A method and system for unmanned aerial vehicle geographic information collection

Abstract

The application discloses a kind of unmanned plane geographic information collection method and system, belong to geographic information collection technical field, including surveying and mapping area basic data acquisition, environment parameter grading quantization, operation parameter coupling operation, no step continuous fine adjustment and disturbance locking and associated data acquisition, blind area discrimination and supplementary acquisition, the operation parameter of the present application can be dynamically adjusted according to the field topography gradient and vegetation shelter state unmanned plane, improve the data acquisition quality under complex environment.

Description

Technical Field

[0001] This invention relates to the field of geographic information acquisition technology, and more specifically, to a method and system for acquiring geographic information using unmanned aerial vehicles (UAVs). Background Technology

[0002] At present, drones have been widely used in geographic information collection operations such as land surveying, mountain topography survey, forest resource survey and geological disaster hazard monitoring. Existing drone geographic information collection operations generally adopt a standardized flight planning mode with fixed flight altitude, fixed flight path spacing and fixed photo sampling parameters preset in advance. The drone completes the collection of full-area images and point cloud data at a constant speed and at the same altitude along the fixed flight path.

[0003] However, my country's surveying and mapping operations are mostly located in complex and composite landforms, including mountains, hills, ravines, steep slopes, and a mix of trees and shrubs. The terrain is varied and the vegetation is unevenly distributed. Conventional fixed-parameter aerial data collection methods cannot dynamically adapt to changes in the terrain and vegetation cover. This can easily lead to blind spots in geographic information collection in areas with large slope differences or dense vegetation cover, resulting in quality defects such as missed surface topographic data, voids and distortions in 3D modeling, and missing texture information of ground features. Subsequent manual data comparison and re-flying are required, which results in long operation cycles, high labor costs, and difficulty in balancing data collection efficiency and surveying data accuracy. Summary of the Invention

[0004] To address the problems existing in the prior art, the present invention aims to provide a method and system for collecting geographic information by unmanned aerial vehicles (UAVs), which can dynamically adjust the operating parameters of the UAV according to the slope of the terrain and the state of vegetation obstruction, thereby improving the quality of data collection in complex environments.

[0005] To solve the above problems, the present invention adopts the following technical solution.

[0006] A method for collecting geographic information by unmanned aerial vehicles (UAVs), characterized by comprising the following steps:

[0007] S1. Basic data collection in the surveying area: Preset benchmark collection operation parameters, including benchmark safe relative flight altitude, benchmark lateral image overlap, and benchmark image shooting time interval; The UAV enters the surveying area with the benchmark collection operation parameters to collect environmental parameters, UAV positioning data, and flight attitude data of the operation point. The environmental parameters include terrain slope parameters and vegetation occlusion status parameters.

[0008] S2. Environmental Parameter Classification and Quantization: Preset slope classification rules and shading classification rules to classify and quantify the collected terrain slope parameters and vegetation shading status parameters to obtain corresponding environmental quantification coefficients. The environmental quantification coefficients include terrain slope weight coefficient, terrain slope safety coefficient, and vegetation shading level coefficient. The terrain slope weight coefficient and terrain slope safety coefficient are matched according to the slope classification rules, and the vegetation shading level coefficient is matched according to the shading classification rules.

[0009] S3. Operational parameter coupling operation: Obtain the baseline acquisition operation parameters and environmental quantization coefficients, and generate adaptive operation parameters in real time according to the corresponding coupling operation relationship to adapt to the current on-site operation conditions. The adaptive operation parameters include real-time flight relative altitude, real-time flight path lateral overlap, and real-time image capture sampling interval.

[0010] S4. Stepless Continuous Fine-Tuning and Disturbance Locking: Each adaptive operation parameter is compared with the current real-time operation parameter item by item, and an error threshold is preset for each. When any difference exceeds the error threshold, a stepless continuous fine-tuning instruction is generated. The stepless continuous fine-tuning instruction includes:

[0011] Continuous and smooth adjustments are made to flight altitude and lateral offset of the flight path;

[0012] When short-term fluctuations or disturbances occur in the flight attitude, the dynamic control of the adaptive operation parameters is temporarily locked, and the lock is released after the flight attitude is restored.

[0013] The image capture sampling interval is updated synchronously via the camera control module;

[0014] S5. Related data acquisition, blind spot identification and supplementary acquisition: The UAV synchronously acquires geographic imagery and terrain point cloud data according to the current adaptive operation parameters, and stores the acquired data after associating and marking it with the corresponding operation parameters.

[0015] Meanwhile, the system can identify blind spots in real time based on the following criteria: the slope angle of the terrain at multiple consecutive work sites changes beyond the preset change threshold, or the vegetation obstruction level changes from the current level to a higher level and the duration exceeds the preset level switching stability time; after the main route data collection is completed, the system can autonomously plan a supplementary data collection route and execute steps S2 to S4 again during the supplementary data collection operation until the data collection of the entire area meets the standard and the drone returns autonomously.

[0016] Preferably, the corresponding coupling operation relationship of each of the adaptive operation parameters is as follows:

[0017] Real-time flight relative altitude = baseline safe relative altitude × terrain slope safety factor;

[0018] Real-time flight path lateral overlap = baseline lateral image overlap × (terrain slope weighting coefficient × vegetation occlusion level coefficient);

[0019] Real-time image capture sampling interval = baseline image capture time interval ÷ (terrain slope weight coefficient × vegetation occlusion level coefficient);

[0020] Wherein, the terrain slope safety factor is greater than or equal to 1; when the product of the terrain slope weight coefficient and the vegetation shading level coefficient causes the calculated value of the real-time flight path lateral overlap to exceed 100%, the upper limit of 100% is taken.

[0021] Preferably, the slope grading rule is divided into multiple gradients according to the slope angle range, and each gradient corresponds to a preset terrain slope weight coefficient and a preset terrain slope safety coefficient; the shading grading rule is divided into multiple levels according to vegetation light transmittance or canopy thickness, and each level corresponds to a preset vegetation shading level coefficient; wherein, the greater the slope, the greater the corresponding terrain slope weight coefficient and safety coefficient, and the more severe the vegetation shading, the greater the corresponding vegetation shading level coefficient.

[0022] Preferably, the error thresholds include an altitude error threshold, an overlap error threshold, and a sampling interval error threshold; the preset disturbance threshold for the flight attitude change rate is pre-calibrated based on the stable flight performance of the UAV; the release condition for the temporary lock is: the time during which the attitude change rate is continuously lower than the disturbance threshold reaches a preset recovery stabilization time.

[0023] Preferably, it also includes

[0024] S6. Operational Anomaly Tolerance Steps: Real-time monitoring of the UAV's remaining battery power and sensor module operating status. When the remaining battery power is lower than the preset safe battery power threshold, the system automatically prioritizes ensuring the integrity of data collection in the core mapping area, reduces the data collection requirements for non-core areas, and plans the shortest return route. When a single sensor module experiences a short-term abnormal data loss, the system continues to adjust the operation by using the product of the effective terrain slope weight coefficient and vegetation occlusion level coefficient from the previous frame. Simultaneously, inertial extrapolation is performed using data from the UAV's inertial measurement unit to fill in the gaps without interrupting the data collection process.

[0025] Preferably, when both the terrain slope weight coefficient and the vegetation occlusion level coefficient obtained in step S2 are greater than their respective error thresholds, the current operating area is determined to be a complex data collection area with overlapping terrain undulations and vegetation occlusion; within this area, the airborne edge computing terminal automatically executes at least one of the following special data collection strategies:

[0026] Force the real-time lateral overlap of flight paths to the maximum value;

[0027] Based on the calculation results of the S3 formula, the sampling interval for real-time image capture is further shortened;

[0028] The onboard lidar is triggered to operate in enhanced mode, and the drone's flight speed is reduced, with lidar point clouds serving as the primary data source for geographic information collection.

[0029] Preferably, both geographic imagery data and terrain point cloud data are synchronously associated with and labeled with real-time geographic coordinates, flight attitude parameters, real-time flight altitude parameters, and vegetation occlusion level parameters.

[0030] A UAV geographic information acquisition system includes:

[0031] The drone platform is equipped with a high-precision RTK positioning module, inertial measurement unit, lidar, and multispectral or visible light camera;

[0032] The airborne edge computing terminal has a built-in terrain slope calculation module, vegetation shading classification module, two-factor coupling calculation module, smooth control command generation module, and blind spot identification and re-sampling planning module.

[0033] Ground monitoring terminals are used to display real-time data acquisition progress, receive abnormal alarms, and issue intervention commands.

[0034] The airborne edge computing terminal communicates with the UAV flight control system in real time via a data bus, and writes the generated stepless continuous fine-tuning instructions into the attitude control loop and camera control module of the flight control system, respectively.

[0035] Preferably, the airborne edge computing terminal also has a built-in fault-tolerant logic unit, which is connected to the power monitoring circuit and the data stream of each sensor. When the loss of any sensor data exceeds the preset loss tolerance time, it automatically switches to the inertial hold mode and issues an alarm; when the power is lower than the preset emergency return threshold, it forcibly executes the safety return mode.

[0036] Compared with the prior art, the advantages of this invention are:

[0037] This invention pre-collects terrain slope parameters and vegetation occlusion status parameters, and then quantifies them to obtain terrain slope weight coefficient, terrain slope safety coefficient, and vegetation occlusion level coefficient. These environmental quantification coefficients are then coupled with baseline operating parameters to generate adaptive operating parameters that are adapted to the current site conditions in real time. This invention can dynamically adjust flight altitude, flight path density, and photo capture frequency according to the site environment. It can improve the integrity of the collected data without increasing human intervention, while reducing the risk of drone collisions by increasing the safe flight altitude.

[0038] After generating adaptive operation parameters, instead of instructing the UAV to make a step-like abrupt adjustment, the adaptive parameters are compared with the real-time operation parameters item by item, and a stepless continuous fine-tuning command is generated only when the error exceeds the preset error threshold. The flight altitude and lateral offset of the flight path are continuously and smoothly adjusted, avoiding the image resolution jump, sudden flight path deviation or gimbal overload caused by parameter abruptness in traditional methods. This ensures the stability of the base-to-height ratio between acquired images, which is convenient for later stitching and 3D modeling.

[0039] When the flight attitude fluctuates briefly due to gusts or other disturbances, the system temporarily locks the dynamic control of adaptive parameters. Once the flight attitude stabilizes, the lock is automatically released, preventing environmental parameters from being misjudged and over-responded to due to the UAV's own attitude disturbances, and further improving the robustness of the control system.

[0040] During the data collection process, based on two criteria, the data collection blind spots are identified in real time: the change in terrain slope angle exceeds the preset change threshold, or the vegetation obstruction level continues to rise for more than the preset stabilization time. After the main route is completed, the airborne edge computing terminal can autonomously plan the supplementary data collection route and perform environmental parameter classification quantification and adaptive parameter calculation again in the supplementary data collection operation until the entire domain is collected, which significantly reduces the number of days of field work and data redundancy.

[0041] When both the slope weight coefficient and the vegetation occlusion level coefficient are greater than their respective base values, the system automatically identifies the area as a complex data collection region where terrain undulations and vegetation occlusion overlap. It then forces a special data collection strategy: forcing the lateral overlap to its maximum value, further shortening the sampling interval, switching to lidar as the primary data source, and reducing flight speed. Compared to existing solutions that simply increase overlap in complex areas, this invention can proactively switch the primary and secondary relationships of data acquisition (point cloud as the main source, imagery as a secondary source) based on the severity of the environment. This allows for the acquisition of terrain data that meets production requirements with lower energy consumption and higher point cloud density, making it particularly suitable for scenarios such as power line inspection and forestry surveys in mountainous areas. Attached Figure Description

[0042] Fig. 1 This is a flowchart illustrating the UAV geographic information collection method according to an embodiment of this application.

[0043] Fig. 2 This is a structural block diagram of the UAV geographic information acquisition system according to an embodiment of this application.

[0044] The labels in the diagram are as follows: 100, UAV platform; 200, airborne edge computing terminal; 300, ground monitoring terminal. Detailed Implementation

[0045] The following will refer to the appendices in the embodiments of the present invention. Figs. 1-2The technical solutions in the embodiments of the present invention are clearly and completely described herein. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0046] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.

[0047] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0048] The terms “comprising” and “including” indicate the presence of the described feature, whole, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0049] The term “and / or” refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes these combinations.

[0050] Firstly, this application provides a method for collecting geographic information using a UAV, applied to a surveying area with typical terrain undulations (such as hilly areas) and local vegetation obstruction (such as sparse forests and shrubs), aiming to obtain high-precision and high-completeness geographic information data, including the following steps:

[0051] S1 Survey Area Basic Data Collection

[0052] Before executing the mission, the operator presets a set of reference acquisition parameters for the UAV (equipped with a high-precision RTK positioning module, inertial measurement unit, lidar, and multispectral or visible light camera) in this embodiment via a ground monitoring terminal, wherein: the reference safe relative flight altitude ( : 80 meters (safe height set for the average elevation of the area); baseline lateral image overlap ( :60% (meets the basic overlap rate requirements for conventional terrain 3D reconstruction); reference image acquisition time interval ( : 2.0 seconds (corresponding to a baseline length of approximately 20 meters at a cruising speed of 10 m / s).

[0053] After takeoff, the UAV enters the first flight path of the surveying area based on the aforementioned benchmark parameters. During flight, the onboard edge computing terminal collects environmental parameters of the work site in real time.

[0054] Terrain slope parameter: The angle between the ground surface normal vector directly below the UAV and the vertical line is calculated by real-time calculation of the point cloud data transmitted back by the airborne lidar. For example, at a certain point, the calculated local terrain slope angle is 28°;

[0055] Vegetation occlusion parameters: By analyzing real-time images from multispectral or visible light cameras, the degree of vegetation occlusion on the ground surface is assessed using a preset vegetation index (NDVI) or a deep learning-based semantic segmentation model. For example, at the same location, analysis shows that the surveyed area is a "dense shrubland area" with a light transmittance of approximately 40%, which is classified as moderate occlusion.

[0056] At the same time, the UAV positioning data (RTK coordinates) and flight attitude data (roll angle, pitch angle, yaw angle) at this location were also recorded simultaneously.

[0057] S2. Classification and quantification of environmental parameters:

[0058] Preset slope grading rules and shading grading rules are used to quantify and grade the collected terrain slope parameters and vegetation shading status parameters, resulting in corresponding environmental quantification coefficients. These environmental quantification coefficients include terrain slope weighting coefficients, terrain slope safety coefficients, and vegetation shading level coefficients. The terrain slope weighting coefficient and terrain slope safety coefficient are matched according to the slope grading rules, while the vegetation shading level coefficient is matched according to the shading grading rules. The slope grading rules are divided into multiple gradients based on the slope angle range, with each gradient corresponding to a preset terrain slope weighting coefficient and a preset terrain slope safety coefficient. The shading grading rules are divided into multiple levels based on vegetation light transmittance or canopy thickness, with each level corresponding to a preset vegetation shading level coefficient. Specifically, the greater the slope, the larger the terrain slope weighting coefficient and safety coefficient; the more severe the vegetation shading, the larger the vegetation shading level coefficient.

[0059] Slope classification rules:

[0060] 0°≤slope<15°: Terrain slope weighting coefficient Safety factor for terrain slope ;

[0061] 15°≤slope<30°: , ;

[0062] 30°≤slope<45°: , ;

[0063] Slope ≥ 45°: , ;

[0064] The collected terrain slope is 28°, which falls within the 15°-30° range. Therefore, the terrain slope weighting coefficient is obtained through matching. Safety factor for terrain slope ;

[0065] Occlusion classification rules:

[0066] Level 0 (No or low shading, light transmittance >80%): Vegetation shading level coefficient

[0067] Level 1 (Moderate shading, 50% < light transmittance ≤ 80%):

[0068] Level 2 (Heavy shading, light transmittance ≤ 50%):

[0069] For "dense shrubland area, light transmittance 40%", the following vegetation shading level coefficient is obtained: .

[0070] S3, Coupled Calculation of Job Parameters:

[0071] The system acquires baseline acquisition parameters and environmental quantization coefficients, and generates adaptive operational parameters in real time according to the corresponding coupling operation relationships. These adaptive operational parameters include real-time flight relative altitude. ), real-time lateral overlap of flight routes ( ), real-time image capture sampling interval ( );

[0072] Baseline data acquisition parameters are obtained through an airborne edge computing terminal. , , ) and environmental quantification coefficient ( , , It generates adaptive job parameters in real time according to the preset coupling operation relationship;

[0073] Real-time relative flight altitude:

[0074] The steeper the terrain, the greater the safety factor, and the adaptive flight altitude is also increased to ensure the safety margin for drone collisions and airflow disturbances;

[0075] Real-time lateral overlap of flight paths: Since the calculated result of 151.2% exceeds the actual physical upper limit of 100%, therefore the final...

[0076] This means that in areas with undulating terrain and severe vegetation obstruction, drones will implement 100% lateral overlap (i.e., adjacent flight paths completely overlap) to maximize data redundancy and ensure a high probability of capturing ground information.

[0077] Real-time image capture sampling interval:

[0078] To address the risk of information loss due to terrain and vegetation, the sampling interval was shortened to approximately 0.79 seconds, which is about 2.5 times higher than the baseline, ensuring that the heading overlap is significantly improved while the flight speed remains constant.

[0079] S4. Stepless continuous fine-tuning and perturbation locking:

[0080] Each adaptive operation parameter is compared with the current real-time operation parameter item by item, and an error threshold is preset for each. The error thresholds include flight altitude error threshold, overlap error threshold, and sampling interval error threshold. When any difference exceeds the error threshold, a stepless continuous fine-tuning command is generated. The stepless continuous fine-tuning command includes:

[0081] Continuous and smooth adjustments are made to flight altitude and lateral offset of the flight path;

[0082] When short-term fluctuations or disturbances occur in the flight attitude, the dynamic control of the adaptive operation parameters is temporarily locked. The preset disturbance threshold of the flight attitude change rate is pre-calibrated based on the stable flight performance of the UAV, and the lock is released after the flight attitude is restored. The condition for releasing the temporary lock is: the attitude change rate is continuously lower than the disturbance threshold for a preset recovery stabilization time.

[0083] The image capture sampling interval is updated synchronously via the camera control module;

[0084] The current real-time relative flight altitude of the drone is 104 meters (just adjusted), the real-time lateral overlap is 100% (achieved by compressing the flight path spacing), and the real-time image capture sampling interval is 0.79 seconds (effective). Assuming that due to gusts, the drone's flight attitude experiences a short-term fluctuation, and the inertial measurement unit (IMU) measures an attitude change rate exceeding a preset disturbance threshold (e.g., roll angle change rate > 30° / s), the onboard edge computing terminal executes disturbance locking logic:

[0085] Temporary lock: Suspend the dynamic adjustment of adaptive operation parameters from step S3. That is, even if the environmental parameters change slightly at this time, no new flight altitude, overlap, or sampling interval commands will be generated, and the current lock value will be maintained. , , ).

[0086] Recovery and unlocking: When the wind weakens and the IMU detects that the attitude change rate has been continuously below the disturbance threshold for a preset recovery stabilization time (e.g., 5 consecutive seconds <30° / s), the lock is released, and real-time monitoring and difference comparison of the output of step S3 are resumed.

[0087] During normal flight (no disturbance or disturbance resolved) and when the difference comparison exceeds the error threshold, a stepless continuous fine-tuning command is generated. For example, when the bank angle increases from 28° to 33° ( It jumped to 1.9. (Jumped to 1.7), the newly calculated The altitude changes to 136 meters. The difference (32 meters) from the current flight altitude of 104 meters exceeds the preset flight altitude error threshold (e.g., 5 meters). At this point, the flight control system will not instruct the UAV to directly climb 32 meters, but will instead continuously and smoothly adjust to 136 meters at a rate of, for example, 2 meters per step over the next few path points, to avoid drastic changes in image resolution or difficulties in image stitching caused by sudden changes in flight altitude.

[0088] S5. Related data collection, blind spot identification and supplementary collection:

[0089] The UAV synchronously collects geographic imagery and terrain point cloud data according to the current adaptive operation parameters, and stores the collected data after associating and marking it with the corresponding operation parameters. Both geographic imagery data and terrain point cloud data are synchronously associated and marked with real-time geographic coordinates, flight attitude parameters, real-time flight altitude parameters and vegetation occlusion level parameters.

[0090] Meanwhile, the following criteria are used to identify blind spots in real time: the slope angle of the terrain at multiple consecutive work sites changes beyond the preset change threshold, or the vegetation obstruction level changes from the current level to a higher level and the duration exceeds the preset level switching stability time; after the main route is completed, the drone will autonomously plan a supplementary route and execute steps S2 to S4 again during the supplementary operation until the full-area collection meets the standard and then the drone will autonomously return.

[0091] The drone follows the adaptive operation parameters generated in step S3. , , Simultaneously acquire geographic imagery and terrain point cloud data. Each image and each point cloud frame is associated with its geographic coordinates, flight attitude, real-time altitude, and vegetation occlusion level. The parameters are associated with the tags and stored in the onboard memory.

[0092] Blind Spot Identification: The airborne edge computing terminal monitors the terrain slope sequence in real time. For example, if the slope angle of five consecutive work points (interval approximately equal to the flight path interval) changes from 15°→28°→33°→35°→31°, and the change exceeds a preset change threshold (e.g., 12° / point), the system determines that the area is a data collection blind spot. Alternatively, if the vegetation obstruction level changes from level 1 (or lower) within 5 seconds... Switch to Level 2 ( If the signal persists for more than the preset level switching stability time (e.g., 3 seconds), blind spot detection will also be triggered.

[0093] Autonomous Supplementary Data Collection: After the main flight path (a "bow"-shaped path covering the entire survey area) is completed, the system autonomously plans a supplementary data collection path (such as a "return"-shaped densified path) based on the recorded blind spot locations. When the UAV performs supplementary data collection, it starts from step S2 again (i.e., reassessing the slope and vegetation coefficient within the blind spot) and generates new adaptive operation parameters. For example, within the blind spot, the flight altitude may be further adjusted to a higher value, or the sampling interval may be shortened to 0.5 seconds. This continues until all identified blind spots are successfully covered, achieving full-area data collection, and the UAV autonomously returns to base.

[0094] When both the terrain slope weight coefficient and the vegetation occlusion level coefficient obtained in step S2 are greater than their respective error thresholds, the current operating area is determined to be a complex data collection area with overlapping terrain undulations and vegetation occlusion; within this area, the airborne edge computing terminal automatically executes at least one of the following special data collection strategies:

[0095] Force the real-time lateral overlap of flight paths to the maximum value;

[0096] Specifically, when both the terrain slope weight coefficient and the vegetation occlusion level coefficient obtained in step S2 are greater than their respective error thresholds, the current operating area is determined to be a complex data collection area with overlapping terrain undulations and vegetation occlusion; within this area, the airborne edge computing terminal automatically executes at least one of the following special data collection strategies:

[0097] Force the real-time lateral overlap of flight paths to the maximum value;

[0098] Based on the calculation results of the S3 formula, the sampling interval for real-time image capture is further shortened;

[0099] The onboard lidar is triggered to operate in enhanced mode, and the drone's flight speed is reduced, with lidar point clouds serving as the primary data source for geographic information collection.

[0100] Based on the calculation results of the S3 formula, the sampling interval for real-time image capture is further shortened;

[0101] The onboard lidar is triggered to operate in enhanced mode, and the drone's flight speed is reduced, with lidar point clouds serving as the primary data source for geographic information collection.

[0102] This application provides a method for collecting geographic information using unmanned aerial vehicles (UAVs), which also includes...

[0103] S6. Error tolerance procedures for operations:

[0104] The system monitors the remaining battery power and sensor module status of the drone in real time. When the remaining battery power is lower than the preset safe battery power threshold, it automatically prioritizes the integrity of data collection in the core mapping area, reduces the data collection requirements for non-core areas, and plans the shortest return route. When a single sensor module data is temporarily missing, the system continues to adjust the operation by using the product of the effective terrain slope weight coefficient and vegetation occlusion level coefficient from the previous frame. At the same time, it uses inertial extrapolation data from the drone's inertial measurement unit to fill in the gaps without interrupting the data collection process.

[0105] During flight, the onboard edge computing terminal monitors the remaining battery power in real time. When the battery power falls below a preset safe battery power threshold (e.g., 20%), fault-tolerant logic is automatically activated:

[0106] Data collection preservation strategy: The drone's flight speed is reduced to 5 m / s, and priority is given to collecting data from the core mapping area (such as the user-preset geographic central rectangle). For peripheral non-core areas, the data collection requirements are reduced by automatically decreasing the real-time flight path lateral overlap from 100% to 80% and extending the sampling interval from 0.79 seconds to 1.2 seconds, thus sacrificing flight endurance to obtain complete data from the core area.

[0107] Sensor Loss Handling: If the multispectral camera experiences a brief malfunction at some point, interrupting the data stream, the onboard edge computing terminal will detect that the data loss exceeds a preset tolerance time (e.g., 0.5 seconds) and automatically lock the valid environmental quantization coefficients from the previous frame. , This product is used to continuously adjust operational parameters, and data from the inertial measurement unit (IMU) is fused for flight status extrapolation to ensure uninterrupted data acquisition. Simultaneously, a "camera module malfunction alarm" is sent to the ground monitoring terminal via data link.

[0108] Secondly, this application provides a UAV geographic information acquisition system, including...

[0109] 100 Drone Platform: Utilizing a hexacopter industrial-grade drone, equipped with:

[0110] High-precision RTK positioning module (horizontal accuracy 1cm+1ppm, vertical accuracy 2cm+1ppm).

[0111] Inertial measurement unit (IMU, which includes a three-axis accelerometer, gyroscope, and magnetometer).

[0112] Mechanical rotating lidar (16 / 32 lines, effective range ≥150m).

[0113] Five-lens tilt camera (downward view + four-way tilt, total pixels ≥ 210 million).

[0114] 200 airborne edge computing terminals: The core is the NVIDIA Jetson AGXXavier module, which has the following built-in software modules:

[0115] Terrain slope calculation module: Calculates local terrain slope in real time based on the input LiDAR point cloud.

[0116] Vegetation occlusion classification module: Based on the input visible light / multispectral image, it uses a deep learning model to output vegetation occlusion level coefficients.

[0117] Two-factor coupled operation module: executes the multiplication and division operation logic in step S3 above to couple the environmental coefficient with the baseline parameter.

[0118] Smooth adjustment instruction generation module: Implements the difference comparison, stepless continuous fine-tuning instruction generation and disturbance locking logic in step S4.

[0119] Blind spot identification and supplementary sampling planning module: realizes the dynamic identification of blind spots and the autonomous generation and execution of supplementary sampling routes in step S5.

[0120] Anomaly-tolerant logic unit: This unit is connected to the power monitoring circuit and the data stream of each sensor of the UAV. When the loss of any sensor data exceeds the preset loss tolerance time (e.g., 1 second), it automatically switches to the inertial hold mode (locks the effective environmental coefficient of the previous frame) and issues an alarm. When the power is lower than the preset emergency return-to-home threshold (e.g., 10%), it forcibly executes the preservation return-to-home mode (abandons the supplementary data collection and returns directly at the best speed).

[0121] 300 Ground Monitoring Terminal: An industrial tablet PC with UAV networking communication capabilities, running customized mission planning software. This terminal communicates with the UAV platform via 4G / 5G or point-to-point data link for:

[0122] Send the mapping area KML / KMZ file, preset benchmark operation parameters and classification rules to the UAV;

[0123] It can receive and display the data collection progress, adaptive operation parameters, RTK trajectory, and alarm information transmitted back by the drone in real time.

[0124] In emergency situations, operators are allowed to intervene remotely, such as issuing priority commands like "return immediately," "hover," or "skip the current blind spot."

[0125] The ground monitoring terminal plans the task and uploads it to the airborne edge computing terminal; the UAV starts up and uses RTK to obtain centimeter-level positioning; during flight, LiDAR and camera data streams enter the airborne edge computing terminal to calculate slope and vegetation coefficient in real time; the dual-factor coupled calculation module calculates adaptive operation parameters according to preset formulas; among them, the smooth control command generation module communicates in real time with the flight control system's data bus to write the smoothly adjusted flight altitude and flight path offset commands into the flight control's attitude control loop, and writes the synchronously updated sampling interval commands into the camera control module; with this closed-loop feedback mechanism, the UAV autonomously and efficiently completes the entire mapping and supplementary data collection task until it returns to base after achieving the target.

[0126] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for collecting geographic information by unmanned aerial vehicles (UAVs), characterized in that, Includes the following steps: S1. Basic data collection in the surveying area: Preset benchmark collection operation parameters, including benchmark safe relative flight altitude, benchmark lateral image overlap, and benchmark image shooting time interval; The UAV enters the surveying area with the benchmark collection operation parameters to collect environmental parameters, UAV positioning data, and flight attitude data of the operation point. The environmental parameters include terrain slope parameters and vegetation occlusion status parameters. S2. Environmental Parameter Classification and Quantization: Preset slope classification rules and shading classification rules to classify and quantify the collected terrain slope parameters and vegetation shading status parameters to obtain corresponding environmental quantification coefficients. The environmental quantification coefficients include terrain slope weight coefficient, terrain slope safety coefficient, and vegetation shading level coefficient. The terrain slope weight coefficient and terrain slope safety coefficient are matched according to the slope classification rules, and the vegetation shading level coefficient is matched according to the shading classification rules. S3. Operational parameter coupling operation: Obtain the baseline acquisition operation parameters and environmental quantization coefficients, and generate adaptive operation parameters in real time according to the corresponding coupling operation relationship to adapt to the current on-site operation conditions. The adaptive operation parameters include real-time flight relative altitude, real-time flight path lateral overlap, and real-time image capture sampling interval. S4. Stepless Continuous Fine-Tuning and Disturbance Locking: Each adaptive operation parameter is compared with the current real-time operation parameter item by item, and an error threshold is preset for each. When any difference exceeds the error threshold, a stepless continuous fine-tuning instruction is generated. The stepless continuous fine-tuning instruction includes: Continuous and smooth adjustments are made to flight altitude and lateral offset of the flight path; When short-term fluctuations or disturbances occur in the flight attitude, the dynamic control of the adaptive operation parameters is temporarily locked, and the lock is released after the flight attitude is restored. The image capture sampling interval is updated synchronously via the camera control module; S5. Related data acquisition, blind spot identification and supplementary acquisition: The UAV synchronously acquires geographic imagery and terrain point cloud data according to the current adaptive operation parameters, and stores the acquired data after associating and marking it with the corresponding operation parameters. Meanwhile, the system can identify blind spots in real time based on the following criteria: the slope angle of the terrain at multiple consecutive work sites changes beyond the preset change threshold, or the vegetation obstruction level changes from the current level to a higher level and the duration exceeds the preset level switching stability time; after the main route data collection is completed, the system can autonomously plan a supplementary data collection route and execute steps S2 to S4 again during the supplementary data collection operation until the data collection of the entire area meets the standard and the drone returns autonomously.

2. The UAV geographic information acquisition method according to claim 1, characterized in that: The corresponding coupling operation relationships of each of the adaptive operation parameters are as follows: Real-time flight relative altitude = baseline safe relative altitude × terrain slope safety factor; Real-time flight path lateral overlap = baseline lateral image overlap × (terrain slope weighting coefficient × vegetation occlusion level coefficient); Real-time image capture sampling interval = baseline image capture time interval ÷ (terrain slope weight coefficient × vegetation occlusion level coefficient); Wherein, the terrain slope safety factor is greater than or equal to 1; when the product of the terrain slope weight coefficient and the vegetation shading level coefficient causes the calculated value of the real-time flight path lateral overlap to exceed 100%, the upper limit of 100% is taken.

3. The UAV geographic information acquisition method according to claim 1, characterized in that: The slope grading rule is divided into multiple gradients according to the slope angle range, and each gradient corresponds to a preset terrain slope weight coefficient and a preset terrain slope safety coefficient; the shading grading rule is divided into multiple levels according to vegetation light transmittance or canopy thickness, and each level corresponds to a preset vegetation shading level coefficient; wherein, the greater the slope, the greater the corresponding terrain slope weight coefficient and safety coefficient, and the more severe the vegetation shading, the greater the corresponding vegetation shading level coefficient.

4. The UAV geographic information acquisition method according to claim 1, characterized in that: The error thresholds include flight altitude error threshold, overlap error threshold, and sampling interval error threshold; the preset disturbance threshold for the flight attitude change rate is pre-calibrated based on the stable flight performance of the UAV; the release condition for the temporary lock is: the time during which the attitude change rate is continuously lower than the disturbance threshold reaches a preset recovery stabilization time.

5. The UAV geographic information acquisition method according to claim 1, characterized in that: Also includes S6. Operational Anomaly Tolerance Steps: Real-time monitoring of the UAV's remaining battery power and sensor module operating status. When the remaining battery power is lower than the preset safe battery power threshold, the system automatically prioritizes ensuring the integrity of data collection in the core mapping area, reduces the data collection requirements for non-core areas, and plans the shortest return route. When a single sensor module experiences a short-term abnormal data loss, the system continues to adjust the operation by using the product of the effective terrain slope weight coefficient and vegetation occlusion level coefficient from the previous frame. Simultaneously, inertial extrapolation is performed using data from the UAV's inertial measurement unit to fill in the gaps without interrupting the data collection process.

6. The UAV geographic information acquisition method according to claim 1, characterized in that: When both the terrain slope weight coefficient and the vegetation occlusion level coefficient obtained in step S2 are greater than their respective error thresholds, the current operating area is determined to be a complex data collection area with overlapping terrain undulations and vegetation occlusion; within this area, the airborne edge computing terminal automatically executes at least one of the following special data collection strategies: Force the real-time lateral overlap of flight paths to the maximum value; Based on the calculation results of the S3 formula, the sampling interval for real-time image capture is further shortened; The onboard lidar is triggered to operate in enhanced mode, and the drone's flight speed is reduced, with lidar point clouds serving as the primary data source for geographic information collection.

7. The UAV geographic information acquisition method according to claim 1, characterized in that: Geographic imagery data and terrain point cloud data are synchronously associated and labeled with real-time geographic coordinates, flight attitude parameters, real-time flight altitude parameters, and vegetation occlusion level parameters.

8. A UAV geographic information acquisition system, characterized in that, include: The drone platform is equipped with a high-precision RTK positioning module, inertial measurement unit, lidar, and multispectral or visible light camera; The airborne edge computing terminal has a built-in terrain slope calculation module, vegetation shading classification module, two-factor coupling calculation module, smooth control command generation module, and blind spot identification and re-sampling planning module. Ground monitoring terminals are used to display real-time data acquisition progress, receive abnormal alarms, and issue intervention commands. The airborne edge computing terminal communicates with the UAV flight control system in real time via a data bus, and writes the generated stepless continuous fine-tuning instructions into the attitude control loop and camera control module of the flight control system, respectively.

9. The UAV geographic information acquisition system according to claim 8, characterized in that: The airborne edge computing terminal also has a built-in fault-tolerant logic unit, which is connected to the power monitoring circuit and the data streams of each sensor. When the loss of any sensor data exceeds the preset loss tolerance time, it automatically switches to the inertial hold mode and issues an alarm; when the power is lower than the preset emergency return threshold, it forcibly executes the safety return mode.

Images