Large water area unmanned aerial photogrammetry image control point layout method and device

Abstract

The application discloses a large water area unmanned aerial vehicle photogrammetry image control point layout method and device, first collects water depth, flow velocity, wind speed and bottom data of each point, calculates branch steel wire length and weight counterweight; then combines aerial survey parameters and water surface texture features to calculate optimal layout spacing to generate grid points; after the assembled device is put into position, the CORS-RTK is used to collect the three-dimensional coordinates of the center reference of the floating plate. Then the unmanned aerial vehicle is controlled to shoot images, and the images are evaluated from two dimensions of target clarity and control point quantity, and the unqualified weak area is matched with encryption coefficient to supplement point positions and re-measure coordinates, and the iteration acceptance is completed through cyclic test flight and supplementary setting. The application has the advantages that the artificial experience extensive control mode is abandoned, the branch distributor is matched with the self-adaptive counterweight, the floating plate is ensured to be stably floated and easily identified, the imaging defects are eliminated in advance through the pre-flight closed loop iteration, the centimeter-level image control point POS data is obtained, the aerial triangulation solution is jointly constrained, and the large water area surveying and mapping precision is greatly improved.

Description

Technical Field

[0001] This invention belongs to the field of surveying and mapping engineering technology, specifically relating to a method and device for setting up image control points for unmanned aerial vehicle (UAV) photogrammetry over large water areas. Background Technology

[0002] With its advantages of mobility, low operating costs, and high data acquisition efficiency, UAV photogrammetry has been widely used in surveying and mapping engineering scenarios such as topographic mapping, water conservancy reservoir area surveying, river topographic monitoring, lake ecological surveying, and reservoir capacity calculation. When conducting aerial surveys in conventional survey areas such as paved roads, flat bare land, and rural construction land, fixed ground control points can be set up by means of on-site spraying of markers, prefabricated hard markers, and foldable ground targets. The markers are stable and the images are highly recognizable. Combined with a standardized grid layout scheme, it can meet the requirements of aerial triangulation (when a drone takes a bunch of aerial photos, each photo only knows the approximate position of the camera, lacking precise coordinates, camera tilt angle, and precise relative position between photos. Directly stitching the images together will result in skew, misalignment, and distortion; aerial triangulation uses the precise ground coordinates of the ground control points to automatically calculate the precise shooting position and camera attitude of each aerial photo, and at the same time calculate the true three-dimensional coordinates corresponding to all pixels on the ground, uniformly calibrating thousands of photos to the same geodetic coordinate system. This is the core process for producing orthophotos, 3D models, and topographic maps) and the accuracy specifications of the output mapping. The relevant deployment technology and marker structure have formed a mature technical system.

[0003] However, when conducting UAV photogrammetry in large water areas such as reservoirs, large lakes, and wide rivers, the special environment of the water area brings multiple inherent technical bottlenecks to the deployment of image control points. Existing land-based image control deployment schemes and conventional floating targets on the water surface are not suitable for the complex working conditions of large water areas. There are a number of long-standing and difficult-to-eradicate technical pain points in the industry, as follows: ① The water surface lacks a stable and fixed benchmark, naturally lacking reliable ground control conditions. The water surface lacks a rigid bearing base, making it impossible to directly bury or spray fixed image control markers. The open water surface has a monotonous texture and large areas of uniform grayscale, lacking naturally identifiable feature points. Relying solely on UAV-borne POS-assisted positioning without ground constraints makes aerial triangulation prone to drift, layering, and splicing misalignment, significantly reducing the planar and elevation accuracy of orthophotos and underwater topographic 3D models, failing to meet the accuracy requirements of large-scale hydraulic surveying standards such as 1:500 and 1:1000. A few near-shore land points can only constrain the edge of the survey area; the central water area and offshore deep water areas lack effective control, resulting in a severely insufficient geometric strength of the overall survey area control network.

[0004] ② Existing floating image control targets have poor anti-disturbance capabilities and cannot meet the required stability. Most commercially available surface aerial survey targets are simple floats with a single counterweight and a single anchor rope, resulting in a simple anchoring structure and poor force balance. Under the impact of water flow and wind, they are prone to horizontal drift, float overturning, and target tilting and twisting. In deep water areas, fast-flowing river sections, and windy weather, target offsets can reach decimeter or even meter levels. There is a systematic deviation between the target center coordinates acquired by RTK (Real-Time Kinematic) and the target's image position in aerial photographs, directly causing excessive errors in the data processing and failure of aerial triangulation adjustment. Furthermore, the single anchor rope structure is prone to problems such as anchor slippage, burying of the counterweight in underwater silt, and rope tilting and imbalance. Without an adaptive anchoring design to suit different bottom materials, water depths, and flow velocities, the stability and reliability of the target cannot be guaranteed.

[0005] ③ Anchoring parameters are set entirely based on the experience of the operators, lacking environmentally adaptive quantitative calculation logic, making it difficult to balance resources and accuracy. Traditional surface image control operations use uniform fixed rope lengths and fixed counterweights for deployment across the entire water area, failing to differentiate between water depth gradients, surface water flow velocities, surface wind speeds, and differences in riverbed sediment within the survey area: silty riverbeds have low friction, and uniform counterweights easily cause anchor slippage; gravel and rocky riverbeds have high friction, and uniform large counterweights result in wasted materials, transportation, and manpower; insufficient rope length in deep water areas causes the float to be dragged by the current, while excessive rope length in shallow water areas causes the buoy to swing widely. The entire set of anchoring parameters lacks standardized calculation formulas and graded matching rules, relying entirely on the field experience of personnel. This leads to anchoring failures at some points and over-deployment at others within the same water area, increasing operational material costs and failing to uniformly guarantee the stability of the buoy across the entire water area.

[0006] ④ The density of image control points (ARPCs) relies on manual experience and grid distribution, without being linked and matched with core parameters of UAV aerial surveying and water surface texture features, resulting in an imbalance in the density of ARPCs. Current water area ARPCs generally use uniformly spaced grids, with point spacing roughly set based on the survey area, without considering factors such as relative flight altitude, camera focal length, sensor pixel size, heading, and lateral image overlap to calculate GSD (Ground Sampling Distance). Furthermore, ARPCs are not differentiated based on the varying degrees of water surface texture density, such as open still water, rippling water, nearshore weed areas, and navigable channels. In actual operations, two types of problems easily arise: ARPCs are too sparsely distributed in the center of open reservoirs with extremely scarce texture, resulting in insufficient coverage of ARPCs per image and broken matching of corresponding points; ARPCs are overly densely distributed in nearshore weed and shallow water areas with rich texture, creating a large number of redundant points, significantly increasing the labor and time costs of RTK survey point deployment, target placement, and subsequent retrieval, resulting in poor operational economics.

[0007] ⑤ The workflow involves a one-time deployment followed by direct aerial photography, lacking a mechanism for feedback on test flight image quality and iterative rework. This results in delayed problem identification and high rework costs. The traditional workflow involves deploying all surface image control beacons at once, completing RTK coordinate acquisition, and then directly executing a full-line aerial photography. There is no pre-verification step to verify the effectiveness of the deployment plan. Due to optical interference such as water surface reflection, wave texture obstruction, and strong backlighting, some areas of the targets may exhibit quality defects in aerial images, such as blurred markings, unidentifiable crosshairs, and insufficient visible control points per image. These problems can only be detected during the data processing stage. Once issues such as voids in aerial triangulation stitching, regional misalignment, or substandard model accuracy occur, a second deployment and re-deployment of targets are necessary, requiring repeated RTK coordinate acquisition and aerial photography. This significantly extends the project duration and increases the redundant investment in vessels, personnel, and buoy consumables.

[0008] ⑥ Existing floating image control devices suffer from uneven stress distribution, weak resistance to current and wind, and insufficient standardization and adaptability. Conventional single-rope, single-point counterweight buoys rely solely on a single steel wire rope to pull the float. Under the influence of water flow and wind, the float experiences concentrated stress on one side, making it extremely prone to deflection and overturning. Without a three-line uniform anchoring structure, it is impossible to offset the impact of multi-directional water flow and wind through multi-directional counterweights. The float marker is always difficult to keep horizontal and directly facing the aerial camera, resulting in severe perspective distortion of the target in the image and further exacerbating the error in the data processing.

[0009] Patent publication number CN110053719A discloses a waterborne positioning buoy and positioning method. It uses a buoy, a retractable rope, and a conical anchor to achieve surface anchoring. A winding mechanism adjusts the rope length to fit the water depth, and a bottom balance block enhances the buoy's stability. This method can be used for deploying control points in waterborne aerial surveys, enabling basic surface positioning and solving the problem of simple buoys having rope lengths that cannot adapt to water depth. However, it only uses a single rope and single-point anchoring structure, relies solely on manual experience to match the counterweight, and uses a one-time uniform grid layout. It only provides a basic surface positioning reference and suffers from drawbacks such as easy buoy drift and overturning, the density of the layout entirely dependent on experience, the lack of prior image quality verification, and high rework costs. It is difficult to adapt to large-scale, high-precision UAV aerial survey operations in complex conditions such as large bodies of water, deep water, rapid currents, and strong winds.

[0010] Patent publication number CN119104037A discloses a method, system, and product for deploying UAV aerial survey image control points in water conservancy projects. By dividing the river survey area into blocks, calculating the baseline length of a single image based on image pixels and overlap, and determining the heading and lateral image control deployment spacing in conjunction with the mapping scale, it can reduce the number of image control points deployed along the riverbank and improve the deployment efficiency in narrow river survey areas. It can also link aerial survey camera parameters to plan the land control point grid. However, it only targets the deployment of fixed image control beacons on the riverbank and does not involve floating image control buoys in large water areas. It lacks an adaptive anchoring parameter calculation system corresponding to water surface depth, current, and wind force, and cannot solve a series of deployment problems caused by the lack of a fixed base in open water, easy buoy drift, and poor water surface image recognition. Therefore, it cannot be directly applied to aerial survey deployment in deep water areas offshore of large water areas such as reservoirs and wide lakes.

[0011] Therefore, the industry urgently needs an intelligent deployment method for UAV photogrammetry image control points in large water areas that can automatically calculate anchoring parameters based on the on-site water environment and optimize the deployment density by combining aerial survey and water surface texture features. At the same time, it should be equipped with a water surface image control device with multi-directional balanced anchoring and high anti-disturbance performance to solve the industry problems of poor stability, rough deployment, lack of quality verification, and high rework rate of existing technologies. Summary of the Invention

[0012] The problem this invention aims to solve is that the current field of large-area drone photography lacks a method that can address positioning difficulties and severe deviations in aerial triangulation calculations.

[0013] The present invention solves the above-mentioned technical problems through the following technical means: A method for setting up image control points for unmanned aerial vehicle (UAV) photogrammetry over large bodies of water includes the following steps: S1. Collection of environmental parameters of the deployment area: Collect environmental parameters of each preset point in the water area to be deployed. The environmental parameters include water depth, surface water flow velocity and wind speed above the water surface. S2. Adaptive Calculation of Deployment Parameters: Anchoring parameters are adaptively calculated based on environmental parameters at each location, including the length of the branch wire rope and the weight of the matching counterweight. A corresponding safety factor is selected based on the type of underwater substrate. Complete the quantitative calculation of the weight of the counterweight; S3. Optimization of Deployment Density and Aerial Survey Parameters: Calculating the optimal deployment spacing of control points by linking UAV aerial survey parameters with water surface texture features. Generate a grid layout scheme for the water area; determine the optimal grid spacing. The calculation is based on flight altitude, ground sampling distance (GSD), and effective ground coverage width of a single image; S4. Initial deployment: Deploy the water surface image control point device according to the deployment grid scheme and adaptive anchoring parameters; S5. RTK coordinate acquisition: The 3D geodetic coordinates (X0, Y0, Z0) of all deployed image control points are acquired using RTK. S6. Flight test image acquisition and quality assessment: Conduct flight test image acquisition and rapid on-site quality assessment of UAVs; hover above each control point and at the center of the grid to capture vertical images, and determine the identifiability of the floating board markings and the number of visible control points at the center of the grid, and delineate unqualified weak areas and areas that need attention and observation. S7. Targeted Supplementation: Address the weak areas identified in the assessment by adjusting the density coefficient. The control points were added in a graded and encrypted manner, and the RTK coordinates were remeasured after the addition was completed. S8. Iterative verification: Iteratively execute steps S6 to S7 until the quality of all water area images meets the standards and the image control points are deployed.

[0014] Breaking away from the traditional operational mode of deploying control points in large water areas based on experience without prior quality verification, this approach automatically matches buoy anchoring conditions to the actual water conditions and intelligently plans the spacing of control points based on aerial flight data and water surface texture, balancing the stability of the buoys and the rationality of the density of control points from the source. Furthermore, it introduces on-site evaluation of test flight images, tiered supplementation of weak areas, and iterative verification mechanisms. This allows for the early identification of control point imaging failures caused by water surface reflections and ripples before formal aerial photography, avoiding rework due to errors in later aerial triangulation calculations. This significantly reduces the costs of field vessels, labor, and consumables, while simultaneously improving the overall accuracy and operational efficiency of the aerial survey control network in large water areas.

[0015] Preferably, the formula for calculating the length of the branch wire rope in step S2 is as follows: The formula for calculating the weight of the hammer is: ;in The angle between the wire rope and the plumb line. Because of the water depth, For the length of the branch wire rope, The underwater equivalent water-blocking area of ​​the branch wire rope and counterweight. The equivalent windward area of ​​the floating platform. The surface water velocity, For wind speed, The weight of the hammer. This is for the safety factor.

[0016] The configuration of buoy weight and rope length has been changed from manual experience to standardized formula quantitative calculation. The weight is determined in combination with the differences in water depth, water flow, wind speed and riverbed sediment to adapt to different water conditions and effectively reduce the problems of buoy drift and weight waste.

[0017] Preferably, the formula for calculating the underwater equivalent water-blocking area of ​​the single branch steel wire rope and the counterweight is as follows: In the formula The diameter of the wire rope, For the length of the rope, The area of ​​the hammer facing the water.

[0018] Preferably, the formula for calculating the ground sampling distance in step S3 is as follows: The formula for calculating the effective area coverage width of a single image is as follows: , The formula for calculating the optimal layout spacing is as follows: In the formula, The pixel dimension of the image. , These are the heading and lateral overlap, respectively. These represent the effective coverage width in the heading and lateral directions, respectively. This represents the minimum equivalent number of pixels required to cover control points within a single image. The altitude at which the drone flies. For camera pixel size, The focal length of the camera; The sparsity coefficient of water texture; Ground sampling distance, The optimal spacing is determined by the spacing between the installations.

[0019] A quantitative correlation is established between flight parameters, camera parameters, water surface texture, and the spacing between points, replacing the traditional equidistant empirical point placement. The point density can be adjusted according to needs, saving deployment costs while ensuring the accuracy of aerial triangulation calculations.

[0020] Preferably, the image quality assessment criteria in step S6 include the identifiability of control points and the number of visible control points in the grid center image; both the identifiability of control points and the number of visible control points in the grid center image are categorized as qualified, pending, and unqualified. The qualified condition for the image control point identifiability is defined as red and white markings with a clear central cross without any reflective obstruction; the unqualified condition is defined as slight reflectivity but the markings are still identifiable; and the unqualified condition is defined as ripples or strong light that make the target completely unidentifiable. The qualified number of visible control points in the grid center image is defined as ≥4, the unqualified number is defined as 2~3, and the unqualified number is defined as 0~1.

[0021] The system sets up a dual three-level judgment standard for target clarity and the number of visible control points, enabling rapid on-site identification of weak areas without the need for complex office software processing, and providing a quantitative basis for fixed-point densification and supplementation.

[0022] Preferably, the formula for calculating the spacing in step S7 is as follows: In the formula For encryption coefficients, To add spacing for encryption.

[0023] By quantifying the encryption spacing using formulas, the encryption calculation logic for weak areas is unified, avoiding arbitrary encryption by humans that leads to redundant or insufficient encryption points, and achieving standardized control over the supplementary spacing.

[0024] Preferably, the The value of is no greater than 1, and it is set as follows: if the image stitching fails or displays holes, it is considered severely weak. Take a value of 0.3~0.4; if there is obvious misalignment in the splicing, it is considered moderately weak. Take a value of 0.4~0.6; if there is slight misalignment in the splicing, it is considered slightly weak, and at this time... Take a value of 0.6 to 0.7.

[0025] The present invention also discloses an apparatus for setting up image control points using the aforementioned method for unmanned aerial vehicle (UAV) photogrammetry in large waters, including a float, a splitter, branch steel wire ropes, and a cast iron counterweight; The splitter includes an upper fixed rod, a lower fixed rod, and an inclined connecting rod; the branch wire ropes include multiple ropes, one end of which is connected to the inclined connecting rod; the center of the inclined connecting rod is connected to the lower end of the lower fixed rod, and the upper end of the lower fixed rod is connected to the lower end of the upper fixed rod; the upper fixed rod passes through the float and is located at the exact center of the float.

[0026] The multi-directional uniform inclined structure of the splitter is adopted to offset the impact of water flow and wind in multiple directions, prevent the float from drifting and overturning, ensure that the buoy marker is horizontal and facing upward, and improve the aerial photography recognition effect.

[0027] Preferably, the length of the branch wire rope is adapted to the changes in water depth and the vertical angle of the wire rope, and the weight of the cast iron hammer is selected according to the water depth, water flow, wind speed, and bottom sediment.

[0028] The device's rope length and counterweight can be selected as needed in accordance with the adaptive calculation process described above. The device is highly compatible with intelligent deployment methods and adaptable to complex and ever-changing aquatic environments.

[0029] Preferably, the float is a square rigid float, and the side floating on the water surface is coated with reflective pigment.

[0030] The square rigid buoy is highly recognizable and not easily deformed; the waterproof paint on the surface prevents the markings from fading and falling off, increases the recognizability of the markings when photographed, extends the lifespan of the buoy in the field, and reduces the cost of replacing consumables.

[0031] The advantages of this invention are: (1) Anchoring parameter environmental adaptive calculation: abandoning the traditional extensive mode of relying on manual experience to uniformly configure rope length and counterweight, the quantitative calculation formula of multiple parameters such as water depth, flow velocity, wind speed and riverbed bottom is used to accurately match the wire rope length and weight of the counterweight at each point; with the balanced force buoy device, the impact of water flow and wind is offset in multiple directions, solving the problem of buoy drift and overturning from both calculation and hardware levels, and adapting to various complex large water conditions such as deep water, rapid current, silt or rocky riverbed; (2) Optimal spacing is calculated in conjunction with the positioning, taking into account both accuracy and deployment cost: Instead of using fixed grid equidistant points, a quantitative model is established based on GSD, image overlap, and water surface texture sparsity coefficient to differentiate the spacing of control points in open still water, rippled channels, and near-shore aquatic plant areas; This ensures that each aerial image contains sufficient control points, enhances the accuracy of aerial triangulation, and avoids excessive deployment of points, which would lead to waste of manpower and buoy materials; (3) Avoid rework: The closed-loop process of "flight test evaluation - regional classification - targeted densification - iterative re-inspection" is used to quickly identify weak imaging areas caused by reflection and ripples on site through dual three-level judgment standards. The corresponding densification coefficient is matched according to the severity of the defects and layered supplementation is carried out. The control point matching defects are eliminated in advance before the formal aerial photography, avoiding the high time and material costs of renting a boat and carrying out construction again after problems are found in the office after the traditional one-time deployment. (4) The device has a multi-directional balancing function: the dedicated buoy structure with uniform circumferential oblique pull of the matching splitter is different from the defect of poor anti-interference ability of traditional single rope buoys; the buoy rope length and counterweight can be fully matched with the adaptive calculation results and selected as needed; the rigid waterproof square buoy has high imaging recognition and strong durability; the complete hardware structure supports the implementation of the whole set of intelligent deployment process, realizing the integration and synergistic effect of method and device. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the overall process of a method for setting up image control points for unmanned aerial vehicle (UAV) photogrammetry in large water areas according to Embodiment 1 of the present invention. Figure 2 This is a top view schematic diagram of a large-area UAV photogrammetry image control point deployment device according to Embodiment 2 of the present invention; Figure 3 This is a left-side schematic diagram of a large-area UAV photogrammetry image control point deployment device according to Embodiment 2 of the present invention; Figure 4 This is a bottom-view schematic diagram of a large-area drone photogrammetry image control point deployment device according to Embodiment 1 of the present invention.

[0033] In the picture: 1. Floating platform; 2. Splitter; 21. Upper fixing rod; 22. Lower fixing rod; 23. Angled connecting rod; 3. Branch wire rope; 4. Heavy hammer. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] Example 1: This invention provides a method for setting up image control points (control points) for UAV photogrammetry in large bodies of water. The method is based on a device for setting up control points for UAV photogrammetry in large bodies of water. To facilitate understanding of the method, this embodiment first provides a preliminary analysis of the device. (See reference...) Figures 2-4 This invention relates to a large-area UAV photogrammetry image control point deployment device, comprising a float 1, a line splitter 2, branch steel wire ropes 3, and a counterweight 4. The line splitter 2 includes an upper fixed rod 21, a lower fixed rod 22, and an oblique connecting rod 23. The float 1 is a square float with a pre-reserved standard bolt connection position in the center as an aerial photography identification reference. The oblique connecting rod 23 is triangular (a triangle is chosen in this embodiment, but the number of sides can be increased appropriately as needed) and is firmly connected to the float 1 through the upper fixed rod 21 and the lower fixed rod 22. The upper ends of multiple branch steel wire ropes 3 are uniformly connected to the oblique connecting rod 24, and each branch steel wire rope 3 is individually equipped with a counterweight 4 at its lower end. The entire structure relies on the line splitter 2 to achieve uniform outward pulling of the branch steel wire ropes 3 at 120°, forming a multi-directional balanced anchoring structure with the underwater counterweight 4. This structure can resist water flow and wind disturbances, ensuring the float 1 floats horizontally and stably, and improving the accuracy of aerial target identification.

[0036] The method for setting up image control points for large-area unmanned aerial vehicle (UAV) photogrammetry in this invention is based on the above-mentioned device, see reference. Figure 1 Specifically, it includes the following steps: S1. Collection of environmental parameters of the deployment area: Collect environmental parameters of each preset point in the water area to be deployed. The environmental parameters include water depth, surface water flow velocity, and wind speed above the water surface.

[0037] The specific process of step S1 includes: ①Survey area segmentation and pre-set point planning In advance, large water areas such as reservoirs, lakes, and wide rivers are divided into regions using tools such as water topographic maps. Combined with the preliminary grid layout, all preset planar points for the deployment of image control devices are marked. Preset collection points are densely deployed in offshore deep water areas, areas with sudden changes in flow velocity, nearshore aquatic plant texture areas, and river bends to ensure that environmental parameter sampling can cover all changes in water depth, flow field, and wind force within the survey area.

[0038] ② Arrival at the sampling site and preparation for sampling After the working vessel travels to the center of a single preset point, it stops and anchors to prevent the vessel from drifting and interfering with the measurement. Three types of data acquisition equipment are prepared: depth sounder, portable current meter, and handheld anemometer. The equipment is calibrated and zeroed.

[0039] ③Water depth sampling Lower the depth sounder probe vertically to the bottom of the water and read the vertical distance from the water surface to the bottom as the water depth at that point. Three sets of data were collected consecutively, and the average value was recorded. Different bottom sediments, such as silt, sand, gravel, or rock, were distinguished and simultaneously marked in the corresponding point parameter log. The sediment type was used for the subsequent safety factor of the jackhammer. Select.

[0040] ④ Surface water flow velocity collection Place the portable current meter probe 0.5m below the water surface (adjust the depth as needed) and allow it to stabilize before reading the surface water velocity. The measurement is taken three times at 10-second intervals (the specific time can be adjusted according to actual needs), and the average value is recorded to accurately reflect the impact intensity of the water flow at the point.

[0041] ⑤ Wind speed collection above the water surface Raise the handheld anemometer to a height of 1.5m above the water surface (adjust the exact height as needed), avoiding obstructions from the ship or the shore, and measure the real-time wind speed above the water surface. After multiple measurements to remove extreme values ​​of instantaneous gusts, the stable average value is recorded to characterize the magnitude of the lateral thrust of the wind on the floating plate.

[0042] ⑥ Parameter archiving and retention The location number, planar position, and water depth of each preset point are recorded. Surface water velocity Wind speed on water surface The types of underwater sediments are matched one by one to form a complete table of environmental parameters for each location, which serves as the sole input for adaptive calculation of S2 deployment parameters.

[0043] Step S1 comprehensively collects hydrodynamic, wind, and bottom friction conditions that affect buoy stability. It abandons the traditional extensive mode of simply relying on experience to uniformly configure rope length and counterweight, and realizes differentiated and accurate calculation of anchoring parameters at each point. This solves problems such as buoy drift, counterweight waste, and anchoring failure under different water conditions from the source.

[0044] S2. Adaptive Calculation of Deployment Parameters: Anchoring parameters are adaptively calculated based on environmental parameters at each location, including the length of the branch wire rope and the weight of the matching counterweight. A corresponding safety factor is selected based on the type of underwater substrate. Complete the quantitative calculation of the weight of the counterweight.

[0045] The specific process of step S2 includes: ①Preparation of data and parameter preprocessing First, extract the complete environmental dataset for a single location: the average water depth at that location. Average surface velocity from three velocity measurements Stable wind speed after removing extreme gusts 1. Classify the riverbed substrate (silty riverbed, sandy riverbed, gravelly rock riverbed); then determine the fixed design parameters of the device: refer to Figures 2-4 This device limits the angle between the branch wire rope 3 and the horizontal plane to a range of 45°~60° (the specific angle may be adjusted according to the actual situation). The angle between the wire rope and the plumb line is obtained through geometric calculation. Finally, input the inherent parameters of the device hardware: diameter of branch steel wire rope 3, projected area of ​​spherical cast iron hammer 4 facing the water, and external dimensions of float plate 1, for subsequent calculation of water blocking and windward area.

[0046] ② Precise calculation and selection of branch wire rope length Due to water depth The vertical height from the underwater counterweight to the float is [height]. Let be the cosine of the angle between the vertical direction and the wire rope. Dividing the two yields the theoretical wire rope length required to satisfy the preset tension angle. The greater the water depth, the required wire rope length increases linearly. Using the formula... Calculate the theoretical length of a single branch wire rope .

[0047] Installation allowance compensation: within the theoretically calculated length An additional 0.3~0.5m margin is added on top of the existing margin (this margin is an empirical value under long-term use, and the specific value should be adjusted according to the actual situation) to offset slight bottom undulations and hull deployment errors, and to prevent the splitter from being forcibly pulled up by the water flow and the float from tilting due to insufficient rope length; at the same time, the margin should be strictly controlled to avoid the steel wire rope being too long and slack, which would cause the float to swing in a wide range and the target imaging to deviate.

[0048] Steel wire rope specification selection: According to the final determined total length, cut 3 branches of 304 stainless steel (or other materials) of the corresponding length. The three branches of steel wire rope at the same point are of the same length to ensure that the tension of the three lines is balanced and to avoid the floating plate tilting due to the rope being too short on one side.

[0049] ③ Calculation of underwater equivalent water-blocking area and floating plate equivalent windward area Using the formula: Calculate the equivalent underwater water-blocking area In the formula The diameter of the branch wire rope, This is the total length of the wire rope calculated for this location. The projected area of ​​a single weight facing the water is 4. The larger the value of the underwater equivalent water-blocking area, the stronger the impact force of the water flow, and the larger the required counterweight.

[0050] Equivalent windward area of ​​floating plate The value selection rule is as follows: Float 1 is a square high-density polyethylene float (in this embodiment, the size of float 1 is set to...). Wind blows towards the square floating plate from various angles. The wind load is quantified by taking the average equivalent windward area, simplifying the on-site calculation process. One-third of the total surface area of ​​the floating plate (the actual equivalent area can be adjusted according to actual needs) is uniformly taken as the equivalent windward area. ; ④ Quantitative solution for the minimum weight of the hammer based on the safety factor of the sediment classification First, determine the safety factor based on the type of bottom sediment. The value range is as follows: In silty riverbeds, the coefficient of friction of the silt at the bottom is extremely low, and the weight is very prone to slippage. The safety factor is... A value of 2.5~4.0 is suitable; the riverbed is sandy with moderate sediment friction. Use a depth of 1.5~2.5; for gravelly or rocky riverbeds, the base is rough with high friction, making anchoring less prone to slippage. Use a value between 1.0 and 1.5; operators should directly select values ​​within this range based on the bottom sediment markings on site. The deeper the water and the greater the flow and wind speed, the better the value will be for the same bottom sediment. Take the upper limit of the range to reserve a higher safety margin.

[0051] Then, the weight constraint formula of the counterweight is used. Calculate the required weight of the counterweight 4 at different locations. The underwater equivalent water-blocking area of ​​the branch wire rope and counterweight. The equivalent windward area of ​​the floating platform. The surface water velocity, For wind speed, The weight of the hammer. For safety factor; Represents the lateral thrust load generated by the water flow. The lateral thrust load generated by wind is represented by the sum of the two loads multiplied by the bottom mass safety factor. To obtain the minimum total counterweight weight required to prevent the buoy from drifting or flipping. Flow rate Wind speed The larger the area, the greater the water-blocking or windward area, the more precise the calculated minimum counterweight. Increase synchronously.

[0052] This embodiment uses three equally weighted spherical cast iron hammers 4. The specifications of each hammer are fixed. The total weight of the three hammers is compared with the lower limit of the calculation. In comparison, select those with a total weight greater than or equal to The counterweight combination; the spherical cast iron hammer 4 includes a standard weight hammer with a screw, increasing in increments of 1 kg to ensure that the anchoring counterweight meets the requirements for resistance to current and wind.

[0053] ⑤ Archiving of Point Anchoring Parameter Ledger The complete calculation results for each point are recorded in a unified ledger, which includes: point number, measured water depth. Calculate the rope length Water-blocking area of ​​a single steel wire rope Equivalent windward area Substrate type and selection Numerical values, calculation of minimum counterweight The actual total weight of the selected weights; the ledger directly serves as the selection basis for the initial deployment operation in step S4, enabling independent configuration at each location without the need for standardized components.

[0054] Step S2 completely abandons the traditional, crude method of relying on operators' experience to estimate rope length and counterweight. It achieves quantitative adaptive calculation of anchoring parameters through a combination of mathematical formulas based on multi-dimensional environmental parameters, adapting to the complex spatial changes in water depth, flow velocity, wind force, and riverbed sediment within the survey area. By matching the length of the steel wire rope with the water depth geometry, it stably maintains a 45°~60° tension angle of the branch steel wire rope, ensuring that the three forces of the splitter 2 are evenly distributed, thus reducing the probability of the float 1 tilting from a structural perspective. Combining the quantitative calculation of counterweight requirements with water flow and wind loads, and with the safety factor of bottom sediment classification, the counterweight is automatically increased to prevent slippage in silty and soft bottoms, while the counterweight is appropriately reduced in rocky and hard bottoms to save cast iron materials. It takes into account both the stability of the buoy and the economy of field deployment, simultaneously solving the core pain points of buoy drifting, overturning, and blurred imaging from both computational and hardware levels.

[0055] S3. Optimization of Deployment Density and Aerial Survey Parameters: Based on the UAV flight hardware parameters, preset operating altitude, heading, and side image overlap, the ground sampling distance (GSD) and the effective area coverage width of a single image are first calculated using imaging geometry formulas. Then, a texture sparsity coefficient is introduced, combining the sparse texture characteristics of water surfaces in different areas within the survey area. A single image must cover at least a certain number of pixel counts. The optimal spacing of image control points in the measurement area blocks was calculated simultaneously. Finally, based on the optimal spacing of all points, a standardized grid layout scheme covering the entire large water area is generated, and the planar deployment coordinates of all image control point devices are divided.

[0056] The specific process of step S3 includes: ① Basic aerial survey parameter input and classification Before the operation, all hardware and flight design parameters for this aerial survey should be collected as underlying calculation constants. The parameter list includes: relative flight altitude of drones The vertical flight altitude of the drone above the water surface is preset by the mapping scale standard. Camera hardware parameters: Lens focal length Sensor single pixel physical size ; Image overlap design parameters: forward overlap Lateral overlap The unit is percentage; Image pixel dimension : Total number of pixels in the heading and lateral directions of a single aerial photograph; Meanwhile, based on the previous water topographic map, the large water area was divided into multiple sub-blocks, and the water surface texture type of each sub-block was individually labeled, which were divided into four categories: open still water (reservoir center), slightly rippling water area, nearshore aquatic plant area, and flowing waterway. Different blocks correspond to a unique texture sparsity coefficient. Minimum number of pixel counts per image .

[0057] ② Calculate the ground sampling distance GSD Using imaging geometry formulas The ground sampling distance is calculated using a formula based on the pinhole imaging model. Higher flight altitude, larger pixel size, and shorter lens focal length result in a larger GSD value, meaning a larger actual ground size corresponding to a single pixel, resulting in a smaller image of the float target and increased recognition difficulty. Conversely, a smaller GSD results in clearer target imaging with sufficient detail. In practice, preset flight altitude, camera pixel size, and focal length parameters are substituted to obtain a unified GSD value for this flight, serving as the core benchmark parameter for subsequent deployment distance calculations. GSD directly determines the pixel size of the float in the aerial image and is the core basis for judging the density of deployment points. In operational conditions with a larger GSD, the deployment spacing must be reduced to ensure the target has sufficient imaging pixels.

[0058] ③ Calculate the effective area coverage width of a single image. Solve for the effective coverage width in the heading direction and the effective coverage width in the lateral direction respectively. The calculation formula is: ,

[0059] In the formula The image pixel dimension is denoted as , and the forward and lateral overlap percentages are denoted as . , These are the heading and lateral overlap, respectively. This represents the proportion of the ground newly covered by a single photo after removing overlapping areas. Using this formula, we can obtain the effective ground coverage area of ​​a single aerial image that does not overlap with adjacent photos, thereby determining how much water surface a single photo can cover and providing geometric boundary constraints for determining grid spacing. The higher the overlap, the smaller the effective coverage width, and more image control points need to be deployed for the same area of ​​water. The program automatically links to reduce the optimal deployment spacing.

[0060] ④ Calculate the optimal spacing of a single block based on water surface texture characteristics.

[0061] The formula for calculating the optimal layout spacing is: In the formula This represents the sparsity coefficient of the water texture. Ground sampling distance, This represents the minimum equivalent number of pixels required to cover control points within a single image. During the calculation, the corresponding value is first selected by looking up a table based on the water surface texture type marked in the sub-block; specifically, if the object is open, still water (the center of a lake or reservoir): very little texture, and a large area of ​​uniform water surface. Take 4~6, The value is set to 0.40~0.50. The water surface in this area lacks natural texture, and only buoys can be used as matching features. Therefore, each image is required to contain more control points and the spacing between them should be minimized. If the object is a gently rippling water area: there is a small amount of ripple texture. Take 2~4, Take a value of 0.50~0.70; If the object is near the shore or in an area with aquatic plants: the aquatic plants and shallow water textures are rich, and there are sufficient automatically matched feature points. Take 1~2, Using a value of 0.70~1.00 allows for larger spacing between installation points and fewer installation locations. If the object is a flowing river: there is a continuous water flow texture. Take 2~3, Take a value of 0.55 to 0.70.

[0062] Therefore, the sparser the texture, the better. The smaller the value, the better the calculated optimal spacing. The smaller the size, the denser the placement of the points.

[0063] Confirm as well as After selecting the values, substitute them into the parameters to solve for the block-specific values. Specifically, the sparsity coefficient of the block texture GSD, minimum equivalent pixel count Substitute into the formula The unique optimal spacing for this sub-block is calculated. Different texture zones within the same body of water will generate multiple sets of different spacing values, enabling differentiated point placement in different areas and abandoning the traditional method of uniform grid spacing across the entire body of water.

[0064] ⑤ Generate a global gridded deployment plan and coordinates of deployment points. Optimal spacing between each sub-block To determine the grid side length, uniform square grids are generated for each sub-block of the water area. All intersections of the grids represent the preset planar deployment points for the image control (ADC) device, and the planar coordinates of each point are output simultaneously. The grid generation process considers the water area boundaries: the outermost grid of the survey area is no more than 0.5 times the optimal spacing from the shoreline to avoid missing control points at the water edge and to prevent excessive aerial triangulation errors at the survey area edges. All sub-block grid points are merged, and a complete list of points is output, with each point in the list simultaneously labeled with its respective block and corresponding optimal deployment spacing. This serves as the basis for the initial planar position of the deployment in step S4.

[0065] Step S3 establishes a quantitative mathematical correlation between UAV flight hardware parameters, flight plans, water surface texture features, and image control deployment spacing. This breaks away from the traditional, rough-and-ready approach of manually setting grid spacing based on experience, providing a unified and reproducible calculation standard for point density. Based on GSD imaging resolution, the point density is adjusted, with automatic densification of points in high-altitude, high-GSD conditions to ensure sufficient recognition pixels for buoys within the image and prevent broken matching of corresponding points due to small targets. Spacing is adjusted according to the richness of water surface texture; the placement distance is appropriately widened in areas with abundant texture, such as aquatic plants and river channels, significantly reducing the investment in buoys, boats, and manpower, thus lowering field costs. In open, still water areas lacking texture, points are automatically densified, strengthening the geometric constraints of the aerial triangulation network and improving the accuracy of horizontal and vertical measurements in the central water area. Standardized grid point coordinates are output, unifying the distribution of control points across the entire water area and preventing arbitrary manual placement that could lead to missing or accumulated points. This provides a complete and orderly point foundation for subsequent environmental parameter collection, buoy deployment, and test flight quality assessment.

[0066] S4. Initial Deployment: Based on the coordinates of the global grid plane points output in step S3, and combined with the exclusive anchoring parameters (length of branch steel wire rope 3, total weight of matching weight 4) obtained by adaptive calculation for each point in step S2, the water surface image control point device is pre-assembled on the work vessel. Then, the vessel sails to each preset grid point in sequence and smoothly deploys the assembled buoy device onto the water surface, completing the initial deployment of all preset points in the entire survey area. After deployment, wait for the weight to sink completely to the bottom and the float to float stably and horizontally before sailing away to the next point, achieving a standardized and adaptable one-time initial deployment of all points.

[0067] The specific process of step S4 includes: ① Pre-assembly of the device at designated locations before deployment The work vessels docked at the shore, and the corresponding specifications of the CVR devices were assembled in batches according to the location log in advance, ensuring that each location had its own dedicated configuration: a. Assembly of Splitter 2: Take the stainless steel splitter body and align and fix the upper fixing rod 21 to the geometric center of the float 1; bolt the upper end of the lower fixing rod 22 to the lower end of the upper fixing rod 21; thread the oblique connecting rod 23 to the lower end of the lower fixing rod 22 to complete the rigid connection between the splitter and the float 1. The float 1 is an 80×80×20cm high-density polyethylene square float (the specific size and material are adjusted according to actual needs), with the red and white diagonal high-strength reflective paint markings on the surface facing upwards, and the upper fixing rod 21 (in this embodiment, an M16 stainless steel nut) serves as the reference for subsequent RTK measurement points.

[0068] b. Matching and installation of branch wire rope 3: Take the total length of the wire rope calculated at point S2, and cut 3 304 stainless steel branch wire ropes 3 of the same length; fix the upper end of the wire rope to the three branch holes of the inclined connecting rod 23, and let the three ropes spread out naturally downwards, with the included angle between adjacent wire ropes maintaining 120°, which meets the design requirement of 45°~60° water surface traction angle.

[0069] c. Optional installation of cast iron counterweight 4: Based on the minimum total weight of the counterweight calculated by S2, select galvanized spherical cast iron counterweight 4 of the corresponding specifications, and tie one counterweight to the lower end of each branch wire rope; the three counterweights are exactly the same in specifications and weight to ensure that the tension in the three directions underwater is equal.

[0070] d. Assembly and verification: After assembly, manually lift the float 1 to observe that the three steel wire ropes are evenly stressed, there is no one-sided tension or slack, and the float is not obviously tilted. After confirming that the assembly is fault-free, stack them according to the point number to avoid mixing rope lengths and counterweights at different points.

[0071] ② Vessel navigation and positioning and pre-deployment water area verification a. Based on the planar coordinates of the grid points generated by S3, guide the working vessel to the center of a single preset point using navigation equipment; the vessel drops a temporary anchor to fix the hull and prevent the hull from shifting due to water flow and wind, thus avoiding horizontal deviation of the deployment point.

[0072] b. On-site simple verification: Quickly verify the water depth, flow velocity, and bottom sediment information collected at point S1 to confirm that the current water conditions do not change drastically from the previously collected data; if there are significant changes in water depth and flow velocity on site, temporarily record the abnormality at this point, and recalculate the anchoring parameters for a second adjustment after all initial deployments are completed.

[0073] ③ Underwater deployment operation of control point device a. Deployment Operation Procedures: Two operators should work together. One person should hold the float 1 to keep it level, while the other person should simultaneously and slowly lower the three branch steel wire ropes 3 that connect to the weight 4. It is forbidden to throw the entire device directly from a height to prevent the steel wire ropes from getting tangled, the weight from being deformed by impact, and the reflective markings on the float surface from being scratched or falling off.

[0074] b. Wait for the bottom to stabilize: Slowly lower the heavy hammer until it fully contacts the riverbed and let it stand still for 2-5 minutes; observe the floating status of the float 1, and use the slight rocking of the hull to simulate the impact of the water flow to confirm that the float does not drift continuously, does not tilt or overturn on one side, and that the three branch steel wire ropes remain taut and do not slack or accumulate.

[0075] c. Criteria for successful deployment: Float 1 is fully exposed above the water surface, the red and white diagonal reflective markings are completely facing upwards and there is no large area of ​​side overturning or obstruction; all three branch steel cables are taut, the weight is in contact with the bottom of the water without any signs of slippage; the reference point of the cross spike in the center of the float is not offset, meeting the conditions for aerial imaging and recognition.

[0076] d. Site marking: After successful deployment, record the deployment completion in the site log and simultaneously take photos of the actual buoy on the water surface for future reference as a basis for comparison of test flight images.

[0077] ④ Global point location cyclic deployment After completing the deployment at the current location and confirming its stability, the temporary anchor is retrieved, and the navigation guides the vessel to the next grid's preset location. The entire process of assembly, verification, placement, deployment, and stability verification is repeated from ① to ③. Deployment is carried out row by row and column by column in strict accordance with the grid zoning order to avoid omissions or duplicate deployments. For deep water areas offshore, river bends, and areas with sudden changes in flow velocity, the deployment pace is slowed down to extend the stabilization time of the weight.

[0078] ⑤ Initial deployment completed, overall inspection After all grid points have been deployed, the vessel will conduct a full inspection by circling the grid loop: checking the floating attitude of each buoy, identifying abnormal points with obvious tilting, severely slack steel cables, or large-scale buoy drift, and marking and recording them separately.

[0079] This step fully matches the S2 adaptive anchoring calculation results, with each point independently configured with its own rope length and counterweight. Unlike traditional all-waterway buoys of uniform specifications, this step ensures buoy stability under different water depths, current velocities, and wind conditions from a hardware assembly perspective. It employs a splitter-type 2-to-3-line 120° uniform traction structure, simultaneously counteracting the lateral thrust of water flow and wind in three directions. Compared to single-rope buoys, this significantly reduces the probability of drifting and capsizing, ensuring the reflective markings on the float always face the aerial camera, improving target image recognition rates. Standardized pre-assembly and stable deployment procedures reduce device damage and point displacement caused by human error. Post-deployment stabilization verification preemptively eliminates anchoring failure points, reducing unnecessary workload in subsequent RTK measurement points and flight test evaluation stages. Simultaneously, it strictly relies on the grid coordinates generated by the S3 aerial survey linkage optimization for precise deployment, ensuring uniform distribution of control points that closely match water surface texture and flight parameters, providing a reasonably distributed, stable, and reliable all-area ground control benchmark for subsequent aerial triangulation calculations.

[0080] S5. RTK Coordinate Acquisition: After all image control point devices have been deployed in step S4 and the float 1 is floating stably on the water surface without continuous drifting, the RTK (Continuously Operating Reference Stations) real-time dynamic differential measurement equipment equipped with network CORS service is used to sequentially reach each deployed buoy point and collect centimeter-level three-dimensional geodetic coordinates (X0, Y0, Z0) at the preset stainless steel spike reference point at the center of the float 1; all image control point three-dimensional coordinate data are stored, numbered and archived point by point.

[0081] The specific process of step S5 includes: ①RTK equipment pre-calibration and parameter configuration Equipment selection and power-on verification: Select a network RTK rover that supports BeiDou + GPS dual-mode; after powering on, connect to the local CORS continuously operating base station network, enter the corresponding account and password to complete the differential link connection, and wait for the equipment to output the fixed solution status. The fixed solution means that the positioning accuracy can reach the centimeter level, which meets the requirements of the aerial survey image control point acquisition specifications. Floating point solutions and single point solutions are not allowed to be used for measurement point operations.

[0082] Set the RTK output coordinate system to the geodetic coordinate system (such as the 2000 National Geodetic Coordinate System) used in this aerial survey, and match the elevation datum to the standard elevation system of the survey area to avoid systematic errors in the later aerial triangulation adjustment caused by inconsistent coordinate datums.

[0083] ② Verification of buoy status before vessel positioning and measurement point setup According to the grid point numbering sequence, the navigation guides the work vessel to the target buoy, slowly stops the vessel, and drops a temporary anchor to secure the hull and eliminate the interference of hull swaying on the RTK data acquisition accuracy; the overall condition of the buoy is visually checked at close range: observe whether the float 1 is level, whether all three branch steel cables 3 are fully tensioned, and whether there is any slippage of the single-sided weight 4 causing the float to tilt; if the buoy has obvious drifting, capsizing, steel cable slack or other abnormalities, mark the point as an abnormal point in the log and conduct key evaluation in the subsequent test flight phase.

[0084] ③Accurate coordinate acquisition of the center reference point of the floating board The bottom circular base of the RTK centering rod is stably attached to the reference surface of the upper fixed rod 21 at the geometric center of the float 1 to ensure that the centering rod is perpendicular to the surface of the float and to avoid the measurement point eccentricity error caused by the tilt of the rod. The float 1 is an 80cm square rigid float with the top surface of the center nut flush with the surface of the plate. It is the only standard puncture point position. It is forbidden to collect coordinates on the edge of the float or in the reflective marking area.

[0085] After the centering rod is installed, keep it stationary and continuously observe the RTK receiver status, maintaining a fixed solution state for no less than 10 seconds; continuously collect 3 sets of coordinate readings. If the difference between the three sets of plane coordinates is less than 2cm and the difference between the three sets of elevations is less than 3cm (the specific difference range is adjusted according to actual needs), the collected data is considered valid; if the readings fluctuate too much, it indicates that the buoy is drifting slightly, extend the stationary waiting time, and collect data again after the buoy stabilizes.

[0086] After data acquisition, the average of the three sets of valid coordinates is taken as the final three-dimensional coordinates (X0, Y0, Z0) of the control point. The coordinate data is bound to the unique point number and stored in the memory of the RTK device. At the same time, the coordinate values ​​are manually entered into the tablet log, and panoramic photos of the buoy are taken simultaneously. The photos are labeled with the corresponding point number to establish the correspondence between the image and the coordinate data.

[0087] ④ Loop data collection and preliminary self-check of data across the entire survey area After the single-point data collection is completed, the RTK equipment is put away, the temporary anchor is taken down, and the vessel sails to the next buoy point. The entire process of vessel positioning, buoy verification, RTK multi-point observation, and data storage is repeated. Measurements are taken point by point in strict accordance with the grid partitioning order to avoid missed or repeated measurements.

[0088] After the full-area data collection is completed, quickly export the coordinate table of all points on-site and conduct a basic self-check. Verify that the total number of points is consistent with the number of points in the S3 grid layout, and check for any missing coordinates or incorrect numbering. Extract the coordinates of previously marked tilted or drifting points separately and add notes explaining the anomalies.

[0089] ⑤ Coordinate data archiving and backup All 3D coordinates of all ground control points, point numbers, on-site photos of buoys, and remarks on abnormal points are integrated and exported to form a complete ground control point coordinate result file. This result is directly used as the core ground control data for indoor aerial triangulation calculation, which is used to constrain the position and attitude of UAV aerial images and eliminate the problems of regional network drift, layering, and splicing misalignment caused by relying solely on airborne POS (Position and Orientation System) data.

[0090] S6. Test Flight Image Acquisition and Quality Assessment: After completing the deployment of control points and RTK coordinate acquisition across the entire area, instead of directly conducting full-coverage aerial photography, a small-scale fixed-point test flight verification process is first implemented. The drone is operated to hover and capture vertical downward images at the center of each buoy 1 and the center of each grid block. The images are viewed in real time on-site using equipment. Standardized grading is performed based on two dimensions: the recognizability of buoy markings and the number of control points visible in the grid center image. The test area is divided into qualified areas, undetermined areas, and unqualified areas and marked for archiving. Imaging defects caused by water surface reflection, ripple obstruction, and buoy drift and tilt are identified in advance.

[0091] The specific process of step S6 includes: ① Equipment and flight parameter configuration before test flight Drone parameter reuse: The flight altitude used in the official aerial photography was completely adopted during the test flight phase. Camera focal length Imaging resolution and shutter parameters are used to ensure that the GSD of the test flight images is completely consistent with that of the official flight, and the evaluation results have real reference value. On-site inspection of equipment preparation: A waterproof tablet equipped with a high-definition display receives images transmitted from the drone in real time via wireless image transmission, allowing for on-site interpretation without waiting for the images to be exported; simultaneously, the location ledger and water area grid distribution map are opened for real-time marking of vulnerable areas.

[0092] ② Two types of fixed-point images need to be captured in batches, covering two major evaluation dimensions: the imaging of a single control point itself and the overall distribution of control points in a block. a) Vertical image capture directly above a single control point The drone is controlled to ascend vertically to the preset altitude and hover precisely above each floating board 1. The lens is then used to capture a single high-definition image vertically downwards. During the shooting, the floating board is ensured to be completely centered in the frame, with no large areas of water surface reflection or waves obscuring the main body of the floating board.

[0093] b. Vertical image capture of the center of the grid block Using the grid generated by S3 as a unit, fly to the geometric center of each grid square and hover to capture vertical images. These images are used to count the number of buoys that can be captured in a single frame, and to intuitively reflect whether the density of the buoys meets the standard.

[0094] ③ First Dimension Judgment: Float Board Marking Identifiability Grading Assessment. Retrieve images taken directly above each float board and observe the clarity of the red and white diagonal reflective markings and the central crosshair reference point on the surface of float board 1. The criteria are divided into three levels: Qualified: The float is fully exposed above the water surface, the red and white reflective areas have clear contrast, the central crosshair outline is complete, there is no strong mirror reflection, no waves completely cover the target, and the software can automatically identify the central spike. Pending (needs attention): There are local water surface reflections and fine ripples that obscure the edges of the buoy, but the central cross reference is still visible to the naked eye and can be used for manual data entry. Unqualified (weak area): Large area of ​​reflection from strong midday sunlight, large waves completely covering the surface of the floating board, cross markings blurred and disappeared, the location of the spikes cannot be located in the image, and they cannot be used as aerial triangulation constraint control points.

[0095] Simultaneously mark the corresponding points on the grid map and classify them as weak areas to be encrypted.

[0096] ④ Second Dimension Interpretation: Grading Determination of the Number of Visible Control Points in the Grid Center Image. Open the vertical image captured at the grid center, manually count the number of complete and identifiable floating platforms within the image, and classify them into three levels according to the number: Acceptable: The number of visible control points in the image is ≥4, the number of control points in a single image is sufficient, the constraint strength of the aerial triangulation network is sufficient, and there is no local drift. Pending (Requires Attention): The number of visible control points in the image is 2-3, which is too few. There is a risk of misalignment during stitching in the later stage. This area is listed as a key area for review. Unqualified (weak area): The number of visible control points in the image is only 0 to 1, the spacing between the control points is too large, and the single image lacks effective surface constraints, which can easily cause aerial triangulation matching to break and image layer misalignment. It is necessary to add more control points.

[0097] ⑤ Global regional hierarchical archiving markers The area level is determined by combining the results of two assessments: if either assessment is unqualified, the entire grid block is classified as an unqualified weak area; if both are pending, it is classified as an area requiring attention and observation; if both are qualified, it is classified as a qualified area. Different colors are used to distinguish three types of areas on the electronic water area grid map, and the defect types of each weak area are recorded simultaneously: target reflection obstruction, sparse buoy deployment, and buoy drift / tilt. A complete assessment log is compiled, including the block number, defect type, weak level, and the basis for the corresponding reinforcement coefficient value, which directly serves as the operational guidance document for targeted supplementary installations in S7.

[0098] ⑥ Backup of original test flight footage All images captured during fixed-point test flights are categorized and stored according to location and grid number; the images correspond one-to-one with the graded evaluation results, and the imaging effects can be compared a second time after subsequent supplementary setup, achieving full traceability.

[0099] This innovative pre-flight test and verification process differs from the traditional one-time deployment followed by direct aerial photography. It exposes issues such as water surface reflection, waves, sparse deployment, and buoy anchoring failure before the official flight, avoiding the need for post-construction defective work requiring secondary vessel rentals and on-site installation, significantly saving on vessel, labor, and time costs. A dual quantitative evaluation standard of "target clarity + control point coverage" is established, eliminating subjective human judgment, unifying the overall assessment scale, and accurately distinguishing between areas with minor and severe defects, providing an objective basis for differentiated grading and densification. The same imaging parameters are reused from the official flight, ensuring that the test flight image resolution is completely consistent with the actual results, guaranteeing reliable evaluation results and preventing discrepancies between successful test flights and ineffective official aerial photography. Simultaneously, weak imaging areas are identified in advance, achieving "precise location of problem areas." This eliminates the need for uniform densification across the entire area, allowing for targeted addition of points, minimizing buoy material costs, and balancing surveying accuracy with deployment economy.

[0100] S7. Targeted Supplementation: Based on the areas to be observed and the unqualified weak areas identified in the image quality assessment in step S6, match the corresponding densification coefficient according to the severity of defects in each area. The calculated encryption and supplementary spacing Based on the original grid, new image control point deployment points were added. The S2 anchoring parameter adaptive calculation process was executed again to match the new points with branch steel wire ropes of specific lengths 3 and cast iron hammers of appropriate weights 4. After the device was assembled, it was transported by boat to the new points to deploy buoys. After the deployment was stable, the RTK was used to remeasure the three-dimensional geodetic coordinates of the center of the new buoy 1. All parameters and coordinate data of the new points were fully recorded to complete the supplementation of control points in weak areas, reduce the spacing between points in defective areas, increase the number of identifiable control points in the image, and solve the aerial triangulation defects caused by unclear targets and too few control points in a single image.

[0101] The specific process of step S7 includes: ① Matching of weak area classification and encryption coefficient Retrieve the S6 assessment log and classify the blocks into three vulnerability levels based on the severity of their defects, corresponding to different encryption coefficients. : Severely weak areas: image stitching failure, large areas of voids exist. Use a value of 0.3~0.4; Moderately weak areas: Significant misalignment exists in the image stitching. Use a value of 0.4~0.6; Slightly weak areas: Image stitching shows only slight misalignment. Set the value to 0.6~0.7; encryption coefficient The value is always less than 1. The smaller the encryption coefficient, the smaller the final spacing and the denser the point layout.

[0102] Substitute into the encryption spacing calculation formula In the formula To determine the original optimal grid spacing for this block, the standard grid spacing after refining the weak block is calculated. The subdivided mesh is regenerated based on the densified spacing. The original buoys are retained at the intersections of the original mesh, and the newly added intersections of the subdivided mesh are the points to be added this time. The planar coordinates of all the newly added points are output to form a list of the added points, and the weak level and corresponding densification coefficient of the points are marked.

[0103] ② Verification of environmental parameters and recalculation of anchoring parameters for newly added locations All the additional measurement points were located within the original survey area, and the measured water depths from the surrounding area during the S1 phase were retrieved from nearby locations. Surface water velocity Wind speed on water surface 1. Bottom sediment data; if the new sampling point is far from the original sampling point or the water conditions have changed significantly, the vessel will be driven to the new sampling point to re-measure all environmental parameters to ensure that the calculation input data is true and accurate.

[0104] The entire adaptive calculation logic of step S2 is reused to calculate the theoretical length of the corresponding branch wire rope and reserve the installation margin, and to calculate the matching weight of the cast iron hammer 4; the rope length and counterweight specifications of each newly added point are entered into the supplementary ledger to realize the independent matching and adaptation of the anchoring components for each supplementary point.

[0105] ③ Pre-assembly and on-site deployment of additional buoy devices On the shore, the image control point devices are assembled in batches according to the assigned point numbers. Branch steel wire ropes of corresponding length are installed, and cast iron counterweights of matching weight are attached. After assembly, the tension of the steel wire ropes is checked to ensure uniformity and that the buoy is not tilted. The vessel is then guided to the target location using the planar coordinates of the assigned point. The vessel is anchored and the entire buoy device is smoothly lowered. It is allowed to stand for 2-5 minutes until the counterweight 4 has completely sunk to the bottom and the buoy is level and stable. After visually inspecting for drift, capsizing, or slack in the steel wire ropes, the point deployment is recorded as complete. This process is repeated until all new points in the vulnerable area are deployed, and photos of the buoys are taken on-site for record-keeping.

[0106] ④ Added buoy RTK remeasurement coordinates The S5 network CORS RTK equipment was reused, and its stable output of centimeter-level fixed solutions was confirmed. The vessel was moored next to the newly added buoy, and temporary anchoring was used to eliminate interference from hull swaying. The RTK centering rod was vertically attached to the stainless steel reference nut at the center of buoy 1, and three sets of coordinate data were collected under static conditions. The average value was taken as the three-dimensional coordinates (X0, Y0, Z0) of the newly added control point, and the coordinates were bound to the supplementary point number for storage. The coordinates of the newly added point, the floating status of the buoy, and on-site photos were simultaneously entered into the supplementary point log to distinguish between the original points and the later-added points, thus establishing a complete global control point database.

[0107] ⑤ Summary of Supplementary Results Archive and integrate all environmental parameters, anchoring configurations, RTK coordinates, and field images of the original grid points and newly added points to generate a unified and complete global image control point result table, clearly marking the point distribution of qualified areas, areas to be observed, and severely weak areas, as the basic data for step S8 iterative review.

[0108] This step involves differentiated encryption based on the severity of image defects, rather than uniform global encryption. Encryption is only applied to areas with substandard imaging, reducing the spacing between buoys to address aerial triangulation matching defects with minimal buoy increments, thus saving on material, vessel, and labor costs. Newly added buoys also undergo environmentally adaptive anchoring calculations to ensure they adapt to local water depth, current, and wind conditions, preventing drifting or flipping and avoiding image failures after installation. All added buoys are simultaneously remeasured with RTK centimeter-level coordinates, and new control points are constrained by standard geodetic coordinates, effectively filling geometric constraint gaps in sparse areas of the original grid control points and eliminating image stitching misalignment and voids. All data from both original and added buoys are fully archived for regional comparison of flight test imaging effects during iterative review phases, ensuring data traceability throughout the deployment and encryption process. S8. Iterative Verification: The entire water area after the hierarchical encryption and supplementation in step S7 is used as a new verification object. The entire process of step S6, test flight image acquisition and quality assessment, and step S7, targeted supplementation is repeated in a loop. The imaging effect of the grid blocks in the entire test area is reviewed round by round. Weak areas with defects are continuously identified and image control points are added. Multiple iterations are carried out until all grid blocks in the entire water area are judged to be qualified after image evaluation. The entire area image control point deployment process is completed.

[0109] Step S8 includes the following process: ① Defining the scope of data integration and verification before the first iteration Integrate the original initial locations and the S7 newly added locations in the logbook, summarize all location numbers, planar coordinates, anchoring configurations, and RTK 3D coordinates, and update the global electronic grid distribution map, distinguishing between the original locations and the first round of added locations on the map; delineate the complete verification scope of this iteration of test flights, covering all qualified areas, areas to be observed, and previously unqualified weak areas, without omitting any water area, to prevent only verifying the added areas while ignoring new defects caused by buoy drift due to water flow and waves in the original qualified areas; reconfirm that the UAV flight parameters, camera parameters, and test flight shooting mode are consistent with the official flight, and ensure that the tablet, RTK, logbook, and image storage devices are all in place, and start a new round of iteration verification.

[0110] ② Iterative process of cyclical flight test evaluation and densification supplementation. Repeat step S6 for the entire test flight operation: the drone flies to the top of all buoys and hovers vertically at the center of all grids to take pictures. On-site, the dual-dimensional classification is completed through image transmission. The qualified, attention-required, and unqualified blocks of this round are re-divided, the weak area log is updated, and the causes of the new defects in this round are recorded simultaneously, such as slight drift of buoys due to long-term water flow, changes in local water surface reflectivity, and insufficient local points after densification.

[0111] If there are no unqualified or weak areas in this round of assessment, and only a few areas to be observed, and the two judgment indicators of the areas to be observed are close to the qualified standard, they can be recorded and retained without the need to be re-established. If any unqualified weak block still exists in this round, or multiple areas under observation are clustered together, it is determined that further encryption is needed, and the process proceeds to step S7: matching the corresponding encryption coefficient. Calculate the spacing between subdivided grids, remeasure the aquatic environmental parameters of newly added points, adaptively calculate the anchoring parameters, assemble and deploy buoys, and collect the coordinates of newly added points using RTK to complete a new round of point addition; After each round of supplementary additions is completed, the global point grid map, coordinate database, and on-site image archives are updated, the iteration rounds are marked, and the points added in the first, second, and multiple rounds are distinguished to facilitate the tracking of each round of optimization process.

[0112] ③ Standards for Acceptance of Standards Across the Entire Region The iteration can be considered complete and the deployment work finished only if all of the following conditions are met: All grid blocks in the entire survey area were evaluated using S6 imagery, and no unqualified or weak areas were found. The number of visible control points in the center image of all grids was ≥4, meeting the requirement of sufficient control points per image. All red and white reflective markings and central crosshairs on all floats were clearly identifiable, with no large-area reflections or targets completely obscured by waves. There were no obvious and continuous drifting, capsizing, or slack steel cables among the buoys in the entire area, and no abnormal shifts in RTK coordinates. The coordinates of the original points, the coordinates of the points added in multiple rounds, the test flight images, the environmental parameters, and the anchoring configuration records were complete and intact, with no missing data.

[0113] ④ The results of the deployment across the entire region will be archived uniformly. After the iteration is accepted, the complete set of operational data is integrated to form a complete deployment result package. All data is backed up in both electronic and paper formats, serving as complete supporting materials for subsequent formal UAV flight operations and indoor aerial triangulation calculations.

[0114] This step constructs a closed-loop iterative process of "flight test evaluation - graded re-deployment - re-flight test and re-inspection", which is different from the traditional one-time deployment without a re-inspection mechanism. It can gradually eliminate the water surface dynamic interference problems (wind and waves, reflection, slight buoy slippage) that are difficult to predict in a single deployment, and continuously optimize the distribution density of control points.

[0115] Example 2: This embodiment further describes the device of the present invention based on Embodiment 1. The large-water unmanned aerial vehicle (UAV) photogrammetry image control point deployment device disclosed in this invention comprises four main parts: a float 1, a line splitter 2, branch steel wire ropes 3, and a counterweight 4.

[0116] For details, please refer to Figure 2 The float 1 is a high-density polyethylene square rigid float with dimensions of 80×80×20cm (the specific dimensions and materials are adjusted according to actual needs). It has an integral molding structure with good buoyancy and water resistance and deformation resistance. The upper surface of the float is painted with red and white diagonal spraying (the specific colors are adjusted according to actual needs, but the colors should be eye-catching). The red area is coated with high-strength reflective waterproof paint to form a high-contrast aerial photography identification mark. A through hole is opened in the geometric center of the float and a fixing rod 21 is pre-installed. In this embodiment, the upper fixing rod 21 is selected as an M16 stainless steel hexagonal nut. The top surface of the nut is flush with the surface of the board. This nut provides a unique standard puncture point position for RTK coordinate acquisition. The entire surface of the float 1 is coated with a waterproof protective coating to resist long-term corrosion from water and ultraviolet rays. The edges are sealed with waterproof to prevent the board from absorbing water and increasing weight. The square and regular shape facilitates automatic recognition of aerial images. The entire underwater anchoring tension is concentrated on the center of the float, which can effectively prevent the float from tilting due to water flow and wind.

[0117] For details, please refer to Figures 2-4The splitter 2 is made entirely of 304 stainless steel (or other rust-resistant materials), possessing the characteristics of long-term underwater rust prevention and resistance to water flow impact. It includes an upper fixed rod 21, a lower fixed rod 22, and an oblique connecting rod 23. The upper end of the lower fixed rod 22 is threaded to the lower end of the upper fixed rod 21, and the lower end of the lower fixed rod 22 is threaded to the center of the oblique connecting rod 23. The oblique connecting rod 23 is an equilateral triangular metal component, with branch holes at each corner for connecting branch wire ropes 3, so that the branch wire ropes are evenly distributed at 120°, stably maintaining a standard tension angle of 45°~60° between the wire rope and the horizontal plane. The entire branch frame unifies and transmits the multi-directional tension generated by multiple underwater branch steel wire ropes 3 to the center of the float, dispersing and offsetting the lateral thrust brought by water flow and wind in different directions, preventing the float from tilting or drifting due to force on one side, ensuring that the reflective markings on the surface of the float 1 always face upward, improving the clarity of aerial photography identification, and at the same time, the connection points of each rod are strong enough to prevent deformation and loosening during long-term underwater operation.

[0118] For details, please refer to Figures 2-4 The branch wire rope 3 is made of three identical 304 stainless steel wire ropes, each with a diameter of 3mm (the specific diameter can be adjusted according to actual needs). The ropes can be cut and processed according to the length calculated based on the water depth in step S2. The three branch wire ropes maintain the same length to ensure balanced underwater force. The upper end of each branch wire rope is uniformly fixed to the evenly distributed branch holes on the inclined connecting rod 23 of the splitter 2. The three ropes naturally spread downwards, with the angle between adjacent branch wire ropes remaining stably at 120°. Combined with the overall assembly structure, this maintains a 45°~60° tension angle between the rope and the water surface. A single weight 4 is attached to the lower end of each branch wire rope to simultaneously tighten the float 1 downwards in three directions. The steel wire rope has excellent resistance to river and lake water corrosion, high tensile strength, and is not easily stretched, deformed or broken under continuous water flow and repeated disturbance of wind and waves. At the same time, the water-facing area of ​​the steel wire rope is included in the underwater equivalent water-blocking area model calculated by the weight of the counterweight in step S2, which can accurately quantify the water flow impact load and match and adapt to the adaptive counterweight calculation scheme. The multi-directional uniform tension structure can offset the lateral hydrodynamic and wind loads from various angles, effectively preventing the float 1 from overturning or drifting, and continuously ensuring that the float marking is horizontal and upward, thus improving the recognition effect of aerial images.

[0119] For details, please refer to Figures 2-4The counterweight 4 is a spherical cast iron component with a galvanized anti-rust treatment on its entire surface, allowing it to be submerged in river and lake water for extended periods, effectively resisting water corrosion and silt erosion. This device is equipped with three identical counterweights, one of which is attached to the lower end of each branch steel wire rope 3. The three counterweights simultaneously provide underwater anchoring tension. The bottom of the counterweight is integrally formed with anti-slip teeth. After sinking to the bottom, the anti-slip teeth embed into the riverbed substrate, increasing the friction between the counterweight and the bottom silt, sand, and rocks, reducing the counterweight's slippage under water flow impact. The spherical shape of the counterweight provides a fixed water-facing projected area. This area value can be directly substituted into the underwater equivalent water-blocking area calculation formula in step S2 to participate in the counterweight solution, and is linked and matched with flow velocity, wind speed, and bottom sediment safety factor to accurately meet the anti-drift requirements under different water conditions. Three equally weighted hammers, together with three equally long branch steel wire ropes 3, form a 120° balanced tension structure, which counteracts the lateral water flow and wind thrust in multiple directions, stabilizes and constrains the float 1, prevents the float from tilting or drifting significantly, and ensures that the reflective markings on the surface of the float are always facing the aerial camera, thereby improving the target's recognition stability in test flights and official aerial images.

[0120] In summary, the present invention has the following advantages: This application provides a method and device for setting up image control points for UAV photogrammetry in large water areas. This application achieves intelligent setting up of image control points in large water areas through a complete closed-loop process of collecting environmental parameters of the deployment area, optimizing aerial survey texture linkage, configuring adaptive buoy anchoring, and iterative re-setting during test flights. Based on measured water depth, flow velocity, wind speed, and bottom sediment data, a dedicated formula is established to match each point with a branch steel wire rope 3 of appropriate length and a corresponding weight of counterweight 4. Simultaneously, differential grid spacing is calculated based on GSD and texture coefficients. Weak areas in the imaging are then located through dual-level judgment during fixed-point test flights. The grid is then layered and iteratively re-inspected according to the degree of defect. This ensures the constraint strength of aerial triangulation calculations and improves the accuracy of water area mapping while reducing buoy consumables and secondary site entry construction costs. The hardware structure and intelligent calculation process support each other, achieving simultaneous improvement in deployment accuracy and operational economy.

[0121] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Terms such as "upper," "lower," "left," "right," "front," and "rear" used in the invention are merely for clarity of description and are not intended to limit the scope of the invention. Changes or adjustments to their relative relationships, without substantially altering the technical content, should also be considered within the scope of the invention.

[0122] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for setting up image control points for unmanned aerial vehicle (UAV) photogrammetry over large bodies of water, characterized in that, Includes the following steps: S1. Collection of environmental parameters of the deployment area: Collect environmental parameters of each preset point in the water area to be deployed. The environmental parameters include water depth, surface water flow velocity and wind speed above the water surface. S2. Adaptive Calculation of Deployment Parameters: Anchoring parameters are adaptively calculated based on environmental parameters at each location, including the length of the branch wire rope and the weight of the matching counterweight. A corresponding safety factor is selected based on the type of underwater substrate. Complete the quantitative calculation of the weight of the counterweight; S3. Optimization of Deployment Density and Aerial Survey Parameters: Calculating the optimal deployment spacing of control points by linking UAV aerial survey parameters with water surface texture features. Generate a grid layout scheme for the water area; optimize the grid spacing. The calculation is based on flight altitude, ground sampling distance (GSD), and effective ground coverage width of a single image; S4. Initial deployment: Deploy water surface image control point devices according to the deployment grid scheme and adaptive anchoring parameters; S5. RTK coordinate acquisition: The 3D geodetic coordinates (X0, Y0, Z0) of all deployed image control points are acquired using RTK. S6. Flight test image acquisition and quality assessment: Conduct flight test image acquisition and rapid on-site quality assessment of UAVs; hover above each control point and at the center of the grid to capture vertical images, and determine the recognizability of the floating board markings and the number of visible control points at the center of the grid, and delineate unqualified weak areas and areas that need attention and observation. S7. Targeted Supplementation: Address the weak areas identified in the assessment by adjusting the density coefficient. The control points were added in a graded and encrypted manner, and the RTK coordinates were remeasured after the addition was completed. S8. Iterative verification: Iteratively execute steps S6 to S7 until the quality of all water area images meets the standards and the image control points are deployed.

2. The method for setting up image control points for UAV photogrammetry in large water areas according to claim 1, characterized in that, The formula for calculating the length of the branch wire rope in step S2 is as follows: The formula for calculating the weight of the hammer is: ;in The angle between the wire rope and the plumb line. Because of the water depth, For the length of the branch wire rope, The underwater equivalent water-blocking area of ​​the branch wire rope and counterweight. The equivalent windward area of ​​the floating platform. The surface water velocity, For wind speed, The weight of the hammer. This is for the safety factor.

3. The method for setting up image control points for UAV photogrammetry in large water areas according to claim 2, characterized in that, The formula for calculating the underwater equivalent water-blocking area of ​​the single branch steel wire rope and the counterweight is as follows: In the formula The diameter of the wire rope, For the length of the rope, The area of ​​the hammer facing the water is the surface area.

4. The method for setting up image control points for UAV photogrammetry in large water areas according to claim 1, characterized in that, The formula for calculating the ground sampling distance in step S3 is as follows: The formula for calculating the effective area coverage width of a single image is as follows: , The formula for calculating the optimal layout spacing is as follows: In the formula, The pixel dimension of the image. , These are the heading and lateral overlap, respectively. These represent the effective coverage width in the heading and lateral directions, respectively. This represents the minimum equivalent number of pixels required to cover control points within a single image. The altitude at which the drone flies. For camera pixel size, The focal length of the camera; The sparsity coefficient of water texture; Ground sampling distance, The optimal spacing is determined by the spacing between the installations.

5. The method for setting up image control points for UAV photogrammetry in large water areas according to claim 1, characterized in that, Step S6, the image quality assessment criteria, includes the identifiability of control points and the number of visible control points in the grid center image; both the identifiability of control points and the number of visible control points in the grid center image are categorized as qualified, pending, and unqualified. The qualified condition for the image control point identifiability is defined as red and white markings with a clear central cross without any reflective obstruction; the unqualified condition is defined as slight reflectivity but the markings are still identifiable; and the unqualified condition is defined as ripples or strong light that make the target completely unidentifiable. The qualified number of visible control points in the grid center image is defined as ≥4, the unqualified number is defined as 2~3, and the unqualified number is defined as 0~1.

6. The method for setting up image control points for UAV photogrammetry in large water areas according to claim 1, characterized in that, Step S7: Formula for calculating the spacing of the encryption and supplementation: In the formula For encryption coefficients, To add spacing for encryption.

7. The method for setting up image control points for UAV photogrammetry in large water areas according to claim 6, characterized in that, The The value of is no greater than 1, and it is set as follows: if the image stitching fails or displays holes, it is considered severely weak. Take a value of 0.3~0.4; If the splicing is obviously misaligned, it is considered moderately weak. Take a value of 0.4~0.6; if there is slight misalignment in the splicing, it is considered slightly weak, and at this time... Take a value of 0.6 to 0.

7.

8. An apparatus employing the image control point deployment method for large-area unmanned aerial vehicle (UAV) photogrammetry according to any one of claims 1 to 7, characterized in that, Includes floats, distributors, branch wire ropes, and cast iron counterweights; The splitter includes an upper fixed rod, a lower fixed rod, and an inclined connecting rod; the branch wire ropes include multiple ropes, one end of which is connected to the inclined connecting rod; the center of the inclined connecting rod is connected to the lower end of the lower fixed rod, and the upper end of the lower fixed rod is connected to the lower end of the upper fixed rod; the upper fixed rod passes through the float and is located at the exact center of the float.

9. A device for setting up image control points for large-area unmanned aerial vehicle (UAV) photogrammetry according to claim 8, characterized in that, The length of the branch wire rope varies according to the water depth and the vertical angle of the wire rope, and the weight of the cast iron hammer is selected according to the water depth, water flow, wind speed and bottom sediment.

10. A device for setting up image control points for large-area unmanned aerial vehicle (UAV) photogrammetry according to claim 8, characterized in that, The float is a square rigid float with reflective pigment coated on the side that floats on the water surface.

Images