Reservoir inspection method and system based on unmanned aerial vehicle integrated AI technology

Abstract

The invention discloses a reservoir routing inspection method and system based on an unmanned aerial vehicle integrated AI technology. The reservoir routing inspection method comprises the steps of routing inspection area division and route planning; flight is planned according to the route, data acquisition is carried out, and acquired data are obtained; carrying out data processing on the acquired data by adopting AI data processing to obtain a processing result; and performing AI data analysis according to the processing result. Automation of reservoir routing inspection is achieved, cost can be saved, risks are reduced, routing inspection efficiency and data analysis efficiency are improved, and a more convenient mode is provided for guaranteeing reservoir safety.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle (UAV) intelligent inspection, and in particular to a reservoir inspection method and system based on unmanned aerial vehicle (UAV) fusion AI technology. Background Art

[0002] The UAV intelligent inspection system is composed of UAVs, UAV airports, and UAV control platforms. It attracts all walks of life with its advantages of efficiency, safety, and intelligence. With the continuous advancement of UAV technology and the advent of the low-altitude economy era, UAV inspection technology will usher in a broader development prospect. In the future, UAV inspection equipment will be more miniaturized and intelligent, and its performance will be greatly improved. At the same time, UAV inspection technology will be deeply integrated with artificial intelligence, big data, the Internet of Things and other technologies, making it widely used in various industries.

[0003] Reservoir inspection is an important part of reservoir safety management. Through regular and irregular inspections, potential safety hazards can be discovered and dealt with in a timely manner, reservoir accidents can be prevented, and important basis can be provided for the maintenance and renovation of reservoirs to ensure the long-term safe operation of reservoirs. However, traditional inspections require manual inspections, which are time-consuming and risky, especially in harsh terrain and climatic conditions. Moreover, the results of manual inspections also have certain errors and limitations.

[0004] The shortcomings of traditional reservoir inspections include:

[0005] 1) High labor cost and low efficiency: The reservoir area is usually large, and the inspection scope is also large. If a full-range inspection is to be carried out, a large amount of manpower and material resources are required to achieve a comprehensive inspection, and the inspection result data also requires manual intervention for analysis and processing; High safety risk: Due to the influence of environmental terrain, the physical fitness and skills of the inspection personnel are required to be high, and there is also the risk of casualties and accidents;

[0006] 2) Limited monitoring scope: Traditional inspection methods often fail to fully cover all key parts of reservoir dams, especially some difficult-to-reach or dangerous areas. This may lead to some important safety hazards being overlooked, and remote monitoring cannot be achieved during the monitoring process, making it difficult to grasp the safety status of reservoir dams when inspection personnel cannot reach or cannot reach them in time. Summary of the invention

[0007] In view of the above problems, the present invention is proposed to provide a reservoir inspection method and system based on drone fusion AI technology to overcome the above problems or at least partially solve the above problems.

[0008] According to one aspect of the present invention, a reservoir inspection method based on drone fusion AI technology is provided, and the reservoir inspection method comprises:

[0009] Inspection area division and route planning;

[0010] Fly according to the route plan and collect data to obtain collected data;

[0011] Using AI data processing to process the collected data to obtain processing results;

[0012] Perform AI data analysis based on the processing results.

[0013] Optionally, the inspection area division and route planning specifically include:

[0014] According to the geographical location information of the inspection area, the number and location of the drone hangar, and the parameter information of the load camera, the inspection area division and route planning are automatically realized.

[0015] Optionally, the flying according to the route planning and collecting data to obtain the collected data specifically includes:

[0016] The aircraft can fly on a scheduled basis according to the established route through mission mode, and collect pictures and videos during the flight by carrying a specific camera.

[0017] Optionally, the AI ​​data processing is used to process the collected data, and the processing results obtained specifically include: real-time recognition and annotation of the video during the flight through image and video processing technology, and recognition and stitching of photos taken during the flight.

[0018] Optionally, performing AI data analysis based on the processing results specifically includes: analyzing historical reservoir data through an algorithm, and forming an analysis report based on the analysis results.

[0019] Optionally, the automatic division of inspection areas and route planning based on the geographical location information of the inspection areas, the number and location of drone hangars, and the parameter information of the payload cameras specifically include:

[0020] Obtain the geographical location information of the pre-inspection reservoir through the GIS interface, and obtain the location information of the river bank and dam;

[0021] According to the geographical location information of the inspection reservoir, the deployment location of the drone airport, and the parameter information of the drone-loaded camera, the area is automatically divided into multiple inspection areas, and the inspection routes within the area are generated;

[0022] Assign multiple areas to the drone airport closest to the area, and subsequent drones will inspect the area according to the planned routes.

[0023] Optionally, the real-time recognition and annotation of the video during the flight by using the image and video processing technology, and the recognition and splicing of the photos taken during the flight specifically include:

[0024] MinIO is used to store the large amount of pictures and video data generated during the flight;

[0025] Separately store the original data of the task execution process and the data after processing the original data;

[0026] For photos and videos taken during inspection tasks, Yolov5 algorithm is used for target detection and recognition. Videos can be identified and displayed in real time, making it convenient to view inspection results in real time.

[0027] For the orthophotos taken during the inspection process, two-dimensional puzzles are realized based on OpenDroneMap to provide a basis for comparative analysis of recognition results.

[0028] Optionally, the raw data includes: real-time dynamic data of the drone airport, photos taken during the flight and recorded videos, laser point cloud data

[0029] Optionally, the data obtained after the original data is processed specifically includes: recognition result annotated pictures, videos and spliced ​​pictures.

[0030] A reservoir inspection system based on drone fusion AI technology, applying the above-mentioned reservoir inspection method based on drone fusion AI technology, the inspection system includes:

[0031] Route planning module, used for inspection area division and route planning;

[0032] A data collection module is used to fly according to the route planning and collect data to obtain the collected data;

[0033] An AI data processing module is used to process the collected data using AI data processing to obtain a processing result;

[0034] The AI ​​data analysis module is used to perform AI data analysis based on the processing results.

[0035] The present invention provides a reservoir inspection method and system based on unmanned aerial vehicle fusion AI technology, the reservoir inspection method includes: inspection area division and route planning; flying according to the route planning and collecting data to obtain collected data; using AI data processing to process the collected data to obtain processing results; and performing AI data analysis based on the processing results. The automation of reservoir inspection is realized, which can save costs and reduce risks, and improve the efficiency of inspection and data analysis, providing a more convenient way to ensure reservoir safety.

[0036] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0038] Figure 1 A flow chart of a reservoir inspection method based on drone fusion AI technology provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0040] The terms "comprises" and "having" and any variations thereof in the description embodiments, claims and drawings of the present invention are intended to cover non-exclusive inclusions, for example, including a series of steps or units.

[0041] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0042] like Figure 1 As shown, a reservoir inspection method based on drone fusion AI technology includes:

[0043] According to the geographical location information of the inspected reservoir area, the drone and the load parameters, the inspection area is automatically divided and the routes within the area are generated, ensuring the maximization of the inspection range and the optimization of the inspection routes, reducing manual operations.

[0044] By formulating inspection tasks and using drones equipped with cameras to replace manual inspections, regular inspections are carried out according to established routes, and photos and videos of the reservoir site are obtained to ensure the real-time nature of inspection data; reduce manpower and material resources, and reduce the risks brought by manual inspections.

[0045] The collected videos and pictures are stored and stitched and identified using algorithms to replace manual processing, improve data processing efficiency, use distributed storage to increase data access speed, and ensure data security and integrity.

[0046] It can realize automatic analysis of inspection results, analyze historical data, analyze the latest data according to established standards, and form a report with certain format and push it to relevant personnel. It can also automatically issue alarms for abnormal phenomena found during the analysis, thus improving the accuracy and real-time performance of inspection results.

[0047] A reservoir inspection system based on unmanned aerial vehicle (UAV) fusion AI technology, including: a route planning module, a data acquisition module, an AI data processing module and an AI data analysis module.

[0048] Route planning module, including:

[0049] The route planning module supports both automatic planning of regional routes and manual drawing of routes through maps.

[0050] Automatic route planning obtains the geographical location information of the pre-inspected reservoir through the GIS interface, and obtains the location information of the river edge and the dam. According to the geographical location information of the inspected reservoir, the deployment location of the drone airport, and the parameter information of the drone-loaded camera, the area is automatically divided into different inspection areas, and the inspection routes within the area are generated. Each area is assigned to the drone airport closest to the area, and the drone will subsequently inspect the area according to the planned route.

[0051] Data acquisition module, including:

[0052] The data collection of this system is carried out by creating collection tasks. The task management module supports both ordinary tasks and scheduled tasks. By setting the expected inspection time and execution cycle of the scheduled tasks, the task management module can control the automatic execution of the tasks and realize the automation of data collection.

[0053] During the data collection process, the drone can be controlled through the flight control module. The flight control module realizes the flight control of the drone and the airport by calling the DJI API. During the flight, the platform's flight control module can be used to control the drone, including mission pause and resume, mission start and end, drone flight direction control (up, down, left, right, left turn, right turn, etc.), gimbal operation, airport operation, and point-to-point flight.

[0054] During the data collection process, video monitoring of the inspection process is achieved through the real-time monitoring module. Real-time videos from the perspective of all drones performing tasks and real-time videos of the airport can be viewed; attitude data information, weather data information and airport status information of the drone during the execution of the task can be received and displayed in real time; the flight trajectory of the drone can be dynamically displayed on the digital map.

[0055] After the inspection task is completed, the data (pictures, videos, etc.) generated during the task execution will be automatically transmitted back to the system. In addition, if the automatic transmission is interrupted, the system supports manual transmission to ensure the integrity of data collection.

[0056] AI data processing module, including:

[0057] The data processing module of this system uses MinIO to store a large amount of pictures, videos and other data generated during the flight. The original data of the task execution process (real-time dynamic data of the drone airport, photos taken during the flight, recorded videos, laser point cloud data, etc.) and the data after the original data is processed (recognition result annotated pictures, videos, spliced ​​pictures, etc.) are stored separately.

[0058] For photos and videos taken during inspection tasks, the Yolov5 algorithm is used for target detection and recognition. Videos can be identified and displayed in real time, making it convenient to view inspection results in real time.

[0059] For the orthophotos taken during the inspection process, two-dimensional puzzles are realized based on OpenDroneMap (ODM) to provide a basis for comparative analysis of recognition results.

[0060] AI data analysis module, including:

[0061] During the inspection, a two-dimensional puzzle screen is generated for the dam body, dam top, overflow dam, gate, upstream and downstream water levels, and the data is automatically identified according to the set identification targets (people, illegal fishing, illegal boats, floating objects on the water surface, illegal swimming, etc.). The identification results are sent to the relevant person in charge of the reservoir in the form of work orders for on-site processing. The processed results can be fed back to the platform system for secondary confirmation.

[0062] Perform image change monitoring and analysis on historical puzzle results, extract change differences, and form difference change curves (such as the change trend of dam cracks, the change trend of building shape, etc.).

[0063] The execution results of each task, the image processing results and the original image, whether a work order has been issued, whether the work order has been completed, and the data analysis results are generated into an inspection report and pushed to relevant personnel.

[0064] If the difference value changes beyond the preset threshold, an alarm will be automatically triggered, and processing suggestions will be given and pushed to relevant personnel.

[0065] The reservoir inspection system that integrates drones with AI technology includes route planning, data collection, AI data processing, and AI data analysis. Route planning is used to automatically realize inspection area division and route planning based on the geographical location information of the inspection area, the number and location of drone hangars, and the parameter information of the load camera; the data acquisition module flies regularly according to the established route through the mission mode, and collects pictures and videos during the flight by carrying a specific camera; the AI ​​data processing module uses image and video processing technology to perform real-time recognition and annotation of videos during the flight, and recognizes and splices photos taken during the flight; the AI ​​data analysis module analyzes the historical data of the reservoir through algorithms, and can form an analysis report based on the analysis results.

[0066] Beneficial effects: By using drones integrated with AI technology, the automatic division of inspection areas, automatic generation of inspection routes, automatic collection of inspection data, and automatic processing and analysis of inspection result data are realized, thus realizing the automation of reservoir inspection, which can not only save costs and reduce risks, but also improve the efficiency of inspection and data analysis, providing a more convenient way to ensure reservoir safety.

[0067] The above specific implementation methods further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above are only specific implementation methods of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A reservoir inspection method based on drone fusion AI technology, characterized in that: The reservoir inspection method comprises: Inspection area division and route planning; Fly according to the route plan and collect data to obtain collected data; Using AI data processing to process the collected data to obtain processing results; Perform AI data analysis based on the processing results.

2. According to claim 1, a reservoir inspection method based on drone fusion AI technology is characterized in that: The inspection area division and route planning specifically include: According to the geographical location information of the inspection area, the number and location of the drone hangar, and the parameter information of the load camera, the inspection area division and route planning are automatically realized.

3. According to claim 1, a reservoir inspection method based on drone fusion AI technology is characterized in that: The flying according to the route planning and collecting data, and obtaining the collected data specifically include: The aircraft can fly on a scheduled basis according to the established route through mission mode, and collect pictures and videos during the flight by carrying a specific camera.

4. According to claim 1, a reservoir inspection method based on drone fusion AI technology is characterized in that: The AI ​​data processing is used to process the collected data to obtain processing results, which specifically include: real-time recognition and annotation of videos during the flight by image and video processing technology, and recognition and splicing of photos taken during the flight.

5. According to claim 1, a reservoir inspection method based on drone fusion AI technology is characterized in that: The AI ​​data analysis based on the processing results specifically includes: analyzing the historical data of the reservoir through an algorithm, and forming an analysis report based on the analysis results.

6. The reservoir inspection method based on drone fusion AI technology according to claim 2 is characterized in that: The automatic division of inspection areas and route planning based on the inspection area geographical location information, the number and location of drone hangars, and the load camera parameter information specifically include: Obtain the geographical location information of the pre-inspection reservoir through the GIS interface, and obtain the location information of the river bank and dam; According to the geographical location information of the inspection reservoir, the deployment location of the drone airport, and the parameter information of the drone-loaded camera, the area is automatically divided into multiple inspection areas, and the inspection routes within the area are generated; Assign multiple areas to the drone airport closest to the area, and subsequent drones will inspect the area according to the planned routes.

7. The reservoir inspection method based on drone fusion AI technology according to claim 4 is characterized in that: The real-time recognition and labeling of the video during the flight by using the image and video processing technology, and the recognition and splicing of the photos taken during the flight specifically include: MinIO is used to store the large amount of pictures and video data generated during the flight; Separately store the original data of the task execution process and the data after processing the original data; For photos and videos taken during inspection tasks, the YoloV5 algorithm is used for target detection and recognition. Videos can be identified and displayed in real time, making it convenient to view inspection results in real time. For the orthophotos taken during the inspection process, two-dimensional puzzles are realized based on OpenDroneMap to provide a basis for comparative analysis of recognition results.

8. The reservoir inspection method based on drone fusion AI technology according to claim 7 is characterized in that: The original data includes: real-time dynamic data of the drone airport, photos taken during the flight, recorded videos, and laser point cloud data.

9. The reservoir inspection method based on drone fusion AI technology according to claim 7 is characterized in that: The data after the original data is processed specifically includes: recognition result annotated pictures, videos and spliced ​​pictures.

10. A reservoir inspection system based on drone fusion AI technology, using a reservoir inspection method based on drone fusion AI technology as described in any one of claims 1 to 9, characterized in that: The inspection system comprises: Route planning module, used for inspection area division and route planning; A data collection module is used to fly according to the route planning and collect data to obtain the collected data; An AI data processing module is used to process the collected data using AI data processing to obtain a processing result; The AI ​​data analysis module is used to perform AI data analysis based on the processing results.

Images