzed to determine whether their design affects the target elements. If not, normal design information is generated. Otherwise, conflicting design information is generated.

[0014] The underground pipe network construction drawing generation and verification system based on image recognition includes: image data acquisition module, data verification and analysis module, pipe network construction conflict analysis module, adaptive processing module and verification information output module; Image data acquisition module, which is used to collect image data of the underground pipe network area, perform preprocessing to obtain preprocessed images, extract image features and build a three-dimensional model, generate construction drawings, and transmit them to the data verification and analysis module; The data verification and analysis module is used to verify and analyze the obtained construction drawings, compare the pipe network elements in the construction drawings with the original data, and generate a conflict verification analysis signal if the two are the same. The signal is transmitted to the pipe network construction conflict analysis module. If the two are different, the different elements are obtained to generate missing information and transmit it to the verification information output module. The pipeline network construction conflict analysis module is used to process the obtained conflict verification analysis signal, label each pipeline network element and determine the corresponding adjacent elements in order, then use the minimum bounding box algorithm to determine the overlap between the pipeline network element and its adjacent elements, generate normal design information or overlap analysis signal, and transmit the overlap analysis signal to the adaptive processing module; An adaptive processing module is used to analyze the acquired overlap analysis signal, obtain overlapping adjacent elements, and classify them into hard overlapping elements and non-hard overlapping elements according to their design properties; analyze the hard overlapping elements, calculate their corresponding overlap distances and compare them with preset distance values, generate conflict design information or impact analysis signals, and generate overlap level information based on the overlap distances, and transmit it to the verification information output module; Analyze non-rigid overlapping elements, determine their impact on pipe network elements, generate normal design information or conflicting design information, and transmit it to the verification information output module; The verification information output module is used to display the obtained omission information, overlapping level information, normal design information and conflicting design information to the corresponding management personnel.

[0015] The present invention provides a method and system for generating and verifying underground pipe network construction drawings based on image recognition. Compared with the existing technology, it has the following advantages: The present invention collects image data through drones and mobile devices, and applies noise reduction, grayscale, and binarization preprocessing technologies, combined with the Canny algorithm and OCR technology, to more accurately extract pipeline network image features and text annotation information, providing a reliable data basis for subsequent modeling and design. Based on the extracted image feature information, the present invention combines GIS technology to construct a three-dimensional model, and automatically generates plan views, elevations, sections, etc. according to the construction drawing drawing specifications, thereby improving the efficiency of construction drawing generation and reducing manual drawing errors.

[0016] The present invention compares the extracted pipe network element data with the original design data element by element, and verifies from the dimensions of element type, quantity, attribute, etc., so as to quickly and accurately determine whether the construction drawings are complete and timely discover missing or erroneous elements. By using the minimum rectangular bounding box and three-dimensional projection overlap judgment algorithm, the present invention can efficiently detect spatial conflicts between pipe network elements, thereby improving the efficiency and accuracy of conflict detection. The overlapping elements are classified according to the elastic modulus of the element material, and the rigid and non-rigid overlapping elements are analyzed respectively. It can not only determine whether there is a design conflict, but also determine the conflict level and impact, providing a more scientific basis for design optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a diagram of the steps and methods of the present invention; Figure 2 This is a system block diagram of the present invention. DETAILED DESCRIPTION

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0019] Example 1: Please refer to Figure 1 This application provides a method for generating and verifying underground pipe network construction drawings based on image recognition, which specifically includes the following steps: Step 1: Use drones equipped with high-definition cameras or handheld mobile devices to collect image data of the underground pipeline network area. Pipeline network image information can also be obtained from existing electronic documents such as CAD drawings. The obtained image data is then preprocessed to produce a preprocessed image. The preprocessing process involves performing noise reduction, grayscale conversion, and binarization on the collected image to improve image quality and facilitate subsequent feature extraction. For example, noise reduction removes random noise from the image, grayscale conversion converts color images to grayscale images, and binarization further converts grayscale images to black and white, highlighting key information such as pipeline network lines. Image recognition algorithms are then used to extract image features from the preprocessed image. For example, edge detection algorithms (such as the Canny algorithm) extract geometric features such as pipeline lines and nodes, and identify information such as pipeline direction, diameter, and connection method. At the same time, optical character recognition (OCR) technology is used to extract text annotation information in the image, such as pipe diameter, material, elevation, slope, etc. Based on the extracted image feature information and combined with geographic information system (GIS) technology, a three-dimensional model of the underground pipeline network is constructed. Professional three-dimensional modeling software or specific three-dimensional modeling algorithms can be used to convert two-dimensional pipeline network information into a model in three-dimensional space. Based on the three-dimensional model, construction drawings are automatically generated in accordance with the drawing specifications and requirements of construction drawings, including plan views, elevations, sections, etc.

[0020] Step 2: Verify and analyze the obtained construction drawings. Use computer vision algorithms (such as YOLO and Faster R-CNN) to extract features from the construction drawings, identify the location, shape, and annotation information of pipeline network elements such as pipelines, valves, and inspection wells, and use optical character recognition technology to extract text annotations from the drawings, including parameter information such as pipe diameter, material, and coordinates. Convert the identified pipeline network elements into structured data, such as storing each element's type, location, and attributes in JSON format. The extracted pipe network element data is compared element by element with the original design data (such as the design database and survey report), and verified according to the following verification rules: Element type verification: Check whether the drawing contains all the pipe network element types required by the design.

[0021] Element quantity verification: Compare the number of similar elements in the drawing and the original data to see if they are consistent.

[0022] Element attribute verification: Check whether the attribute parameters of each element (such as pipe diameter, material) match the original data; If all network elements are identical to the original data, the construction drawing is considered complete, a conflict check analysis signal is generated, and the next conflict detection process is entered. If there are any inconsistencies (missing elements, attribute errors, etc.), the construction drawing is considered incomplete and an incomplete analysis signal is generated. Through comparative analysis, missing or incorrect pipe network elements in the construction drawings are identified. Based on the original data, a detailed missing information report is generated, including the type, location, and attribute requirements of the missing elements. The missing or incorrect locations are marked on the drawings and highlighted in red to facilitate designers to quickly locate and modify them.

[0023] Step 3: Get the generated conflict check analysis signal and analyze it. Get all the pipe network elements and label them as i, and i=1, 2, ..., j, where j represents the type of pipe network element. Then get the adjacent elements corresponding to the pipe network element i in the order of the labels. The adjacent elements here are represented as the straight-line distance between the pipe network elements is less than the preset value. At the same time, the obtained adjacent elements are labeled as i. a , and a=1, 2, …, b, where b represents the type of adjacent elements; Then, the pipe network element i is taken as the target element, and its corresponding minimum rectangular bounding box is established as the target bounding box. The rectangular bounding box here is represented as enclosing the target element as a whole. At the same time, a three-dimensional coordinate axis is established with the center point of the target bounding box as the origin. The maximum vertex and minimum vertex corresponding to the target bounding box and the corresponding three-dimensional coordinates are obtained as max(x i ,y i , z i ) and min(x i ,y i , z i ), similarly, establish the minimum rectangular bounding box corresponding to the adjacent elements and record it as the bounding box to be analyzed, and at the same time obtain the maximum and minimum vertex coordinates max (x ia ,y ia , z ia ) and min(x ia ,y ia , z ia ); According to the adjacent element number i a Analyze the size relationship between its projection on the x, y, and z axes and the projection of the target bounding box in turn, and determine whether there is overlap between the two based on the projection size relationship. The specific judgment method is as follows: x-axis projection overlap: min i .x≤max ia .x and max i .x≥min ia .x; y-axis projection overlap: mini .y≤ma ia .y and max i .y≥min ia .y; z-axis projection overlap: min i .z≤ma ia .z and max i .z≥min ia .z; If any set of projections on the x, y, and z axes do not overlap, it means that there is no overlap between the two, and normal design information is generated. Conversely, if the projections on the x, y, and z axes all overlap, it means that there is overlap between the two, and an overlap analysis signal is generated.

[0024] After receiving the conflict verification analysis signal, the system uses image recognition or structured data parsing technology to extract all pipeline network elements (pipelines, valves, manholes, etc.) from the construction drawings. These elements are divided into j categories by type, labeled i=1, 2, ..., j (for example, i=1 represents category 1 pipelines, i=2 represents valves). A distance threshold D is set (for example, the default D=2m for municipal pipelines). The straight-line distance between elements is calculated using the Euclidean distance formula. Adjacent elements with a distance to element i less than D are selected and labeled ia (a=1, 2, ..., b, where b is the number of adjacent elements). Generate a minimum axial bounding box (AABB) for element i, recorded as the target bounding box, whose minimum vertex coordinate is min i (x i ,y i , z i ), the maximum vertex coordinate is max i (x i ,y i , z i ), and the center point of the bounding box is O i Establish a local 3D coordinate system for the origin.

[0025] For each adjacent element i a Repeat the above steps to generate the bounding box to be analyzed and obtain its minimum / maximum vertex coordinates min(x ia ,y ia , z ia ) and max(x ia ,y ia , z ia ).

[0026] Single-axis projection judgment: for each adjacent element i a , respectively check whether its projection on the x, y, and z axes overlaps with the target bounding box projection: x-axis overlap condition: mini .x≤max ia .x and max i .x≥min ia .x; Y-axis overlap condition: min i .y≤max ia .y and max i .y≥min ia .y; Z-axis overlap condition: min i .z≤max ia .z and max i .z≥min ia .z; If the projections of the three coordinate axes all meet the overlapping conditions, it is determined that there is a spatial conflict between elements i and ia, and an overlapping analysis signal is generated; If any coordinate axis projections do not overlap, it is determined that there is no conflict and normal design information is generated.

[0027] Assume that an underground pipe network project includes the following elements: Sewage pipeline (element i=1): pipe diameter DN300, buried depth 2m, length 10m, coordinates (100, 100, -2); Gas pipeline (adjacent element i1): pipe diameter DN150, buried depth 1.5m, horizontal distance from the sewage pipeline 1.8m (less than the preset value D=2m), coordinates (101, 100, -1.5).

[0028] Sewage pipeline (target element i=1): Target bounding box: minimum vertex min i (95, 98.5, -2.15), maximum vertex max i (105, 101.5, -1.85); Center point coordinates O1(100, 100, -2).

[0029] Gas pipeline (adjacent element i1): Bounding box to be analyzed: minimum vertex min i1(100.25, 99.25, -1.65), maximum vertex max i1(101.75, 100.75, -1.35).

[0030] X-axis projection: min i .x=95≤maxi1.x=101.75 and max i .x=105≥mini1.x=100.25, which satisfies the overlap condition; Y-axis projection: min i .y=98.5≤maxi1.y=100.75 and max i .y=101.5≥mini1.y=99.25, which satisfies the overlap condition; z-axis projection: min i .z=-2.15≤maxi1.z=-1.35 and max i .z=-1.85≥mini1.z=-1.65, which satisfies the overlap condition.

[0031] Since the x-, y-, and z-axis projections all overlap, the system determines that there is a spatial conflict between the sewage pipeline and the gas pipeline and generates an overlap analysis signal.

[0032] Step 4: Process the generated overlapping analysis signal to obtain adjacent elements that overlap with the target element, record them as elements to be analyzed, and analyze the design properties of the elements to be analyzed. The design properties here represent the hardness or softness of the corresponding element material. Specifically, they are distinguished by the elastic modulus E. Elements with an elastic modulus E greater than 5 GPa are classified as hard elements, and elements with an elastic modulus E less than 5 GPa are classified as non-hard elements. Based on the design properties, hard overlapping elements and non-hard overlapping elements are obtained. For example, valves and manholes are hard overlapping elements, while pipelines are non-hard overlapping elements. The hard overlapping elements and non-hard overlapping elements obtained by classification are then analyzed separately. Analyze the obtained hard overlapping elements, calculate the overlapping distance between the hard overlapping elements and the target element, and express the overlapping distance as the maximum straight-line distance of the overlapping parts of the two. Compare the obtained overlapping distance with the corresponding preset distance value, and the specific value of the preset distance value is set by the actual situation of the corresponding hard overlapping elements and the target element. If the overlapping distance is greater than the preset distance value, it indicates that the current hard overlapping element has a design conflict, and generates conflict design information. If the overlapping distance is less than the preset distance, it indicates that the current hard overlapping element can be designed normally, and generates an impact analysis signal. Then the generated impact analysis signal is analyzed, and the overlap level is determined according to the obtained overlap distance. The overlap level is divided into level one, level two and level three. At the same time, the overlap distance corresponding to level one to level three gradually increases, and the overlap level information is generated at the same time.

[0033] The obtained non-rigid overlapping elements are analyzed to determine whether their design affects the target elements. If not, normal design information is generated. Otherwise, conflicting design information is generated.

[0034] Example 2: Please refer to Figure 2This application provides an underground pipe network construction drawing generation and verification system based on image recognition, which includes: an image data acquisition module, a data verification and analysis module, a pipe network construction conflict analysis module, an adaptive processing module and a verification information output module, and combines Figure 2 It can be known that the functional modules are electrically connected in a unidirectional manner.

[0035] An image data acquisition module is used to collect image data of the underground pipe network area, perform preprocessing to obtain a preprocessed image, extract image features and establish a three-dimensional model, generate construction drawings, and transmit them to the data verification and analysis module. The specific processing method is similar to the processing process of step 1 in embodiment 1; The data verification and analysis module is used to verify and analyze the obtained construction drawings, compare the pipe network elements in the construction drawings with the original data, and generate a conflict verification analysis signal if the two are the same, which is transmitted to the pipe network construction conflict analysis module. If the two are different, the different elements are obtained to generate missing information and transmitted to the verification information output module. The specific processing method is the same as the processing process of step 2 in embodiment 1; The pipeline construction conflict analysis module is used to process the obtained conflict verification analysis signal, label each pipeline network element and determine the corresponding adjacent elements in order, then use the minimum bounding box algorithm to determine the overlap between the pipeline network element and its adjacent elements, generate normal design information or overlap analysis signal, and transmit the overlap analysis signal to the adaptive processing module. The specific processing method is similar to the processing process of step 3 in embodiment 1; An adaptive processing module is used to analyze the acquired overlap analysis signal, obtain overlapping adjacent elements, and classify them into hard overlapping elements and non-hard overlapping elements according to their design properties; analyze the hard overlapping elements, calculate their corresponding overlap distances and compare them with preset distance values, generate conflict design information or impact analysis signals, and generate overlap level information based on the overlap distances, and transmit it to the verification information output module; Analyze the non-rigid overlapping elements, determine the impact on the pipe network elements, generate normal design information or conflicting design information, and transmit it to the verification information output module. The specific processing method is the same as the processing process of step 4 in embodiment 1; The verification information output module is used to display the obtained omission information, overlapping level information, normal design information and conflicting design information to the corresponding management personnel.

[0036] Some of the data in the above formulas are calculated based on their numerical values and are not substituted into parameter units for calculation. At the same time, the contents not described in detail in this specification belong to the existing technology known to those skilled in the art.

[0037] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A method for generating and verifying underground pipe network construction drawings based on image recognition, characterized in that: The method specifically comprises the following steps: Collect image data of the underground pipe network area and preprocess it to obtain preprocessed images. At the same time, extract image features and build a three-dimensional model to generate construction drawings; Verify and analyze the construction drawings, compare the pipe network elements in the construction drawings with the original data, and generate conflict verification analysis signals if the two are the same. If the two are different, obtain the different elements to generate missing information; Process the conflict check analysis signal, label each pipe network element and determine the corresponding adjacent elements in order, then use the minimum bounding box algorithm to determine the overlap between the pipe network element and its adjacent elements, and generate normal design information or overlap analysis signal; Analyze the acquired overlap analysis signal, obtain overlapping adjacent elements and classify them into hard overlapping elements and non-hard overlapping elements according to their design properties, analyze the hard overlapping elements, calculate their corresponding overlap distances and compare them with preset distance values, generate conflict design information or impact analysis signals, and generate overlap level information based on the overlap distances; Analyze non-rigid overlapping elements, determine their impact on pipe network elements, and generate normal design information or conflicting design information.

2. The method for generating and verifying underground pipe network construction drawings based on image recognition according to claim 1, characterized in that: The specific method of generating the construction drawings is as follows: A high-definition camera equipped with a drone is used to collect image data of the underground pipeline network area, and noise reduction, grayscale, and binarization preprocessing operations are performed to obtain preprocessed images. Image recognition algorithms are used to extract image features. Combined with geographic information system technology, a three-dimensional model of the underground pipeline network is constructed, and construction drawings are automatically generated in accordance with the drawing specifications and requirements of construction drawings.

3. The method for generating and verifying underground pipe network construction drawings based on image recognition according to claim 1, characterized in that: The specific method of verifying and analyzing the construction drawings is as follows: Obtain all pipe network elements in the construction drawing and compare them with the original data. If the pipe network elements are the same as the original data, the construction drawing is complete and a conflict detection analysis signal is generated. Conversely, if the pipe network elements are different from the original data, the construction drawing is incomplete and an incomplete analysis signal is generated. For incomplete analysis signal analysis, the corresponding incomplete elements are obtained based on the original data, and the missing information is generated based on it.

4. The method for generating and verifying underground pipe network construction drawings based on image recognition according to claim 1, characterized in that: The specific method of processing the conflict check analysis signal is as follows: Get all the pipe network elements and label them as i, and i=1, 2, ..., j, where j represents the type of pipe network element. Then, according to the order of pipe network element i, get the adjacent elements whose distance is less than the threshold and label them as i. a , and a=1, 2, …, b, where b represents the type of adjacent elements; Then, the pipe network element i is taken as the target element, and its corresponding minimum rectangular bounding box is established as the target bounding box. The three-dimensional coordinate system is established with the center point of the bounding box as the origin, and the maximum vertex coordinate max (x i ,y i , z i ) and the minimum vertex coordinate min(x i ,y i , z i ), similarly, establish the minimum rectangular bounding box corresponding to the adjacent elements and record it as the bounding box to be analyzed, and at the same time obtain the maximum fixed point coordinates and minimum vertex coordinates min (x ia ,y ia , z ia ), and analyze based on the established minimum rectangular bounding box.

5. The method for generating and verifying underground pipe network construction drawings based on image recognition according to claim 4, characterized in that: The specific method of performing analysis based on the established minimum rectangular bounding box is: According to the adjacent element number i a Analyze the size relationship between its projection on the x, y and z axes and the projection of the target bounding box in turn. The specific judgment relationship is as follows: x-axis projection overlap: min i .x≤max ia .x and max i .x≥min ia .x; y-axis projection overlap: min i .y≤ma ia .y and max i .y≥min ia .y; z-axis projection overlap: min i .z≤ma ia .z and max i .z≥min ia .z; If any set of projections on the x, y, and z axes do not overlap, it means that there is no overlap between the two, and normal design information is generated. Conversely, if the projections on the x, y, and z axes all overlap, it means that there is overlap between the two, and an overlap analysis signal is generated.

6. The method for generating and verifying underground pipe network construction drawings based on image recognition according to claim 5, characterized in that: The specific method of analyzing the acquired overlapping analysis signals is as follows: Adjacent elements that overlap with the target element are obtained and recorded as elements to be analyzed. At the same time, the design properties of the elements to be analyzed are analyzed, and they are classified according to the design properties to obtain hard overlapping elements and non-hard overlapping elements. Specifically, elements with an elastic modulus E greater than 5 GPa are classified as hard elements, and elements with an elastic modulus E less than 5 GPa are classified as non-hard elements. The classified hard overlapping elements and non-hard overlapping elements are analyzed separately.

7. The method for generating and verifying underground pipe network construction drawings based on image recognition according to claim 6, characterized in that: The specific method of separately analyzing the hard overlapping elements and non-hard overlapping elements obtained by the classification is as follows: Analyze the obtained hard overlapping elements, take the maximum straight-line distance between the overlapping parts of the two elements and compare it with the preset distance value. If the overlapping distance is greater than the preset distance value, generate conflict design information; if the overlapping distance is less than the preset distance, generate an impact analysis signal; Then, the generated impact analysis signal is analyzed, the overlap level is determined according to the obtained overlap distance, and overlap level information is generated at the same time; The obtained non-rigid overlapping elements are analyzed to determine whether their design affects the target elements. If not, normal design information is generated. Otherwise, conflicting design information is generated.

8. An underground pipe network construction drawing generation and verification system based on image recognition, used to execute the underground pipe network construction drawing generation and verification method according to any one of claims 1 to 7, characterized in that: include: Image data acquisition module, data verification and analysis module, pipeline network construction conflict analysis module, adaptive processing module and verification information output module; Image data acquisition module, which is used to collect image data of the underground pipe network area, perform preprocessing to obtain preprocessed images, extract image features and build a three-dimensional model, generate construction drawings, and transmit them to the data verification and analysis module; The data verification and analysis module is used to verify and analyze the obtained construction drawings, compare the pipe network elements in the construction drawings with the original data, and generate a conflict verification analysis signal if the two are the same. The signal is transmitted to the pipe network construction conflict analysis module. If the two are different, the different elements are obtained to generate missing information and transmit it to the verification information output module. The pipeline network construction conflict analysis module is used to process the obtained conflict verification analysis signal, label each pipeline network element and determine the corresponding adjacent elements in order, then use the minimum bounding box algorithm to determine the overlap between the pipeline network element and its adjacent elements, generate normal design information or overlap analysis signal, and transmit the overlap analysis signal to the adaptive processing module; An adaptive processing module is used to analyze the acquired overlap analysis signal, obtain overlapping adjacent elements, and classify them into hard overlapping elements and non-hard overlapping elements according to their design properties; analyze the hard overlapping elements, calculate their corresponding overlap distances and compare them with preset distance values, generate conflict design information or impact analysis signals, and generate overlap level information based on the overlap distances, and transmit it to the verification information output module; Analyze non-rigid overlapping elements, determine their impact on pipe network elements, generate normal design information or conflicting design information, and transmit it to the verification information output module; The verification information output module is used to display the obtained omission information, overlapping level information, normal design information and conflicting design information to the corresponding management personnel.

Images