Unmanned aerial vehicle intelligent inspection mode and monitoring method for reservoir bank slope

Abstract

The invention discloses a reservoir bank slope unmanned aerial vehicle intelligent inspection mode and a monitoring method, and relates to the technical field of water conservancy and hydropower engineering reservoir bank slope safety monitoring. According to the invention, four inspection modes (annual large-range inspection, conventional key area inspection, special situation inspection and emergency situation inspection) based on the unmanned aerial vehicle are provided, and the full life cycle demand of reservoir bank slope monitoring is covered; through combination of different inspection modes, monitoring requirements in different scenes can be flexibly met, annual large-scale overall monitoring is ensured, high-frequency and high-precision monitoring can be performed on key areas, and meanwhile, quick response can be performed under special conditions and emergency conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of safety monitoring of reservoir bank slopes in water conservancy and hydropower projects, and specifically relates to an intelligent inspection mode and monitoring method for reservoir bank slopes using unmanned aerial vehicles (UAVs). The aim is to achieve intelligent monitoring and early warning of slope conditions by proposing different inspection modes and selecting targeted inspection paths and monitoring methods. Background Art

[0002] The statements in this section only provide background information related to the present disclosure and may not constitute prior art.

[0003] The stability of reservoir bank slopes is an important guarantee for the safety of water conservancy and hydropower projects. Traditional slope monitoring mainly relies on manual inspections or fixed monitoring equipment, which has problems such as low efficiency, high cost, limited coverage, and is difficult to cope with complex terrains and adverse weather conditions. With the rapid development of UAV technology, the application of UAVs in slope monitoring has gradually become a research hotspot. However, existing UAV inspection technologies lack systematic inspection modes and efficient monitoring methods, and are difficult to meet the requirements of large-scale, high-precision, and high-efficiency slope monitoring.

[0004] In the prior art, "An Automatic Inspection and Data Processing Method for Dike Diseases Using Unmanned Aerial Vehicles" (Applicant: Nanjing Automation Institute of Water Conservancy and Hydrology, Ministry of Water Resources) proposed a full-coverage inspection method for dikes based on customized flight trajectories, and realized the identification of dike diseases through UAV fixed-orbit, fixed-speed, and fixed-attitude cruising, training of disease image detection models, and digital warehousing. However, this technology has the following limitations:

[0005] High complexity of data processing: Its image enhancement, segmentation, and disease extraction rely on complex algorithms, which have strict requirements for computing resources and storage capabilities, and are prone to data processing delays and storage pressure during large-scale inspections;

[0006] Lack of real-time performance: The inspection and data processing processes take a long time, and it is difficult to achieve real-time disease feedback, especially the response ability is limited in emergency scenarios such as flood season emergencies;

[0007] Single inspection mode: It adopts a full-coverage automatic inspection scheme and does not design differentiated inspection strategies for different working conditions (such as regular monitoring, special meteorological conditions, emergency disasters), resulting in limited monitoring accuracy and efficiency. Specifically, for dam body deformation and crack monitoring, higher-resolution images and weather adaptability are required, and the existing technology does not involve flight parameter optimization and multi-mode collaboration mechanisms for high-precision monitoring requirements.

[0008] In addition, although DL / T 5796-2019 "Technical Specification for Slope Safety Monitoring of Hydropower Projects" mentions that drones can be used for inspection, it lacks specific implementation standards, resulting in a lack of systematic methods for drone monitoring of reservoir bank slopes. It is difficult to balance the requirements of large-scale coverage, high-frequency monitoring of key areas, and emergency rapid response. The defects of the existing technology indicate that there is an urgent need for an intelligent drone inspection system for reservoir bank slopes that combines multiple modes and balances accuracy and efficiency. Summary of the Invention

[0009] The purpose of the present invention is to provide an intelligent drone inspection mode and monitoring method for reservoir bank slopes in view of the problems existing in the prior art. By proposing different inspection modes and selecting targeted inspection paths and monitoring methods, intelligent monitoring and early warning of slope conditions are realized.

[0010] The technical solution of the present invention is as follows:

[0011] An intelligent drone inspection mode and monitoring method for reservoir bank slopes, comprising:

[0012] Step S1: Design four inspection modes, including: annual large-scale inspection mode, regular key area inspection mode, special situation inspection mode, or emergency situation inspection mode;

[0013] Step S2: Configure the flight parameters of the drone for the selected inspection mode and conduct inspections and monitoring.

[0014] Further, the annual large-scale inspection mode specifically includes:

[0015] Set the flight height to 60-100m, the forward overlap rate to 85%, and the side overlap rate to 75%;

[0016] Adopt the single-lens vertical photography mode to obtain three-dimensional aerial photography data of the entire slope;

[0017] Construct a three-dimensional real-scene model including a digital orthophoto map;

[0018] Execute the overall slope stability assessment once a year.

[0019] Further, the regular key area inspection mode specifically includes:

[0020] Set at least 2 main monitoring sections, and the auxiliary monitoring sections are dynamically adjusted according to the slope range;

[0021] Based on the three-dimensional real-scene model, conduct flight path planning to form an automatic aerial photography inspection path. Adopt a flight height of 16-20m, implement strip-shaped flight lines at the monitoring sections, and implement grid flight lines or circular flight lines in key areas with potential hazards.

[0022] Furthermore, the conventional key area inspection mode is used to inspect the disease conditions and abnormal conditions of the slope. The disease conditions include: cracks, collapse, and apparent water seepage of the slope; the abnormal conditions include: whether the slope drainage channel is blocked.

[0023] Furthermore, the selection principles for key monitoring objects in the conventional key area inspection mode are as follows:

[0024] Crack monitoring objects: mainly the areas where obvious cracks appear in the slope shotcrete, and also include crack areas on some natural rock masses that may affect the safety of the overall project;

[0025] Monitoring objects for apparent water seepage in concrete: select unobstructed areas with signs of apparent water seepage in concrete or areas with whitening and yellowing due to calcium precipitation in concrete;

[0026] Collapse monitoring object: The riverside slope of Mabukan, with priority given to areas with complete gully aerial data;

[0027] Drainage channel anomaly detection object: In the collected aerial data, the aerial perspective provided by the drone aerial data.

[0028] Furthermore, the monitoring method under the conventional key area inspection mode includes:

[0029] After the first image collection, the images of each waypoint were analyzed and classified according to the key monitoring items, and the photos were numbered according to the location of the route and the number of each route photo;

[0030] When the data of the 2nd, 3rd, ... nth periods are collected, the data with the same number are compared in pairs and multiple periods are compared to determine whether the cracks have widened, whether there is collapse, whether there is water seepage disease, and to evaluate the stability of the slope.

[0031] Furthermore, for cracks and collapse diseases, the background difference method is used to monitor and analyze whether there are changes in the calculation area. The monitoring results are displayed in the form of a broken line and can be exported according to the required data presentation format.

[0032] Furthermore, the background difference method specifically includes:

[0033] Establish the grayscale background model of the first image;

[0034] Calculate the pixel-level grayscale difference matrix of subsequent images;

[0035] Set the dynamic threshold function to extract disease features.

[0036] Furthermore, the special situation inspection mode specifically includes:

[0037] It is started after felt earthquakes, rainstorms or sudden drops in reservoir water levels occur;

[0038] Reuse the track planning parameters of conventional key areas.

[0039] Furthermore, the specific emergency inspection mode includes:

[0040] Start an emergency response after landslides or collapses occur;

[0041] Implement the UAV hovering shooting mode in the disaster area to obtain three-dimensional morphological data of the disaster body;

[0042] Calculate the secondary disaster risk index and generate an emergency risk removal plan.

[0043] Compared with the existing technologies, the beneficial effects of the present invention are:

[0044] 1. Through innovative technical routes such as a multi-mode inspection system, three-dimensional real-scene model construction, intelligent disease identification, and rapid response mechanism, the present invention realizes the high efficiency, accuracy, and intelligence of reservoir bank slope monitoring, significantly improving the comprehensiveness of slope monitoring and the emergency response ability.

[0045] 2. The present invention proposes four inspection modes based on UAVs (annual large-scale inspection, conventional key area inspection, special situation inspection, and emergency situation inspection), covering the full life cycle requirements of reservoir bank slope monitoring. Through the combination of different inspection modes, it can flexibly meet the monitoring needs in different scenarios, ensuring both the overall annual large-scale monitoring and the high-frequency and high-precision monitoring of key areas, and being able to respond quickly in special situations and emergencies.

[0046] 3. Using the three-dimensional aerial photography data collected by UAVs, the present invention constructs a high-precision three-dimensional real-scene model and a digital orthophoto map (DOM), and conducts track planning based on this to achieve automated inspection. The three-dimensional real-scene model can not only be used for macroscopically judging the overall changes of the slope, but also provide an accurate track planning basis for subsequent key area inspections, improving the efficiency and accuracy of inspections.

[0047] 4. The present invention proposes an intelligent identification method for diseases such as cracks and collapses based on the background difference method. Through the comparative analysis of multi-period data, it automatically calculates the expansion trend of cracks, changes in water seepage diseases, etc. This method can quickly and accurately identify the small changes of the slope, especially the monitoring of diseases such as cracks and water seepage, being able to early warn the potential risks of the slope and improving the intelligent level of monitoring.

[0048] 5. In case of emergency, the drone can hover and shoot, quickly complete shooting, transmission, analysis and result output, achieving a fast response effect. This mechanism can quickly determine whether secondary disasters will occur when landslides, collapses and other disasters occur on the slope, providing timely data support for emergency rescue and slope reinforcement, and improving the efficiency of emergency response. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the intelligent inspection mode of the drone for the reservoir bank slope;

[0050] Figure 2 It is a schematic diagram of the specific frequency, equipment and functions of the annual large-scale inspection;

[0051] Figure 3 It is a schematic diagram of the specific frequency, equipment and functions of the routine inspection of key areas;

[0052] Figure 4 It is a working flow chart of the inspection and monitoring method for the routine key areas;

[0053] Figure 5 It is a schematic diagram of the specific frequency, equipment and functions of the inspection in special cases;

[0054] Figure 6 It is a schematic diagram of the specific frequency, equipment and functions in case of emergency. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] It should be noted that relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the phrase "including a..." does not exclude the existence of additional identical elements in the process, method, article or device including the element.

[0056] The features and performance of the present invention will be further described in detail below in conjunction with the embodiments.

[0057] Embodiment 1

[0058] Please refer to Figure 1 , an intelligent inspection mode and monitoring method of the drone for the reservoir bank slope, including:

[0059] Step S1: Design four inspection modes, including: annual large-scale inspection mode, regular key area inspection mode, special situation inspection mode, or emergency situation inspection mode;

[0060] Step S2: Execute the UAV flight parameter configuration for the selected inspection mode, and conduct inspections and monitoring.

[0061] In this embodiment, specifically, please refer to Figure 2 , the annual large-scale inspection mode specifically includes:

[0062] Set the flight altitude to 60 - 100m, the forward overlap rate to 85%, and the side overlap rate to 75%;

[0063] Adopt the single-lens vertical photography mode to obtain the three-dimensional aerial photography data of the entire slope;

[0064] Construct a three-dimensional real-scene model including a digital orthophoto map;

[0065] Execute the overall slope stability assessment once a year.

[0066] That is, the annual large-scale inspection refers to the inspection of the entire slope. Since its scope is often large, the number of photos taken by UAV aerial photography is large, and the data processing time is long. Therefore, the inspection is generally carried out once a year. In this inspection mode, the flight altitude of the UAV can be set to 60m - 100m, the flight line overlap rate can be set to 85% forward and 75% side, and the single-lens vertical photography mode is used to obtain the three-dimensional aerial photography data of the entire slope. In this inspection mode, a three-dimensional real-scene model needs to be constructed, that is, use software such as DJI Terra or ContextCapture to generate the required three-dimensional real-scene model and digital orthophoto map (DOM). The main functions of this inspection mode include: first, by querying the constructed three-dimensional model, it is possible to understand whether there are changes in the entire slope every year, especially by comparing the data of multiple years, to understand the overall situation of the reservoir bank slope, and to make a macroscopic judgment on the safety and stability of the entire slope; second, this three-dimensional model can be used as the data basis for the inspection track planning of key areas of the reservoir bank high slope.

[0067] In this embodiment, specifically, please refer to Figure 3 , the regular key area inspection mode specifically includes:

[0068] Set no less than 2 main monitoring sections, and the auxiliary monitoring sections are dynamically adjusted according to the slope range;

[0069] Based on the three-dimensional real scene model, flight path planning is carried out to form an automatic aerial survey and inspection path. A flight altitude of 16 - 20 m is adopted. Strip-shaped flight lines are implemented at the monitoring sections, and grid flight lines or circular flight lines are implemented in key areas with potential hazards.

[0070] In this embodiment, specifically, the conventional key area inspection mode is used to inspect the disease conditions and abnormal conditions of the slope. The disease conditions include cracks, collapses, and apparent water seepage on the slope. The abnormal conditions include whether the slope drainage channels are blocked.

[0071] In this embodiment, specifically, the selection principles for key monitoring objects in the conventional key area inspection mode are as follows:

[0072] Crack monitoring object: mainly the areas where obvious cracks appear in the slope shotcrete, and at the same time, it should also include the crack areas on some natural rock masses that may affect the overall project safety;

[0073] Concrete apparent water seepage disease monitoring object: select the areas without obstruction that have traces of concrete apparent water seepage or areas where concrete shows calcium precipitation, whitening, or yellowing;

[0074] Collapse monitoring object: for the slope on the riverside side of Mabakand, preferentially select the areas that contain complete gully aerial survey data;

[0075] Drainage channel anomaly detection object: in the aerial survey data collected, the aerial perspective provided by the unmanned aerial vehicle aerial survey data.

[0076] In this embodiment, it should be noted that in the conventional key area inspection mode, first refer to the inspection specifications, set no less than 2 main detection sections for the reservoir bank slope, and set auxiliary monitoring sections as needed. The larger the slope range, the more auxiliary monitoring sections can be set. At the same time, for some key areas such as potential landslide bodies and deformed bodies, separate monitoring sections also need to be set;

[0077] Flight path planning is carried out using the three-dimensional model constructed by the annual large-scale inspection mode to form an automatic aerial survey and inspection path. The inspection flight altitude is 16 - 20 meters. Strip-shaped flight lines can be used for the monitoring sections, and grid flight lines or circular flight lines can be used for key areas such as potential landslide bodies and deformed bodies; for the conventional key area inspection, the inspection frequency is 1 to 2 times per month.

[0078] In this embodiment, it should be noted that for the conventional key area inspection, key inspections are carried out on disease conditions such as cracks, collapses, and apparent water seepage on the slope, as well as abnormal conditions such as whether the slope drainage channels are blocked. The selection principles for these key monitoring objects are:

[0079] a. Principle for selecting objects for crack monitoring: In the collected aerial photography data, the main areas are those where obvious cracks appear in the slope shotcrete, and at the same time, it should also include the crack areas on natural rock masses that may affect the overall project safety. Try to ensure that the cracks occupy a large proportion in the aerial photography data, and the crack trend can ensure that the whole crack is within the aerial photography data, with obvious cracks and relatively large widths. This is to ensure that when using crack identification software, obvious broken lines can be fitted, or when making comparative judgments, conclusions can be quickly drawn.

[0080] b. Principle for selecting objects for monitoring the apparent water seepage disease of concrete: In the collected aerial photography data, mainly select the areas where the apparent water seepage disease points of the shotcrete can be clearly observed, with clear traces of apparent water seepage of the concrete or areas of calcium leaching, whitening, or yellowing of the concrete in the pictures, and few vegetation or other interfering or covering elements. When making comparative judgments, it can quickly judge whether the water seepage traces have changed and whether the calcium leaching phenomenon of the concrete has worsened, so as to draw an effective conclusion on whether the apparent water seepage disease of the concrete is abnormal.

[0081] c. Principle for selecting objects for collapse monitoring: In the collected aerial photography data, there is shotcrete on the left bank slope, and the risk of collapse disasters is small. It is mainly the slope on the riverside side of the horse-step embankment. Priority is given to selecting aerial photography data containing complete gullies. Through multi-period data comparison analysis and the analysis of collapse identification software, judge whether there is a risk of collapse and debris flow in these areas.

[0082] d. Principle for selecting objects for abnormal detection of drainage channels: In the collected aerial photography data, the UAV aerial photography data provides an aerial perspective. Through aerial photography data at different times and positions, the water volume changes of the drainage channel caused by different times and spaces can be intuitively analyzed, as well as whether there is a blockage phenomenon.

[0083] In this embodiment, specifically, the monitoring method in the conventional key area inspection mode includes:

[0084] After the images are collected for the first time, analyze and classify the images of each flight point according to the key monitoring items, and number these photos according to the position of the flight line and the number of each flight line photo;

[0085] When the data of the 2nd, 3rd,... nth periods are collected back, make pairwise comparisons and multi-period data comparisons of the data with the same number, so as to judge whether the cracks become wider, whether there is a collapse phenomenon, whether there is water seepage disease, and evaluate the stability of the slope.

[0086] In this embodiment, specifically, for crack and collapse diseases, the background difference method is used to monitor and analyze whether there are changes in the calculation area, the monitoring results are displayed in the form of a broken line, and can be exported according to the required data presentation format. That is, the analysis results of the image data of each monitoring section of the slope can be summarized to obtain the situation of key monitoring items, such as cracks, collapses, seepage diseases, whether the drainage channels are abnormal, etc. The results are comprehensively judged to determine whether the slope is abnormal. If it is abnormal, further detection or on-site investigation can be carried out to determine whether a warning needs to be issued.

[0087] In this embodiment, specifically, the background difference method specifically includes:

[0088] Establish a grayscale background model for the first-phase image;

[0089] Calculate the pixel-level grayscale difference matrix of subsequent images;

[0090] Set a dynamic threshold function for disease feature extraction.

[0091] That is, after the images of each flight point are collected for the first time, the images of each flight point are analyzed and classified according to the situation of key monitoring items (such as cracks, collapses, seepage diseases, whether the drainage channels are abnormal, etc.), and these photos are numbered according to the position of the flight line, the number of each flight line photo, etc.;

[0092] When the data of the second phase, the third phase,... the nth phase are collected, the data with the same number are compared pairwise and compared with multi-phase data, so as to judge whether the cracks become wider, whether there are collapse phenomena, and whether there are seepage diseases, and evaluate the stability of the slope;

[0093] For diseases such as cracks and collapses, the background difference method is adopted and analyzed using the compiled software. The background difference method is the background subtraction method. Its principle is to perform differential processing on the established background model and the grayscale image of the current frame, and judge whether the area where the current frame pixel point is located is in a moving state or belongs to the background according to the comparison result of the grayscale difference between the current frame pixel point and the background model pixel point and the set threshold. In this embodiment, the first-phase or previous-phase image is used as the background base map. At the same time, the crack measurement area or the collapse concern area is used as the key point for comparison, and the obvious feature points in these areas are used as the comparison objects, which are used as the key comparison and calculation areas for the subsequent algorithm. Then, the monitoring maps of the second phase, the third phase, etc. of the same monitoring part are imported. Each monitoring map received by the software will be automatically classified under the monitoring base map at the same location according to the geographical coordinates, and then the crack monitoring algorithm will be automatically run to calculate the expansion trend of the cracks. Finally, the monitoring results are automatically displayed in the form of a broken line and can be exported according to the required data presentation format. The analysis flow chart is as Figure 4 shown.

[0094] In this embodiment, please refer to Figure 5 , the special situation inspection mode specifically includes:

[0095] Start after a felt earthquake, heavy rain, or sudden drop in reservoir water level;

[0096] Reuse the track planning parameters of the conventional key areas.

[0097] That is, the special situation inspection mainly refers to the inspection carried out when there is a felt earthquake, precipitation intensity above heavy rain, sudden drop in reservoir water level, and other unsafe signs appear on the slope. After these special situations occur, timely inspect the key attention areas of the slope. The key inspection areas are the same as those of the conventional key areas in terms of inspection steps, methods, and the inspection frequency is that when a special situation occurs, inspection should be carried out in a timely manner.

[0098] In this embodiment, please refer to Figure 6 , the emergency situation inspection mode specifically includes:

[0099] Start an emergency response after a landslide or collapse occurs;

[0100] Implement the UAV hover shooting mode in the disaster area to obtain three-dimensional shape data of the disaster body;

[0101] Calculate the secondary disaster risk index and generate an emergency risk removal plan.

[0102] That is, the emergency situation inspection refers to the inspection of the disaster area when disasters such as landslides and collapses have occurred on the reservoir bank slope. The key monitoring is whether secondary disasters will occur in these areas, providing safety guarantees for emergency rescue and slope reinforcement and risk removal. In this inspection mode, the UAV should perform hover shooting, complete shooting - transmission - analysis - result output for the slope in the disaster area, achieve a rapid response effect, and timely judge whether secondary disasters will occur.

[0103] In this embodiment, it should also be noted that the method of using a multi-rotor UAV + visible light camera is used as the inspection equipment, such as the DJI M350 UAV + P1 camera.

[0104] The above-described embodiments only represent the specific implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation of the protection scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the technical solution of the present application, several deformations and improvements can still be made, and these all belong to the protection scope of the present application.

[0105] This Background of the Invention section is provided to generally present the context of the present invention. Work of the presently named inventors, to the extent it is described in this Background of the Invention section, and aspects of the work that are not prior art as of the filing date of this application, are neither expressly nor impliedly admitted to be prior art to the present invention.

Claims

1. An intelligent inspection mode and monitoring method for the bank slope of a reservoir by an unmanned aerial vehicle, characterized in that, Including: Step S1: Design four inspection modes, including: annual large-scale inspection mode, regular key area inspection mode, special situation inspection mode or emergency situation inspection mode; Step S2: Execute the drone flight parameter configuration for the selected inspection mode, and conduct inspections and monitoring.

2. The intelligent inspection mode and monitoring method for the reservoir bank slope drone according to claim 1, wherein The annual large-scale inspection mode specifically includes: Set the flight altitude at 60 - 100m, the forward overlap rate at 85%, and the side overlap rate at 75%; Adopt the single-lens vertical photography mode to obtain the three-dimensional aerial photography data of the entire slope; Construct a three-dimensional real-scene model containing a digital orthophoto map; Execute the overall slope stability assessment with a cycle of once a year.

3. The intelligent inspection mode and monitoring method of the reservoir bank slope drone according to claim 2, characterized in that, The regular key area inspection mode specifically includes: Set no less than 2 main monitoring sections, and the auxiliary monitoring sections are dynamically adjusted according to the slope range; Based on the three-dimensional real-scene model, conduct flight path planning to form an automatic aerial photography inspection path. Adopt a flight altitude of 16 - 20m, implement strip-shaped flight lines at the monitoring sections, and implement grid flight lines or circular flight lines in the key areas with potential hazards.

4. The intelligent inspection mode and monitoring method of a reservoir bank slope drone according to claim 3, characterized in that, The regular key area inspection mode is used to inspect the disease conditions and abnormal conditions of the slope. The disease conditions include: cracks, collapses, and apparent seepage of the slope. The abnormal conditions include: whether the slope drainage channels are blocked.

5. The intelligent inspection mode and monitoring method of the reservoir bank slope drone according to claim 4, characterized in that, The selection principles for the key monitoring objects in the regular key area inspection mode are as follows: Crack monitoring objects: mainly focus on the areas where obvious cracks appear in the slope shotcrete, and also include some crack areas on the natural rock mass that may affect the overall project safety; Concrete apparent seepage disease monitoring objects: Select the areas without occlusion that have traces of concrete apparent seepage or areas with calcium leaching, whitening, or yellowing of the concrete; Collapse monitoring objects: For the slope on the riverside side of Mabakand, preferentially select the areas containing complete gully aerial photography data; Drainage channel abnormal detection objects: In the aerial photography data collected, the aerial perspective provided by the drone aerial photography materials.

6. The intelligent inspection mode and monitoring method of a reservoir bank slope by an unmanned aerial vehicle according to any one of claims 3-5, characterized in that, The monitoring methods in the regular key area inspection mode include: After the images at each flight point are collected for the first time, analyze and classify the images according to the key monitoring items, and number these images according to the position of the flight line and the number of each flight line photo; When the data of the 2nd period, the 3rd period,... the nth period are collected, conduct pairwise comparison and multi-period comparison of the data with the same number, so as to judge whether the cracks become wider, whether there is a collapse phenomenon, whether there is a seepage disease, and evaluate the stability of the slope.

7. The intelligent inspection mode and monitoring method of the reservoir bank slope by drone according to claim 6, characterized in that, For crack and collapse diseases, use the background difference method for monitoring and analyze whether there are changes in the calculation area, display the monitoring results in the form of a broken line, and can be exported according to the required data presentation format.

8. The intelligent inspection mode and monitoring method of the reservoir bank slope drone according to claim 7, characterized in that, The background difference method specifically includes: Establish the gray background model of the first-phase image; Calculate the pixel-level gray difference matrix of the subsequent images; Set the dynamic threshold function for disease feature extraction.

9. The intelligent inspection mode and monitoring method of the reservoir bank slope drone according to claim 1, characterized in that The special situation inspection mode specifically includes: Start after an induced earthquake, heavy rain weather, or sudden drop in reservoir water level; Reuse the flight path planning parameters of the regular key area.

10. The intelligent inspection mode and monitoring method of a reservoir bank slope drone according to claim 1, characterized in that, The emergency situation inspection mode specifically includes: Start the emergency response after a landslide or collapse; Implement the drone hovering shooting mode in the disaster area to obtain the three-dimensional morphological data of the disaster body; Calculate the secondary disaster risk index and generate an emergency risk elimination plan.

Images