Method and system for assisting post-disaster road damage detection and repair by intelligent unmanned aerial vehicle

Multi-scale data is obtained through high and low altitude flight of drones and combined with GIS technology for analysis, which solves the problem that a single data cannot reflect the real situation of road damage, and achieves rapid and accurate detection and repair planning of post-disaster road damage, providing post-disaster rescue Scientific basis.

CN119091325APending Publication Date: 2024-12-06WUHAN UNIV
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202410842811.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-27
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

In the prior art, single high-altitude or low-altitude drone data is difficult to fully reflect the real situation of road damage, and the lack of effective data management and analysis tools leads to low efficiency and insufficient accuracy of post-disaster road detection.

Method used

Combined with the high and low altitude flight of UAVs to obtain multi-scale road damage data, data integration, analysis and spatial analysis are carried out through image recognition algorithms and geographic information system (GIS), and detailed damage detection reports and repair planning schemes are generated.

Benefits of technology

It has achieved rapid, accurate and comprehensive inspection and evaluation of post-disaster road damage, improved work efficiency and safety, and provided a scientific and reasonable basis for post-disaster restoration planning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119091325A_ABST
    Figure CN119091325A_ABST
Patent Text Reader

Abstract

The invention belongs to unmanned aerial vehicle, image processing and road engineering technologies, and particularly relates to an intelligent unmanned aerial vehicle assisted post-disaster road damage detection and restoration method and system, and the method comprises the following steps: carrying out the high-altitude and low-altitude aerial photographing of a post-disaster road through an unmanned aerial vehicle, and obtaining the image data of the surface of the road; carrying out preprocessing on the acquired image data; an image recognition algorithm is adopted to recognize the preprocessed image data, crack and pit damage conditions on the surface of the road are recognized, and a damage detection report is generated; in combination with a geographic information system, spatial analysis is carried out on the damage detection report, and spatial distribution and severity of the damaged road are determined; and according to a spatial analysis result, making a repair planning scheme including a repair sequence, a repair method and a repair material. According to the method, the data can be efficiently integrated, managed and analyzed, visual display and spatial analysis of the damaged road are realized, and visual and comprehensive road damage information is provided for decision makers.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the fields of unmanned aerial vehicles, image processing and road engineering technology, and in particular to a method and system for intelligent unmanned aerial vehicles to assist in post-disaster road damage detection and repair. Background Art

[0002] In the context of frequent natural disasters, rapid and accurate detection of road damage is crucial for post-disaster rescue and reconstruction. Traditional road damage detection mainly relies on manual field surveys, but this method is not only time-consuming and labor-intensive, but also often faces many safety and environmental restrictions after a disaster. Therefore, seeking an efficient and safe road damage detection method has become an urgent need in the industry.

[0003] With the rapid development of drone technology, it is widely used in various fields, including disaster monitoring and assessment. Drones have the ability to fly at high altitudes and can quickly capture large areas of road damage, providing important data for preliminary assessments. [1-2] At the same time, the drone's low-altitude flight capability also enables it to accurately detect minor cracks and other damage on the road surface. [3-4] , thereby providing more comprehensive damage information.

[0004] However, the data obtained from a single high-altitude or low-altitude flight often cannot fully reflect the true situation of road damage. In order to obtain more accurate and multi-scale road damage information, multiple high-altitude and low-altitude flights are required to collect data with different spatial resolutions. The large amount of image data generated in this process will bring great challenges to subsequent work if there is a lack of effective management and analysis tools.

[0005] Geographic Information System (GIS) technology provides strong support for the integration and analysis of road damage data [5] GIS can efficiently store, manage, query and analyze spatial data. By associating data with geographic locations, it can realize the visualization and spatial analysis of damaged roads. [6-7] Integrating the detection results of drones into GIS can not only improve the efficiency of data processing, but also provide decision makers with more intuitive and comprehensive road damage information, which helps to formulate more scientific and reasonable repair plans. Summary of the invention

[0006] The present invention provides a method and system for intelligent drone-assisted post-disaster road damage detection and repair, which is used to solve the defect in the prior art that single data often cannot fully reflect the actual situation of road damage, and realize comprehensive and efficient detection of post-disaster road damage, providing strong support for repair planning.

[0007] The present invention provides a method for detecting and repairing road damage after a disaster using an intelligent drone, comprising the following steps:

[0008] Use drones to take high-altitude and low-altitude aerial photos of roads after disasters to obtain image data of the road surface;

[0009] Preprocessing the acquired image data;

[0010] Use image recognition algorithms to identify pre-processed image data, identify cracks and potholes on the road surface, and generate damage detection reports;

[0011] Combined with geographic information system, spatial analysis of damage inspection reports was conducted to determine the spatial distribution and severity of damaged roads;

[0012] Based on the results of spatial analysis, a restoration plan is developed, including the restoration sequence, restoration methods and restoration materials.

[0013] According to a method for intelligent drone-assisted post-disaster road damage detection and repair provided by the present invention, preprocessing the acquired image data includes:

[0014] The road surface image data obtained by the drone is deeply processed and analyzed, and cracks, potholes and damage on the road surface are identified through feature extraction, classification recognition and target detection steps.

[0015] According to the method provided by the present invention, an intelligent drone-assisted post-disaster road damage detection and repair method uses an image recognition algorithm to identify pre-processed image data, identify cracks and potholes on the road surface, and generate a damage detection report, including:

[0016] The pre-processed image data is deeply processed and analyzed using image recognition algorithms. Through feature extraction, classification recognition and target detection, cracks, potholes and damage on the road surface are identified, and a damage detection report is generated.

[0017] According to a method for detecting and repairing road damage after a disaster assisted by an intelligent drone provided by the present invention, a spatial analysis of the damage detection report is performed in combination with a geographic information system to determine the spatial distribution and severity of the damaged roads, including:

[0018] An effective dense attention fusion network based on channel-related loss is constructed, including: feature extraction module, attention module, fusion module and classification and positioning module; the fusion module and feature extraction module use the pre-trained convolutional neural network ResNet to extract the features of the road image; the attention module weights the features through the attention mechanism to highlight the features of the damaged area; the fusion module fuses the features from different layers to obtain contextual information; the classification and positioning module classifies the fused features and determines the location and range of the damage; considering the correlation between feature channels to ensure that the network can learn the difference between different damage types, its mathematical expression is:

[0019]

[0020] C is the number of channels, f i and f j is the feature vector of the i-th and j-th channels, δ ij is the Kronecker function, which is 1 when i=j and 0 otherwise.

[0021] The present invention also provides an intelligent UAV-assisted post-disaster road damage detection and repair system, comprising: a UAV, an image acquisition device, a ground control station, a data processing and analysis module, and a geographic information system;

[0022] Image acquisition equipment is carried on the drone, including high-definition cameras and infrared sensors, which are used to capture images and temperature distribution data of road damage;

[0023] The ground control station is used for UAV flight control and real-time reception of image data;

[0024] Data processing and analysis module: processes and analyzes the collected image data to identify road damage;

[0025] Geographic Information Systems are used to integrate, store, manage and analyze road damage data.

[0026] According to an intelligent drone-assisted post-disaster road damage detection and repair system provided by the present invention, the data processing and analysis module adopts an effective dense attention fusion network based on channel-related loss, including: a feature extraction module, an attention module, a fusion module, and a classification and positioning module;

[0027] The fusion module and feature extraction module use the pre-trained convolutional neural network ResNet to extract the features of road images; the attention module weights the features through the attention mechanism; the fusion module fuses the features from different layers to obtain contextual information; the classification and positioning module classifies the fused features and determines the location and scope of the damage.

[0028] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for post-disaster road damage detection and repair assisted by an intelligent drone as described above is implemented.

[0029] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for intelligent drone-assisted post-disaster road damage detection and repair as described in any one of the above is implemented.

[0030] The present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned methods for intelligent drone-assisted post-disaster road damage detection and repair.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] The present invention combines the rapid, multi-scale road damage detection capabilities of UAV technology with the efficient data integration and analysis functions of the Geographic Information System (GIS), achieving rapid, accurate, and comprehensive detection and assessment of road damage. Through the high and low altitude flights of UAVs, the present invention can capture road damage data of different spatial resolutions, providing an important basis for the preliminary assessment of road damage. At the same time, with the support of GIS technology, the present invention can efficiently integrate, manage, and analyze these data, achieve visual display and spatial analysis of damaged roads, and provide decision makers with intuitive and comprehensive road damage information, thereby helping to formulate more scientific and reasonable post-disaster rescue and reconstruction plans. This comprehensive detection method not only improves work efficiency, but also greatly enhances safety and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0034] Figure 1 It is a schematic flow chart of a method for intelligent drone-assisted post-disaster road damage detection and repair provided by an embodiment of the present invention;

[0035] Figure 2 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] This embodiment combines the high- and low-altitude flight capabilities of drones and GIS technology to develop a method and system that can comprehensively and efficiently detect post-disaster road damage and provide strong support for repair planning.

[0038] This embodiment is implemented by the following technical solution, a method for intelligent drone-assisted post-disaster road damage detection and repair planning, including the following steps:

[0039] Step 1: Use a drone to take high-altitude and low-altitude aerial photos of the post-disaster roads to obtain image data of the road surface;

[0040] Step 2, preprocessing the acquired road surface image data;

[0041] Step 3: Use image recognition algorithm to identify the preprocessed image, automatically identify cracks, potholes and other damage on the road surface, and generate a damage detection report;

[0042] Step 4: Combined with Geographic Information System (GIS) technology, spatial analysis of the damage inspection report is performed to determine the spatial distribution and severity of damaged roads;

[0043] Step 5: Based on the results of spatial analysis, formulate a repair plan to guide post-disaster road repair work.

[0044] In step 1, the high-altitude and low-altitude flight capabilities of the drone are used to achieve rapid detection of road damage. In high-altitude flight mode, the drone can quickly cover the disaster area and obtain macroscopic road damage conditions; in low-altitude flight mode, the drone can capture minor cracks and other damage on the road surface, providing an important basis for detailed assessment of road damage.

[0045] Wherein, in said step 2, the acquired road surface image data is preprocessed, including operations such as denoising and enhancement, so as to improve the image quality.

[0046] Among them, in step 3, advanced image recognition and processing algorithms are used to deeply process and analyze a large amount of image data obtained by the drone aerial photography system. The road damage detection model is used to detect the distribution, severity and influencing factors of road damage, providing strong support for subsequent repair planning. An effective dense attention fusion network based on channel-related losses is constructed to enhance the performance of road damage detection. The main components of the network include: feature extraction module, attention module, classification and positioning module. Fusion module The feature extraction module uses a pre-trained convolutional neural network ResNet to extract features of road images. The attention module weights the features through the attention mechanism to highlight the features of the damaged area. The fusion module fuses features from different layers to obtain richer contextual information. The classification and positioning module classifies the fused features and determines the location and range of the damage. The correlation between feature channels is considered to ensure that the network can learn the difference between different types of damage. Its mathematical expression is:

[0047]

[0048] C is the number of channels, f i and f j is the feature vector of the i-th and j-th channels, δ ij is the Kronecker function, which is 1 when i=j and 0 otherwise.

[0049] In step 4, the processed damage data is integrated into the geographic information system. The spatial analysis function of GIS is used to perform multi-spatial resolution analysis on the damaged roads, including accurate positioning of the damage location, determination of the damage range, and visualization of the damage extent.

[0050] In step 5, based on the output results of the image processing and analysis system and the GIS system, combined with road engineering knowledge and repair experience, the system can develop a scientific repair plan. Develop a targeted repair plan, including the repair sequence, repair methods, and forecasts of required resources. Generate detailed inspection reports and repair recommendations based on the repair plan. Provide decision makers with comprehensive data support and scientific basis to help efficient post-disaster road repair work.

[0051] This embodiment also provides an intelligent UAV-assisted post-disaster road damage detection and repair planning system, including a UAV, an image acquisition device, a ground control station, a data processing and analysis module, and a geographic information system (GIS).

[0052] Drones: Select drones with high stability and long endurance to support long-term high and low altitude flight missions.

[0053] Image acquisition equipment: carried on the drone, including high-definition cameras, infrared sensors, etc., used to capture high-definition images and temperature distribution data of road damage.

[0054] Ground control station: responsible for the flight control of the UAV and the real-time reception of image data.

[0055] Data processing and analysis module: Process and analyze the collected image data to automatically identify road damage.

[0056] Geographic Information System (GIS): Used to integrate, store, manage and analyze road damage data.

[0057] When implementing it, 1. Plan the flight route and altitude of the drone according to the actual situation in the post-disaster area. Ensure that the drone can cover all roads that need to be inspected.

[0058] 2. High- and low-altitude flights and data collection. First, high-altitude flights are conducted to quickly capture large-scale road damage. High-definition cameras are used to capture the road panorama and obtain macroscopic damage information. Next, based on the data obtained from high-altitude flights, specific damaged areas are located and low-altitude flights are conducted. During low-altitude flights, high-definition cameras, infrared sensors and other equipment are used to accurately capture subtle damage such as road cracks and potholes.

[0059] 3. The collected image data is transferred to the data processing and analysis module. Using image recognition and processing algorithms, the damage on the road surface, including cracks, potholes, etc., is automatically identified. The damage is classified and rated according to the type and degree of damage. An effective dense attention fusion network based on channel-related loss is constructed to enhance the performance of road damage detection. The main components of the network include: feature extraction module, attention module, classification and positioning module. Fusion module The feature extraction module uses a pre-trained convolutional neural network ResNet to extract features of road images. The attention module weights the features through the attention mechanism to highlight the features of the damaged area. The fusion module fuses features from different layers to obtain richer contextual information. The classification and positioning module classifies the fused features and determines the location and range of the damage. The correlation between feature channels is considered to ensure that the network can learn the difference between different types of damage. The mathematical expression is:

[0060]

[0061] Where C is the number of channels, f i and f j is the feature vector of the i-th and j-th channels, δ ij is the Kronecker function, which is 1 when i=j and 0 otherwise.

[0062] 4. Integrate the processed damage data into the geographic information system. Using the spatial analysis function of GIS, conduct multi-spatial resolution analysis of damaged roads, including precise positioning of the damage location, determination of the damage range, and visualization of the damage extent.

[0063] 5. Based on the results of GIS analysis, assess the severity of road damage and repair needs. Develop targeted repair plans, including repair sequence, repair methods, and forecasts of required resources. Generate detailed inspection reports and repair recommendations based on the repair plan. Provide decision makers with comprehensive data support and scientific basis to help efficient post-disaster road repair work.

[0064] Figure 2 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 2 As shown, the electronic device may include: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus. The processor may call the logic instructions in the memory to execute the method for intelligent drone-assisted post-disaster road damage detection and repair.

[0065] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0066] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the methods for intelligent drone-assisted post-disaster road damage detection and repair provided by the above methods.

[0067] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the method for intelligent drone-assisted post-disaster road damage detection and repair provided by the above-mentioned methods.

[0068] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0069] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0070] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0071] References:

[0072] [1] Zhou Wenqing, Liu Gang. A review of overhead line defect inspection based on deep learning and drone images [J]. Electric Power Engineering Technology, 2024, 43(02): 73-82.

[0073] [2] She Jie, Shen Aihong, Shi Yun, et al., Desert steppe vegetation classification based on UAV remote sensing images and object-oriented technology [J / OL]. Journal of Grassland Science, 1-14 [2024-04-01].

[0074] [3] Li Ruxia, Research on post-earthquake damaged road identification based on UAV images[D], Wuhan University, 2019. DOI: 10.27379 / d.cnki.gwhdu.2019.000797.

[0075] [4] Wang Jiannan. Research on the application of UAV optical images in post-earthquake road clearance[D]. National University of Defense Technology, 2016.

[0076] [5] Wang Jiannan. Research on the application of UAV optical images in post-earthquake road clearance[D]. National University of Defense Technology, 2016. [J]. Automation and Instrumentation, 2023, (7): 1-8.

[0077] [6] Xu Yang, Li Qingquan, Tang Luliang. Road damage detection using remote sensing images based on road network vector data [J]. Bulletin of Surveying and Mapping, 2011, (04): 14-16+22.

[0078] [7] Sui Haigang, Hua Feng, Fan Yida, et al. Extraction of road damage information from high-resolution SAR images using GIS and Bayesian network [J]. Journal of Wuhan University (Information Science Edition), 2016, 41(05): 578-583. DOI: 10.13203 / j.whugis20140323.

Claims

1. A method for detecting and repairing road damage after a disaster using an intelligent drone, characterized in that: The following steps are involved: Use drones to take high-altitude and low-altitude aerial photos of roads after disasters to obtain image data of the road surface; Preprocessing the acquired image data; Use image recognition algorithms to identify pre-processed image data, identify cracks and potholes on the road surface, and generate damage detection reports; Combined with geographic information system, spatial analysis of damage inspection reports was conducted to determine the spatial distribution and severity of damaged roads; Based on the results of spatial analysis, a restoration plan is developed, including the restoration sequence, restoration methods and restoration materials.

2. The method for detecting and repairing road damage after a disaster using an intelligent drone as aid according to claim 1 is characterized in that: The preprocessing of the acquired image data includes: The road surface image data obtained by the drone is deeply processed and analyzed, and cracks, potholes and damage on the road surface are identified through feature extraction, classification recognition and target detection steps.

3. The method for intelligent drone-assisted post-disaster road damage detection and repair according to claim 2 is characterized in that: The image recognition algorithm is used to identify the pre-processed image data, identify cracks and potholes on the road surface, and generate a damage detection report including: The pre-processed image data is deeply processed and analyzed using image recognition algorithms. Through feature extraction, classification recognition and target detection, cracks, potholes and damage on the road surface are identified, and a damage detection report is generated.

4. The method for intelligent drone-assisted post-disaster road damage detection and repair according to claim 3 is characterized in that: Combined with the geographic information system, the damage inspection report is spatially analyzed to determine the spatial distribution and severity of damaged roads, including: An effective dense attention fusion network based on channel-related loss is constructed, including: feature extraction module, attention module, fusion module and classification and positioning module; the fusion module and feature extraction module use the pre-trained convolutional neural network ResNet to extract the features of the road image; the attention module weights the features through the attention mechanism to highlight the features of the damaged area; the fusion module fuses the features from different layers to obtain contextual information; the classification and positioning module classifies the fused features and determines the location and range of the damage; considering the correlation between feature channels to ensure that the network can learn the difference between different damage types, its mathematical expression is: C is the number of channels, f i and f j is the feature vector of the i-th and j-th channels, δ ij is the Kronecker function, which is 1 when i=j and 0 otherwise.

5. An intelligent drone-assisted post-disaster road damage detection and repair system, characterized in that: The system is used for the method for intelligent drone-assisted post-disaster road damage detection and repair as described in any one of claims 1 to 4; comprising: a drone, an image acquisition device, a ground control station, a data processing and analysis module, and a geographic information system; Image acquisition equipment is carried on the drone, including high-definition cameras and infrared sensors, which are used to capture images and temperature distribution data of road damage; The ground control station is used for UAV flight control and real-time reception of image data; Data processing and analysis module: processes and analyzes the collected image data to identify road damage; Geographic Information Systems are used to integrate, store, manage and analyze road damage data.

6. The intelligent drone-assisted post-disaster road damage detection and repair system according to claim 5 is characterized in that: The data processing and analysis module adopts an effective dense attention fusion network based on channel-related loss, including: feature extraction module, attention module, fusion module and classification and positioning module; The fusion module and feature extraction module use the pre-trained convolutional neural network ResNet to extract the features of road images; the attention module weights the features through the attention mechanism; the fusion module fuses the features from different layers to obtain contextual information; the classification and positioning module classifies the fused features and determines the location and scope of the damage.

7. An electronic 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 method for intelligent drone-assisted post-disaster road damage detection and repair as described in any one of claims 1 to 4 is implemented.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, the method for intelligent drone-assisted post-disaster road damage detection and repair as described in any one of claims 1 to 4 is implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by the processor, the method for intelligent drone-assisted post-disaster road damage detection and repair as described in any one of claims 1 to 4 is implemented.

Citation Information

Patent Citations

  • Method for testing road damage based on post-disaster high-resolution remote sensing image

    CN101614822A

  • Unmanned aerial vehicle vision-based autonomous inspection system along mountain road

    CN114089786A

  • Integrated intelligent vehicle-mounted detection system and method for surface cracks of asphalt pavement

    CN116563810A

  • Road damaged target detection method based on attention mechanism ensemble learning network

    CN116597270A

  • Gas supply apparatus having transference portion and assembling method of connector of gas supply apparatus

    KR102233470B1