Low-error surveying and mapping unmanned aerial vehicle and surveying and mapping method

Abstract

The invention discloses a low-error surveying and mapping unmanned aerial vehicle and a surveying and mapping method, relates to the technical field of distance measurement, and can solve the technical problems of low data processing precision and poor efficiency in a surveying and mapping process due to the fact that real topographic relief and unmanned aerial vehicle bumping errors cannot be effectively distinguished under complex topographic conditions. Comprising the steps that a ground height data set obtained by measuring a target area through a surveying and mapping unmanned aerial vehicle is acquired, and the ground height data set comprises ground height data of multiple measurement points; screening out a plurality of alternative error data from the ground height data set, wherein the alternative error data are measurement point data deviating from a normal fluctuation range; multiple pieces of target error data are identified from the multiple pieces of alternative error data, wherein the target error data are error data caused by jolting of the surveying and mapping unmanned aerial vehicle in the multiple pieces of alternative error data; performing data correction on the target error data to obtain a corrected ground height data set; and determining a ground height map according to the corrected ground height data set.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of distance measurement, in particular to a low-error surveying and mapping unmanned aerial vehicle and a surveying and mapping method. BACKGROUND

[0002] In recent years, unmanned aerial vehicle surveying and mapping technology has been widely used in land surveying, terrain modeling and other fields. It can quickly obtain large-scale geographic information through aerial photography and remote sensing means, significantly improving the surveying and mapping efficiency and reducing the labor cost. In the prior art, the unmanned aerial vehicle is prone to be disturbed by air flow during flight, resulting in deviation of the collected height data. In order to identify such errors, a threshold is usually set to judge the difference between the data points and the adjacent data, so as to filter out the abnormal values. However, this method is difficult to effectively distinguish the real height change caused by the terrain fluctuation from the measurement error caused by the unmanned aerial vehicle pitching when facing complex terrain, resulting in insufficient data processing accuracy and the need for frequent rework and remeasurement, so the overall surveying and mapping efficiency and accuracy still need to be improved. SUMMARY

[0003] In order to solve the technical problems that the real terrain fluctuation and the unmanned aerial vehicle pitching error cannot be effectively distinguished under complex terrain conditions at present, resulting in low data processing accuracy and poor efficiency in the surveying and mapping process, the purpose of the present application is to provide a low-error surveying and mapping unmanned aerial vehicle and a surveying and mapping method, and the technical solutions adopted are as follows: In a possible implementation manner, the surveying and mapping unmanned aerial vehicle measures the target region according to a plurality of preset flight lines during the measurement, and each preset flight line includes a plurality of measurement points. The plurality of candidate error data are filtered out from the ground height data set, specifically including: for each measurement point data in the ground height data set, determining a deviation degree value of the measurement point data; wherein the deviation degree value is used to represent the local deviation of the measurement point data and the adjacent measurement point data; in the case that the deviation degree value of the measurement point data is greater than or equal to a first preset threshold, the measurement point data is determined as the candidate error data.

[0004] In a possible implementation manner, the surveying and mapping unmanned aerial vehicle measures the target region according to a plurality of preset flight lines during the measurement, and each preset flight line includes a plurality of measurement points. The plurality of candidate error data are filtered out from the ground height data set, specifically including: for each measurement point data in the ground height data set, determining a deviation degree value of the measurement point data; wherein the deviation degree value is used to represent the local deviation of the measurement point data and the adjacent measurement point data; in the case that the deviation degree value of the measurement point data is greater than or equal to a first preset threshold, the measurement point data is determined as the candidate error data.

[0005] In a possible implementation, the bias degree value of the measurement point data is determined, specifically comprising: determining, for each measurement point data in the ground height data set, a first difference value and a second difference value of the measurement point data and adjacent measurement point data; the first difference value is a difference value of the measurement point data and a previous adjacent measurement point data, and the second difference value is a difference value of the measurement point data and a next adjacent measurement point data; and determining the bias degree value of the measurement point data according to the first difference value and the second difference value.

[0006] In a possible implementation, the target error data is identified from the plurality of candidate error data, specifically comprising: determining a bump error quantification value of each candidate error data; the bump error quantification value is used to represent a probability that the candidate error data belongs to the error data caused by the bump of the unmanned aerial vehicle; and in a case where the bump error quantification value of the candidate error data is greater than or equal to a second preset threshold, the candidate error data is determined as the target error data.

[0007] In a possible implementation, the bump error quantification value of each candidate error data is determined, specifically comprising: determining a significant degree value and a dispersion degree value of each candidate error data; the significant degree value is used to represent a bias significance of the candidate error data in the corresponding preset flight line, and the dispersion degree value is used to represent an aggregation degree of the candidate error data in the ground height data set; and the bump error quantification value of each candidate error data is determined according to the bias degree value, the significant degree value and the dispersion degree value of each candidate error data.

[0008] In a possible implementation, the significant degree value of each candidate error data is determined, specifically comprising: for each candidate error data, calculating a third difference value; the third difference value is an absolute value of a sum of difference values of the candidate error data and its left and right adjacent candidate error data; and calculating a fourth difference value; the fourth difference value is an absolute value of a difference value of the candidate error data and an average value of the measurement point data in a local window, and the local window is a continuous measurement point data set in a preset area range centered on the candidate error data; and the significant degree value is determined according to the third difference value, the fourth difference value and the number of measurement points in the local window.

[0009] In a possible implementation, the dispersion degree value of each candidate error data is determined, specifically comprising: for each candidate error data, expanding a neighborhood range according to a preset step length until there is no other candidate error data at a boundary of the neighborhood range, or the neighborhood radius after expansion reaches a preset maximum search radius, with the candidate error data as the center; and the total number of candidate error data in the neighborhood range is counted as the dispersion degree value.

[0010] In a possible implementation, the jolt error quantification value of each candidate error data is determined according to the significance value and the dispersion value of each candidate error data, specifically comprising: determining a trend matching degree value of each candidate error data according to the significance value of each candidate error data; wherein the trend matching degree value is used to represent the consistency degree of the change trend of the candidate error data and the surrounding measurement point data; and calculating the jolt error quantification value of each candidate error data according to the bias degree value, the significance value, the dispersion value and the trend matching degree value of each candidate error data.

[0011] In a possible implementation, the data correction is performed on the target error data, specifically comprising: for each target error data, obtaining a plurality of accurate measurement point data of the target error data in a preset two-dimensional neighborhood in a two-dimensional plane data set; wherein the two-dimensional plane data set is obtained by converting the ground height data set; constructing a plurality of line segments according to the plurality of accurate measurement point data; wherein each line segment connects two accurate measurement point data; for each line segment, calculating a perpendicular distance of the target error data to the line segment and a projection ratio of the target error data on the line segment in the two-dimensional plane; for each line segment, calculating an actual data estimation value corresponding to the line segment according to the projection ratio and the data values corresponding to the two endpoints of the line segment; and calculating a correction value of the target error data according to the actual data estimation values corresponding to all line segments and estimation value weights; wherein the estimation value weights are determined according to the perpendicular distances.

[0012] In a second aspect, the present application provides a low-error surveying and mapping unmanned aerial vehicle, comprising: a data acquisition module, a candidate error data screening module, a target error data identification module, a data correction module and an image output module; the data acquisition module is used to obtain a ground height data set obtained by measuring a target area of the surveying and mapping unmanned aerial vehicle; wherein the ground height data set comprises ground height data of a plurality of measurement points; the candidate error data screening module is used to screen a plurality of candidate error data from the ground height data set; wherein the candidate error data is measurement point data deviating from a normal fluctuation range; the target error data identification module identifies a plurality of target error data from the plurality of candidate error data; wherein the target error data is error data caused by jolting of the surveying and mapping unmanned aerial vehicle in the plurality of candidate error data; the data correction module is used to perform data correction on the target error data to obtain a corrected ground height data set; and the image output module is used to determine a ground height map according to the corrected ground height data set.

[0013] In a third aspect, the present application provides an electronic device, comprising: a processor and a memory; wherein the memory is configured to store one or more programs, the one or more programs comprising computer-executable instructions, and when the electronic device is running, the processor executes the computer-executable instructions stored in the memory, so that the electronic device executes the low-error mapping method as described in the first aspect and any possible implementation manner of the first aspect.

[0014] In a fourth aspect, the present application provides a computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by the electronic device of the present application, cause the electronic device to execute the low-error mapping method as described in the first aspect and any possible implementation manner of the first aspect.

[0015] In a fifth aspect, the present application provides a computer program product comprising instructions, which when executed on a computer, cause the electronic device of the present application to execute the low-error mapping method as described in the first aspect and any possible implementation manner of the first aspect.

[0016] In a sixth aspect, the present application provides a chip system applied to a low-error mapping unmanned aerial vehicle; the chip system comprises one or more interface circuits and one or more processors. The interface circuit and the processor are interconnected through a circuit; the interface circuit is configured to receive a signal from a memory of the low-error mapping unmanned aerial vehicle and send the signal to the processor, the signal comprising computer instructions stored in the memory. When the processor executes the computer instructions, the low-error mapping unmanned aerial vehicle executes the low-error mapping method as described in the first aspect and any possible design manner thereof.

[0017] The present application has the following beneficial effects: by constructing a multi-stage processing flow comprising bias degree value calculation, alternative error data screening, jolt error value determination and target error data correction, the present application realizes accurate identification and effective correction of error data caused by jolt in the mapping process of the unmanned aerial vehicle, significantly improves the accuracy and reliability of the ground height data set, while greatly reducing the repeated mapping workload under the premise of ensuring the mapping accuracy, and comprehensively improves the mapping efficiency and result quality under complex terrain conditions. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the drawings needed in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0019] Figure 1 This is a schematic diagram of the architecture of a low-error mapping drone provided in one embodiment of the present invention; Figure 2 This is a schematic diagram of the architecture of a data acquisition module provided in one embodiment of the present invention; Figure 3 This is a schematic diagram of the architecture of a target error data identification module provided in one embodiment of the present invention; Figure 4 This is a schematic flowchart illustrating a low-error mapping method provided in one embodiment of the present invention. Figure 5 This is a flowchart illustrating another low-error mapping method provided in one embodiment of the present invention. Figure 6 This is a schematic flowchart illustrating another low-error mapping method provided in one embodiment of the present invention. Detailed Implementation

[0020] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the specific implementation methods, structures, features, and effects of the present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0022] The following description, in conjunction with the accompanying drawings, details the specific scheme of a low-error mapping drone and mapping method provided by the present invention.

[0023] For example, such as Figure 1 The diagram shown illustrates the architecture of a low-error mapping drone according to an embodiment of the present invention. The low-error mapping drone 10 includes: a data acquisition module 11, a candidate error data filtering module 12, a target error data identification module 13, a data correction module 14, and an image output module 15. The modules are described below in sequence: (1) Data acquisition module 11.

[0024] The data acquisition module 11 is responsible for collecting ground elevation data from multiple measurement points in the target area according to a preset route, forming a ground elevation dataset, which will serve as the input basis for the alternative error data filtering module 12.

[0025] Optionally, the data acquisition module 11 is configured to acquire a ground height data set of the target region measured by the surveying unmanned aerial vehicle.

[0026] As shown in the figure, the data acquisition module 11 can include three sub-modules, i.e., a route planning sub-module 111, an acquisition execution sub-module 112, and a data storage sub-module 113. The three sub-modules are described as follows: Figure 2 (1.1) Route planning sub-module 111.

[0027] The route planning sub-module 111 is configured to plan a plurality of preset routes according to the terrain and range of the target region. The surveying unmanned aerial vehicle 10 measures the target region at a constant speed and a constant flight height.

[0028] Specifically, the route planning sub-module 111 acquires the boundary coordinates and the measurement accuracy requirement of the target region according to the user interaction, and generates a plurality of preset routes based on these information by using a corresponding algorithm, for example, a grid-like uniform distribution algorithm. During the route generation process, the route planning sub-module 111 ensures that each route is parallel and has a consistent spacing, and each route includes a plurality of measurement points. The preset routes generated by the route planning sub-module 111 are simultaneously sent to the flight control module mentioned later, and the flight control module drives the unmanned aerial vehicle to fly along the preset routes.

[0029] In addition, the preset routes generated by the route planning sub-module 111 are also simultaneously sent to the data storage sub-module 113 for subsequent matching and association of the ground height data and the routes.

[0030] (1.2) Acquisition execution sub-module 112.

[0031] The acquisition execution sub-module 112 is responsible for measuring the ground height data of the measurement points in each preset route.

[0032] Specifically, when the surveying unmanned aerial vehicle flies to a measurement point, the acquisition execution sub-module 112 performs the acquisition of the ground height data at the measurement point. For example, a laser radar sensor is arranged in the acquisition execution sub-module 112. The laser radar sensor emits a laser beam to the ground and records the time when the laser is emitted and the time when the ground reflection signal is received. The vertical distance between the unmanned aerial vehicle and the ground measurement point is calculated by the formula of light speed x (receiving time - emitting time) / 2.

[0033] (1.3) Data storage sub-module 113.

[0034] The data storage sub-module 113 is responsible for storing all associated data of the data acquisition module 11 and outputting the ground height data set to the alternative error data screening module 12.​

[0035] Specifically, the data storage submodule 113 receives the information of the plurality of preset flight paths generated by the flight path planning submodule 111, and the ground height data collected by the execution collecting submodule 112, and combines these data into a two-dimensional ground height data set according to the dimensions of the flight path number and the measurement point number. When the measurement of all the preset flight paths is completed, the data storage submodule 113 outputs the two-dimensional ground height data set to the following alternative error data screening module 12.

[0036] (2) The alternative error data screening module 12.

[0037] The alternative error data screening module 12 is responsible for screening the alternative error data from the ground height data set output by the data collection module 11.

[0038] Optionally, the alternative error data screening module 12 is configured to determine, for each measurement point data in the ground height data set, a deviation degree value of the measurement point data, wherein the deviation degree value is used to represent the local deviation size of the measurement point data and adjacent measurement point data.

[0039] Specifically, after receiving the ground height data set, the alternative error data screening module 12 performs the following operations on each measurement point data: first, calculates the first difference value and the second difference value; then, quantifies the sum of the absolute values of the first difference value and the second difference value by using a preset type function to obtain the deviation degree value of the measurement point data. The preset type function can be a sigmoid function.

[0040] Optionally, the alternative error data screening module 12 is further configured to determine the measurement point data as the alternative error data in a case where the deviation degree value of the measurement point data is greater than or equal to a first preset threshold value. For example, the first preset threshold value can be set to an empirical value of 0.7.

[0041] (3) The target error data identification module 13.

[0042] The target error data identification module 13 is responsible for identifying the target error data from the alternative error data output by the alternative error data screening module 12, which is the error data caused by the bumping of the surveying unmanned aerial vehicle.

[0043] Optionally, the target error data identification module 13 identifies a plurality of target error data from a plurality of alternative error data. The target error data is the error data caused by the bumping of the surveying unmanned aerial vehicle in the plurality of alternative error data.

[0044] For example, as shown in FIG. 6, the target error data identification module 13 identifies the target error data 601 from the alternative error data 602. Figure 3As shown, the target error data identification module 13 may include two sub-modules: a bump error quantification sub-module 131 and an error type judgment sub-module 132. These two sub-modules are described below: (3.1) Bump error quantification submodule 131.

[0045] Optionally, the turbulence error quantization submodule 131 is used to determine the turbulence error quantization value of each candidate error data; wherein the turbulence error quantization value is used to characterize the probability that the candidate error data belongs to the error data caused by the turbulence of the UAV.

[0046] Specifically, the turbulence error quantification submodule 131 first determines the significance and dispersion values ​​of each candidate error data point. Then, based on the significance value, it determines the trend matching value for each candidate error data point. Finally, it calculates the quantified turbulence error value based on the significance, dispersion, and trend matching values. The significance value characterizes the significance of the deviation of the candidate error data within the corresponding preset flight path; the dispersion value characterizes the degree of clustering of the candidate error data within the ground altitude dataset; and the trend matching value characterizes the consistency between the changing trends of the candidate error data and the surrounding measurement point data.

[0047] (2.2) Error type judgment submodule 132.

[0048] Optionally, the error type determination submodule 132 is further configured to determine the candidate error data as the target error data if the quantized value of the bump error of the candidate error data is greater than or equal to a second preset threshold. For example, the second preset threshold can be set to an empirical value of 0.75.

[0049] Finally, the error type determination submodule 132 outputs the target error data set to the data correction module 14, and at the same time feeds back the non-target error data to the data storage submodule 113 for storage.

[0050] (4) Data correction module 14.

[0051] The data correction module 14 is responsible for correcting the target error data output by the target error data identification module 13, generating a corrected ground height dataset, and providing a precise data foundation for the image output module 15.

[0052] Specifically, the data correction module 14 first defines a neighborhood within a preset range in the two-dimensional plane dataset centered on each target error data, and extracts all non-target error data from this neighborhood as precise measurement point data.

[0053] Following this, the data correction module 14 constructs multiple valid line segments based on the determined precise measurement point data, with each line segment connecting two precise measurement point data points. For each line segment, the module calculates the vertical distance from the target error data to the line segment in a two-dimensional plane, as well as the projection ratio of the target error data onto the line segment. Then, based on the height data of the two endpoints of the line segment and the projection ratio, it calculates the estimated value of the actual data corresponding to that line segment. Simultaneously, based on the vertical distance, a preset weighting function determines the weight of the estimated value corresponding to each line segment.

[0054] Finally, the data correction module 14 performs a weighted calculation based on the estimated actual data values ​​and their weights corresponding to all line segments to obtain the corrected value for the target error data. This corrected value replaces the corresponding target error data in the original data and is integrated with the remaining non-error data to form the corrected ground height dataset. The corrected dataset is sent to the image output module 15 for mapping and is simultaneously stored in the data storage submodule 113 for later retrieval.

[0055] (5) Image output module 15.

[0056] The image output module 15 is responsible for outputting a practically valuable ground elevation map based on the corrected ground elevation dataset provided by the data correction module 14 through data processing and graphical workflow.

[0057] Specifically, the image output module 15 first converts the corrected ground elevation dataset into a two-dimensional plane dataset corresponding to the coordinates of the actual surveyed area. During this process, the module maps the elevation data of each measurement point to its corresponding two-dimensional coordinate position based on preset flight path coordinate information, thereby establishing a set of elevation data with spatial relationships.

[0058] Subsequently, the image output module 15 generates a contour map based on the two-dimensional planar dataset. The module uses a preset contour generation algorithm to interpolate the height data and draws contour lines with corresponding height values ​​according to the contour interval parameters set by the user, forming a contour map that accurately reflects the terrain undulation characteristics.

[0059] Finally, the image output module 15 performs coordinate alignment and overlay fusion of the generated contour map with the base map of the target area, ensuring accurate matching of the contour lines and the geographical location of the base map through a unified coordinate system. The overlaid ground elevation map can be transmitted to the user terminal via the UAV's wireless communication unit, or stored in the data storage submodule 113 for later retrieval.

[0060] It should be noted that, in addition to the four functional modules mentioned above, the low-error mapping UAV 10 also includes a basic flight control module and a power module (not shown in the figure). The flight control module is the flight hub of the low-error mapping UAV 10, primarily responsible for receiving and executing the flight path requirements of the core functional modules, while simultaneously adjusting the UAV's flight attitude in real time to adapt to the mapping operation. Specifically, the flight control module receives multiple preset flight paths output by the data acquisition module 11 (such as uniform flight paths planned for a square target area, ensuring that the flight altitude of each path remains constant), and drives the UAV to fly stably along the preset flight paths by controlling the UAV's servos, gyroscopes, and other components.

[0061] Furthermore, the power module provides the necessary power for the low-error mapping UAV 10 to fly and operate, and mainly consists of components such as batteries, motors, and propellers. Its core function is to cooperate with the flight control module to ensure that the UAV can continuously fly along the route planned by the data acquisition module 11. When the data acquisition module 11 performs ultrasonic measurements, the power module needs to maintain the UAV's stable flight altitude (to avoid altitude fluctuations affecting the calculation of ultrasonic propagation distance). At the same time, when the alternative error data screening module 12, the target error data identification module 13, and the data correction module 14 perform data processing, the power module needs to support the UAV's hovering or low-speed cruise to ensure that the UAV's position is stable during data processing and does not affect subsequent possible re-measurement operations. If the target error data identification module 13 identifies a large amount of target error data and re-acquisition is required, the power module needs to quickly respond to the re-measurement route command from the flight control module.

[0062] The above describes the low-error mapping UAV 10 and its included modules.

[0063] For example, such as Figure 4 The diagram shown is a flowchart illustrating a low-error mapping method according to an embodiment of the present invention, comprising the following steps: S401. Obtain the ground height dataset from the target area measured by the surveying UAV. The ground height dataset includes ground height data from multiple measurement points.

[0064] For example, this step can be performed by the data acquisition module 11 in the surveying drone 10 described above. It should be noted that during the measurement process, the surveying drone 10 measures the target area according to multiple preset flight paths, each flight path including multiple measurement points. The data acquisition module 11 performs the acquisition of the ground altitude dataset, specifically including the following steps: (1) Based on the terrain and scope of the target area, plan multiple preset routes.

[0065] Optionally, this sub-step is performed by the route planning sub-module 111 in the data acquisition module 11.

[0066] Specifically, the flight path planning submodule 111 obtains the boundary coordinates and mapping accuracy requirements of the target area based on user interaction, and generates multiple preset flight paths using appropriate algorithms, such as a grid-like uniform distribution algorithm. During the flight path generation process, the flight path planning submodule 111 ensures that each flight path is parallel and has consistent spacing, and each flight path includes multiple measurement points. The preset flight paths generated by the flight path planning submodule 111 are simultaneously sent to the flight control module (mentioned later), which then drives the UAV to fly along the preset flight paths.

[0067] (2) Measure the ground height data of the measurement points in each preset route.

[0068] Optionally, this sub-step is executed by the acquisition execution sub-module 112 in the data acquisition module 11.

[0069] Specifically, when the surveying drone flies to a certain measurement point, the data acquisition and execution submodule 112 collects ground altitude data at that measurement point. For example, the data acquisition and execution submodule 112 is equipped with an ultrasonic transceiver. This ultrasonic transceiver emits ultrasonic waves to the ground and records the time of ultrasonic wave emission and the time of receiving the ground reflection wave. The vertical distance between the drone and the ground measurement point is calculated using the formula: ultrasonic wave propagation speed × (reception time - emission time) / 2.

[0070] (3) Integrate and store the ground height data of multiple preset routes and each measurement point.

[0071] Optionally, this sub-step is performed by the data storage sub-module 113 in the data acquisition module 11.

[0072] Specifically, the data storage submodule 113 receives information related to multiple preset routes generated by the route planning submodule 111 and ground elevation data collected by the data collection and execution submodule 112. It then integrates this data into a two-dimensional ground elevation dataset according to the route number and measurement point number. Once all preset route measurements are completed, the data storage submodule 113 outputs the two-dimensional ground elevation dataset to the alternative error data filtering module 12.

[0073] S402. Select multiple candidate error data points from the ground elevation dataset. These candidate error data points are measurement point data points that deviate from the normal fluctuation range.

[0074] For example, this step can be performed by the candidate error data filtering module 12 in the surveying UAV described above, specifically including: first, determining the deviation value of each measurement point data in the ground altitude dataset; then, if the deviation value of the measurement point data is greater than or equal to a first preset threshold, determining the measurement point data as candidate error data. The deviation value is used to characterize the magnitude of the local deviation between the measurement point data and adjacent measurement point data. It should be noted that the specific process of the candidate error data filtering module 12 identifying target error data according to the aforementioned sub-steps can be found in S501-S502 below, and will not be repeated here.

[0075] In another possible implementation, the alternative error data filtering module 12 can use the sliding window variance method to calculate the deviation degree value. This method involves defining a sliding window of a preset size for each measurement point data and calculating the variance of the data within the window as the deviation degree value. Alternatively, it can use the local extreme value comparison method, which determines the deviation degree value by judging whether the measurement point data is a local extreme value and calculating the difference between its extreme value and the surrounding data.

[0076] S403. Identify multiple target error data from multiple candidate error data. Among them, the target error data are error data caused by the turbulence of the surveying UAV.

[0077] For example, this step can be performed by the target error data identification module 13 in the mapping UAV described above, specifically including: first, determining the turbulence error quantization value of each candidate error data; wherein, the turbulence error quantization value is used to characterize the probability that the candidate error data belongs to the error data caused by the turbulence of the UAV; finally, if the turbulence error quantization value of the candidate error data is greater than or equal to a second preset threshold, the candidate error data is determined as the target error data. It should be noted that the specific process of the target error data identification module 13 in identifying the target error data according to the aforementioned sub-steps can be found in S601-S602 below, and will not be repeated here.

[0078] In another possible implementation, in determining the turbulence error quantification value, the target error data identification module 13 can also use the flight state-assisted method to calculate the turbulence error quantification value by introducing the flight state data fed back by the UAV flight control module and weighting and fusing it with parameters such as the significance value; the multi-route cross-validation method extracts the measurement point data near the same position in adjacent routes, calculates the deviation consistency and uses it as a new dimension to participate in the fusion calculation, thereby improving the accuracy of the judgment.

[0079] Therefore, based on S402-S403, this embodiment of the invention can effectively distinguish between the actual altitude change caused by terrain undulation and the measurement error caused by the turbulence of the UAV by setting a two-level screening mechanism of deviation degree value and turbulence error quantification value.

[0080] S404. Correct the target error data to obtain the corrected ground height dataset.

[0081] For example, this step can be performed by the data correction module 14 in the mapping UAV 10 described above, and specifically includes the following steps: (1) For each target error data, acquire multiple precise measurement point data in a preset two-dimensional neighborhood around the target error data in the two-dimensional plane dataset.

[0082] The two-dimensional plane dataset is obtained by converting the ground height dataset.

[0083] For example, for the first For each target error data, the data correction module 14 obtains its surrounding 24 neighborhoods on the two-dimensional plane dataset. The window contains all measurement point data, from which the precise data that does not belong to the target error data is filtered out and used as multiple precise measurement point data.

[0084] (2) Construct multiple line segments based on data from multiple precise measurement points. Each line segment connects two precise measurement points.

[0085] In this step, the data correction module 14 connects the multiple precise measurement point data obtained above in pairs to obtain several line segments, that is, constructs multiple line segments.

[0086] (3) For each line segment, calculate the vertical distance from the target error data to the line segment in the two-dimensional plane, and the projection ratio of the target error data onto the line segment.

[0087] Optionally, the data correction module 14 selects line segments from the constructed multiple line segments that satisfy the following geometric conditions: there exists a point on the line containing the line segment such that the line connecting the target error data point (i.e., the measurement point corresponding to the target error data) and this point is perpendicular to the line segment; and the foot of the perpendicular from the target error data point to the line segment lies between the two endpoints of the line segment (i.e., the foot of the perpendicular lies on the line segment, not its extension). The total number of selected line segments is recorded as follows: .

[0088] For the selected first A line segment, with the data values ​​of its two endpoints recorded as follows: as well as Draw a straight line perpendicular to the first target error data point. Line segments, and calculate the target error data point to the first line segment. The perpendicular distance between the two line segments is denoted as . .

[0089] At the same time, record the above-mentioned line perpendicular to the first... The foot of the perpendicular when there are two line segments, the foot of the perpendicular will be the first... The line segment is divided into two parts. Calculate the distance from the foot of the perpendicular to the endpoint with the smaller height value (i.e.,...). The length of the line segment (corresponding endpoints) and the length of the line segment of the first endpoint. The ratio of the total length of the line segments is denoted as [the ratio of the lengths of the line segments]. That is, the projection ratio of the target error data onto the line segment.

[0090] (4) For each line segment, calculate the estimated actual data value corresponding to the line segment based on the projection scale and the data values ​​corresponding to the two endpoints of the line segment.

[0091] For example, the data correction module 14 adjusts the data values ​​of the two endpoints of the line segment and the position of the target error data point relative to the line segment (i.e., ), which can be calculated from the first The actual data estimate of the line segment .

[0092] (5) Calculate the correction value of the target error data based on the estimated value of the actual data corresponding to all line segments and the weight of the estimated value. The weight of the estimated value is determined based on the vertical distance.

[0093] Furthermore, the data correction module 14 first calculates the average of all vertical distances as a normalization factor, denoted as . It can be calculated using the following formula: in, Indicates the first The target error data is mapped to any one of the multiple line segments. vertical distance, This indicates the total number of line segments.

[0094] Following this, according to the The distance from each line segment to the target error data point can be used to derive the first... The reference value of the estimated value of the actual data from the line segment to the target error data point, i.e., the weight of the estimated value. Specifically, it is calculated using the following formula: In the above formula, Indicates the first The estimated weights of the line segments, Indicates the first The target error data is transferred to the first... perpendicular distance between line segments This indicates the total number of line segments. Indicates the first The target error data is mapped to any one of the multiple line segments. vertical distance, This represents the inverse proportional normalized function. Through normalization, the input to the exponential function becomes dimensionless, conforming to mathematical norms. It should be noted that, because... The range of the function does not include 0, and Since the denominator is greater than 0, the above formula will not result in a situation where the calculation is meaningless due to the denominator being zero.

[0095] Finally, the data correction module 14 calculates the correction value for the target error data based on the actual data estimates and estimate weights corresponding to all line segments, denoted as . , The calculation formula is as follows: In the above formula, Indicates the first Correction values ​​for each target error data, Indicates the first The estimated actual data value of the line segment. Indicates the first Weights of the estimated values ​​of line segments.

[0096] In summary, according to the first Accurate data surrounding the target error data, for the first target error data, The correction value is obtained by estimating the actual data of each target error. Use this correction value for the first The target error data is recovered. This process is repeated for all target error data to obtain the corrected ground height dataset.

[0097] In another possible implementation, when correcting the target error data, the data correction module 13 can also use an interpolation correction method based on accurate neighborhood data. This involves extracting accurate measurement point data within a preset two-dimensional neighborhood surrounding the target error data, constructing a continuous height model using an interpolation algorithm that conforms to terrain features, and then substituting the spatial coordinates of the target error data into this model to calculate the correction value. This method is suitable for scenarios with gentle terrain changes and uniform distribution of accurate data, and can effectively reflect the overall height trend of the neighborhood.

[0098] Alternatively, data correction module 13 can use a terrain trend-based fitting correction rule to fit the precise neighborhood data into a mathematical model that conforms to the terrain characteristics through least squares fitting algorithms. After verifying the model's accuracy, the spatial coordinates of the target error data are substituted into the fitted model to obtain the correction value. This method is suitable for areas with obvious overall terrain trends and can effectively eliminate the influence of local disturbances.

[0099] Therefore, by utilizing the spatial continuity characteristics of the precise measurement point data in the neighborhood surrounding the target error data, the data correction module 13 constructs a correction algorithm based on geometric relationships or statistical models, which can accurately recover the altitude data distorted by the instantaneous turbulence of the UAV.

[0100] S405. Determine the ground elevation map based on the corrected ground elevation dataset.

[0101] For example, this step can be performed by the image output module 15 in the mapping UAV 10 described above. Specifically, the image output module 15 first converts the corrected ground height dataset into a two-dimensional plane dataset corresponding to the coordinates of the actual mapping area. In this process, the module maps the height data of each measurement point to the corresponding two-dimensional coordinate position according to the preset flight path coordinate information, thereby establishing a height data set with spatial positional relationships.

[0102] Subsequently, the image output module 15 generates a contour map based on the two-dimensional planar dataset. The module uses a preset contour generation algorithm to interpolate the height data and draws contour lines with corresponding height values ​​according to the contour interval parameters set by the user, forming a contour map that accurately reflects the terrain undulation characteristics.

[0103] Finally, the image output module 15 performs coordinate alignment and overlay fusion of the generated contour map with the base map of the target area, ensuring accurate matching of the contour lines and the geographical location of the base map through a unified coordinate system. The overlaid ground elevation map can be transmitted to the user terminal via the UAV's wireless communication unit, or stored in the data storage submodule 113 for later retrieval.

[0104] Based on the above technical solution, the embodiments of the present invention construct a multi-level processing flow that includes deviation degree value calculation, candidate error data screening, turbulence error quantification value determination, and target error data correction. This enables accurate identification and effective correction of error data caused by turbulence during UAV surveying, significantly improving the accuracy and reliability of ground altitude datasets. At the same time, while ensuring surveying accuracy, it greatly reduces the amount of repetitive surveying work, comprehensively improving surveying efficiency and result quality under complex terrain conditions.

[0105] For example, in combination Figure 4 ,like Figure 5The diagram shown illustrates a flowchart of another low-error mapping method provided by an embodiment of the present invention. In this method, multiple candidate error data are selected from the ground elevation dataset, specifically including the following steps: S501. For each measurement point in the ground height dataset, determine the deviation value of the measurement point data. The deviation value characterizes the magnitude of the local deviation between the measurement point data and the data of adjacent measurement points.

[0106] For example, the alternative error data filtering module 12 determines the deviation value of the measurement point data, specifically including: (1) For each measurement point in the ground height dataset, determine the first difference value and the second difference value between the measurement point data and the adjacent measurement point data. The first difference value is the difference between the measurement point data and the previous adjacent measurement point data, and the second difference value is the difference between the measurement point data and the next adjacent measurement point data.

[0107] In this step, the ground height dataset obtained in S401 is denoted as... ,in Indicates the number of preset routes. This indicates the number of measurement points on each route.

[0108] Specifically, the alternative error data filtering module 12 processes the measurement point data corresponding to each measurement point in the ground height dataset and calculates the first difference value and the second difference value: for the first... The first of the routes Data from measurement points In contrast, using the difference method, the difference between a given data point and its previous neighbor is calculated and denoted as . (That is, the first difference value), the difference between it and the next adjacent data is denoted as (That is, the second difference value).

[0109] (2) Determine the degree of deviation of the measurement point data based on the first difference value and the second difference value.

[0110] Optionally, the alternative error data filtering module 12 calculates the deviation value using the following formula: In the above formula, Represents measurement point data The degree of deviation, Indicates the first difference value. Indicates the second difference value. This represents the logistic function. When... When it approaches 1, it indicates If the local deviation from the data at adjacent measurement points is extremely large, it is highly likely to be error data; while when When it approaches 0, it indicates The local deviation from the data of adjacent measurement points is extremely small, and it is highly likely that the data is normal.

[0111] It should be noted that the above formula uses the absolute value of the sum of the first difference value and the second difference value as the core metric when calculating the deviation of the measurement point data. This calculation method is based on the characteristics of real terrain: real terrain has spatial continuity, and its height data changes gradually; while the error caused by the drone's turbulence is isolated and abrupt.

[0112] Specifically, the deviation value essentially reflects the curvature intensity of the local sequence formed by the measurement point and its neighboring points. For continuous inflection points of real terrain (such as smooth ridges or valleys), the first and second difference values ​​are usually opposite in sign (e.g., the former is positive and the latter is negative). When summed, positive and negative cancellation effects occur, resulting in a smaller calculated result, which is then identified as a normal terrain feature and preserved.

[0113] Conversely, isolated error points caused by random turbulence of the drone will be abnormally high or low, while the adjacent points remain normal terrain points. In this case, the first and second difference values ​​usually have the same sign (for example, for a raised error point, the difference values ​​before and after it are both positive), and the absolute value after summing will increase significantly, thus effectively identifying it as potential error data.

[0114] Therefore, the strategy for calculating the degree of deviation cleverly utilizes the positive and negative cancellation phenomenon in the difference summation process to achieve efficient and preliminary differentiation between normal inflection points that conform to the trend of terrain change and isolated error points that do not conform to continuity, laying a reliable foundation for subsequent fine screening. S502, If the degree of deviation of the measurement point data is greater than or equal to the first preset threshold, the measurement point data is determined as candidate error data.

[0115] For example, the first preset threshold can be set to an empirical value of 0.7.

[0116] Alternatively, the first preset threshold can be determined through analysis of historical surveying data. Specifically, ground elevation datasets under different terrain conditions from historical operations can be collected, the deviation values ​​of each measurement point in each dataset can be calculated, and the distribution characteristics of these deviation values ​​can be statistically analyzed. Based on the distribution characteristics, the first preset threshold can be set to a specific percentile value, such as the 85th or 90th percentile, to ensure that measurement point data that significantly deviates from the normal fluctuation range can be effectively filtered out. This setting method based on historical data statistical analysis allows the threshold to better adapt to different surveying environments and terrain features, improving the adaptability and accuracy of the error identification process.

[0117] Based on the above technical solution, this embodiment of the invention determines the degree of deviation by calculating the first and second difference values ​​between each measurement point's data and its adjacent data, and filters candidate error data based on a first preset threshold, thus constructing an efficient and reliable preliminary error data identification mechanism. This solution can accurately capture local data anomalies caused by UAV turbulence and effectively eliminate interference from normal terrain undulations, laying a solid foundation for subsequent accurate identification of target error data and significantly improving the accuracy and efficiency of error data identification.

[0118] For example, in combination Figure 4 ,like Figure 6 The diagram shown illustrates a flowchart of another low-error mapping method provided by an embodiment of the present invention. In this method, multiple target error data are identified from multiple candidate error data, specifically including the following steps: S601. Determine the turbulence error quantization value for each candidate error data. The turbulence error quantization value characterizes the probability that the candidate error data belongs to the error data caused by UAV turbulence.

[0119] Understandably, executing S502 yields some data with significant deviations (i.e., candidate error data) as potential target error data. However, if a large deviation is found in the data, it could be due to drone turbulence, or it could be due to the complexity and unevenness of the measured terrain itself. Therefore, a large deviation in the data is only a necessary but not sufficient condition for drone turbulence to cause error data. It is not sufficient to conclude that the data is due to drone turbulence solely based on the presence of deviation. This step also requires calculating a quantified value of the turbulence error to provide more accurate evidence for subsequent steps in determining whether the candidate error data is indeed the target error data.

[0120] For example, this step can be performed by the turbulence error quantification submodule 131 in the target error data identification module 13, specifically including: S6011. Determine the significance value of each alternative error data.

[0121] In this step, the turbulence error quantification submodule 131 calculates the significance value of each candidate error data, which is used to characterize the significance of the deviation of the candidate error data within the corresponding preset route.

[0122] Specifically, the turbulence error quantification submodule 131 will... The data values ​​of the candidate error data are denoted as follows: The adjacent deviation data on the left and right sides of the same preset route are respectively recorded as and .in For route number, These represent the numbers of three adjacent measurement points on the flight path, with the following order of importance: The following steps (1) to (3) will be used to explain the calculation process of the significance value: (1) The turbulence error quantification submodule 131 calculates a third difference value for each candidate error data. The third difference value is the absolute value of the sum of the differences between the candidate error data and its left and right adjacent candidate error data. For example, the third difference value is expressed as... .

[0123] It should be noted that the third difference value is used to characterize the degree of local abruptness of the candidate error data among all candidate error data within the corresponding route; the larger the value, the more prominent the point is compared with other outliers, and the more likely it is to originate from random turbulence rather than continuous terrain changes.

[0124] (2) The bump error quantification submodule 131 calculates the fourth difference value; wherein, the fourth difference value is the absolute value of the difference between the candidate error data and the average value of the measurement point data in the local window, and the local window is a dataset of continuous measurement points within a preset area centered on the candidate error data.

[0125] Optionally, the bump error quantization submodule 131 quantifies the bump error based on the measurement points. and measurement points From the flight route A local dataset is delineated from the corresponding ground elevation data. This refers to the aforementioned local window. The turbulence error quantification submodule 131 then calculates the fourth difference value based on the local window, specifically using the following formula: In the above formula, Represents a local window The average value of the data. This represents the fourth difference value.

[0126] (3) Finally, the bump error quantification submodule 131 determines the significance value based on the third difference value, the fourth difference value, and the number of measurement points within the local window. For example, the bump error quantification submodule 131 can calculate the significance value using the following formula: In the above formula, Indicates the first The significance value of each alternative error data point This indicates the number of measurement points within the local window. This represents an adjustment factor, a very small positive value, such as 10 to the power of negative 5, used to prevent the denominator from being zero. The numerator in the formula uses... and The geometric mean (i.e.) This geometric mean is used to characterize the overall bias of the candidate error data within a local window, and it has the same characteristics as... , Same unit of length (meter). Denominator The number of dimensionless measurement points is used to reflect the data size of a local window. Therefore, the significance value... It has the dimension of length (meter), and the larger its value, the longer the length of the first unit. The greater the degree of local deviation of each candidate error data, the greater the degree of local deviation.

[0127] It should be noted that, in conjunction with the example in step (2) above, Specifically, The span of the local window used to calculate the fourth difference value is defined. The smaller the value, the more significantly the candidate error data appears relative to its neighboring data within a shorter measurement distance. This increases the likelihood that the error is caused by random fluctuations, thus making the significance level more pronounced. The larger the calculation result, the better.

[0128] Therefore, the bump error quantification submodule 131 can determine the significance value of each candidate error data according to the above three sub-steps.

[0129] S6012. Determine the dispersion value for each candidate error data. The dispersion value characterizes the degree of clustering of the candidate error data within the ground elevation dataset.

[0130] Understandably, considering the randomness and instantaneity of turbulence during UAV flight, the resulting error data typically exhibits isolated or scattered distribution characteristics; while data changes caused by terrain undulations, due to their spatial continuity, often show a trend of continuous distribution within a certain area. Based on these differences, by evaluating the dispersion of candidate error data in the ground altitude dataset, it is possible to effectively distinguish between the two different types of data bias.

[0131] Specifically, the assessment of dispersion needs to be performed in two-dimensional space. By constructing a ground altitude dataset containing multiple flight paths, the distribution and clustering characteristics of each candidate error data in planar space are analyzed, thus providing an important basis for accurately identifying error data caused by UAV turbulence. The calculation process of the significance value is explained in detail below through sub-steps (a) to (b): (a) For each candidate error data, the neighborhood range is expanded with the candidate error data as the center according to a preset step size until there are no other candidate error data at the boundary of the neighborhood range, or the expanded neighborhood radius reaches the preset maximum search radius (e.g., 10 grid units). The maximum search radius can be preset according to the size of the survey area and the characteristics of the error distribution to ensure that the dispersion value can effectively characterize the local clustering characteristics, while avoiding excessive computation.

[0132] In this step, the turbulence error quantization submodule 131 will quantify the ground height dataset. ,by For action, Consider the column as a two-dimensional planar dataset, and label the candidate error data within it. The total number of labeled candidate error data is denoted as . indivual.

[0133] right The first of the candidate error data The candidate error data is defined by setting its surrounding area as the neighborhood range, specifically as follows: Corresponding to the previous example, the first... The candidate error data are denoted as Then the circle around it is arrive The square region is the diagonal area, and this square region is the first... The neighborhood range of each alternative error data.

[0134] Next, the first Each candidate error data point and its neighboring candidate error data are considered as a single data unit. Then, according to a preset step size, the candidate error data within the neighboring range of this single data unit are merged back into the single data unit, and this process continues until there are no candidate error data points within the neighboring range of the current single data unit, or the expanded neighborhood radius reaches the preset maximum search radius. For example, the preset step size is set to 1. In practical applications, it can be dynamically adjusted according to the size of the data. This embodiment of the invention does not impose specific limitations.

[0135] (b) The total number of candidate error data within the neighborhood range is used as the dispersion value.

[0136] In this step, referring to the example in step (a), the turbulence error quantification submodule 131 records the total number of candidate error data within the current data set until there are no candidate error data in the neighborhood of the current data set. This number is used as the first... Dispersion value of each candidate error data .

[0137] Therefore, the turbulence error quantification submodule 131 can determine the dispersion value of each candidate error data according to the above two sub-steps.

[0138] S6013. Determine the turbulence error quantification value for each candidate error data based on the deviation degree value, significance value, and dispersion value of each candidate error data.

[0139] For example, the turbulence error quantification submodule 131 determines the turbulence error quantification value of each candidate error data based on the significance value and dispersion value of each candidate error data, specifically including the following steps (I) and (II): (I) Determine the trend matching degree value for each candidate error data based on the significance value of each candidate error data. The trend matching degree value is used to characterize the degree of consistency between the changing trend of the candidate error data and the data of surrounding measurement points.

[0140] Optionally, the bump error quantization submodule 131 first calculates the first... The gradient of each candidate error data point with its eight adjacent measurement points on the two-dimensional plane is used to obtain the first... The unit vector of the gradient direction at each candidate error data point. Average gradient of data from the other eight neighboring measurement points The angle between the two vectors is denoted as . .

[0141] Furthermore, the turbulence error quantification submodule 131 calculates the first... The trend matching degree value of the candidate error data: In the above formula, Indicates the first The degree of trend matching of the candidate error data; and These represent the current gradient magnitude and the surrounding average gradient magnitude, respectively. This represents the angle between the two vectors mentioned above. This is a measure of the consistency of the directions of two vectors. The first term (magnitude ratio) in the formula ranges from [0, 1], reflecting the similarity of their magnitudes; the second term (normalized cosine value) also ranges from [0, 1], reflecting the consistency of their directions. Therefore, The value range is [0, 1]. The larger the value, the more consistent the terrain change trend between the current point and the surrounding points is, and the lower the possibility of error caused by drone turbulence.

[0142] (II) Calculate the turbulence error quantification value for each candidate error data based on the deviation value, significance value, dispersion value and trend matching value of each candidate error data.

[0143] Optionally, the bump error quantization submodule 131 calculates the bump error quantization value using the following formula: In the above formula, Indicates the first The quantification value of the turbulence error of the candidate error data. Indicates the first The degree of deviation of each alternative error data. Indicates the first The degree of trend matching among the candidate error data. Indicates the first The dispersion value of each candidate error data. Indicates the first The significance value of each alternative error data point This represents the logistic function. This represents the adjustment factor, which is a very small positive value. For example, it can be 10 to the power of negative 5, used to prevent the denominator from being zero.

[0144] It is not difficult to see that in the above formula, The larger the value, The larger the value, the more likely it is that the first... The more likely the alternative error data is to be caused by drone turbulence; and , The larger the value, the better. The smaller the value, the more likely it is that the first... The fewer alternative error data points there are, the less likely they are to be caused by drone turbulence. Therefore, in the formula... right The value is adjusted to obtain the final value. Quantitative value of bump error for each candidate error data .

[0145] S602. If the quantized value of the bump error of the candidate error data is greater than or equal to the second preset threshold, the candidate error data shall be determined as the target error data.

[0146] For example, the second preset threshold can be set to an empirical value of 0.75.

[0147] Alternatively, the second preset threshold can be determined by analyzing confirmed error data samples from historical mapping data, calculating the statistical characteristics of the turbulence error quantification values ​​corresponding to these samples, and setting the second preset threshold at a critical level that can effectively distinguish error data from normal terrain undulation data. Another implementation method is to train a historical dataset using a preset classification model, and dynamically adjust the second preset threshold based on the probability distribution characteristics output by the model to adapt to the recognition needs under different terrain conditions. Furthermore, the second preset threshold can be adaptively adjusted based on the fluctuation characteristics of real-time collected data and the UAV flight status parameters to improve the accuracy of error recognition and environmental adaptability.

[0148] Furthermore, the error type determination submodule 132 quantizes the bump error value calculated by S601. If the quantified value of the bump error of the candidate error data is greater than or equal to the second preset threshold, the candidate error data is determined as the target error data.

[0149] Based on the above technical solution, this invention constructs a multi-dimensional evaluation system including deviation value, significance value, dispersion value, and trend conformity value, and employs a two-level screening mechanism to accurately identify error data caused by UAV turbulence. This solution effectively overcomes the difficulty of distinguishing between real terrain undulations and measurement errors under complex terrain conditions, significantly improving the accuracy and reliability of error data identification. Simultaneously, while ensuring data quality, it greatly reduces the amount of repetitive surveying work caused by misjudgments, thereby comprehensively improving the overall efficiency and accuracy of surveying operations.

[0150] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0151] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A low-error surveying method, characterized in that, The method includes: Obtain a ground height dataset from a target area measured by a surveying drone; wherein the ground height dataset includes ground height data from multiple measurement points; Multiple candidate error data are selected from the ground height dataset; wherein, the candidate error data are measurement point data that deviate from the normal fluctuation range; Multiple target error data are identified from the multiple candidate error data; wherein, the target error data are the error data caused by the turbulence of the mapping UAV among the multiple candidate error data; The target error data is corrected to obtain the corrected ground height dataset; A ground elevation map is determined based on the corrected ground elevation dataset.

2. The low-error mapping method according to claim 1, characterized in that, During the measurement process, the surveying UAV measures the target area according to multiple preset routes, each preset route including multiple measurement points; Multiple candidate error data points were selected from the ground elevation dataset, specifically including: For each measurement point in the ground height dataset, a deviation value is determined for the measurement point data; wherein, the deviation value is used to characterize the magnitude of the local deviation between the measurement point data and the data of adjacent measurement points; If the deviation value of the measurement point data is greater than or equal to a first preset threshold, the measurement point data is determined as candidate error data.

3. The low-error mapping method according to claim 2, characterized in that, Determining the degree of deviation of the measurement point data specifically includes: For each measurement point in the ground height dataset, a first difference value and a second difference value are determined between the measurement point data and the adjacent measurement point data; wherein, the first difference value is the difference between the measurement point data and the previous adjacent measurement point data, and the second difference value is the difference between the measurement point data and the next adjacent measurement point data; The degree of deviation of the measurement point data is determined based on the first difference value and the second difference value.

4. The low-error mapping method according to claim 1, characterized in that, Multiple target error data are identified from the plurality of candidate error data, specifically including: Determine the turbulence error quantization value for each of the candidate error data; wherein the turbulence error quantization value is used to characterize the probability that the candidate error data belongs to the error data caused by the turbulence of the UAV; If the quantized value of the bump error of the candidate error data is greater than or equal to the second preset threshold, the candidate error data is determined as the target error data.

5. The low-error mapping method according to claim 4, characterized in that, Determining the turbulence error quantization value for each of the candidate error data specifically includes: Determine the significance value and dispersion value of each candidate error data; wherein, the significance value is used to characterize the deviation significance of the candidate error data within the corresponding preset flight path, and the dispersion value is used to characterize the degree of clustering of the candidate error data in the ground altitude dataset; Based on the deviation value, significance value, and dispersion value of each candidate error data, a quantified value of the turbulence error is determined for each candidate error data.

6. The low-error mapping method according to claim 5, characterized in that, Determining the significance value of each of the candidate error data points specifically includes: For each of the candidate error data, a third difference value is calculated; wherein the third difference value is the absolute value of the sum of the differences between the candidate error data and its left and right adjacent candidate error data; Calculate the fourth difference value; wherein the fourth difference value is the absolute value of the difference between the candidate error data and the average value of the measurement point data within the local window, and the local window is a dataset of continuous measurement points within a preset area centered on the candidate error data; The significance value is determined based on the third difference value, the fourth difference value, and the number of measurement points within the local window.

7. The low-error mapping method according to claim 5, characterized in that, Determining the dispersion value of each of the candidate error data specifically includes: For each of the candidate error data, the neighborhood range is expanded according to a preset step size, with the candidate error data as the center, until there are no other candidate error data at the boundary of the neighborhood range, or the expanded neighborhood radius reaches the preset maximum search radius. The total number of candidate error data within the neighborhood range is counted and used as the dispersion value.

8. The low-error mapping method according to claim 5, characterized in that, Based on the significance and dispersion values ​​of each candidate error data point, a quantized value for the turbulence error of each candidate error data point is determined, specifically including: Based on the significance value of each candidate error data, a trend matching degree value is determined for each candidate error data; wherein, the trend matching degree value is used to characterize the degree of consistency between the changing trend of the candidate error data and the data of surrounding measurement points; Based on the deviation value, significance value, dispersion value, and trend matching value of each candidate error data, calculate the quantified value of the turbulence error for each candidate error data.

9. The low-error surveying method according to any one of claims 1-8, characterized in that, The target error data is corrected by performing data correction, specifically including: For each target error data, multiple precise measurement point data within a preset two-dimensional neighborhood surrounding the target error data are acquired in a two-dimensional plane dataset; wherein, the two-dimensional plane dataset is obtained by converting the ground height dataset; Based on the data from the multiple precise measurement points, multiple line segments are constructed; wherein each line segment connects two precise measurement point data points. For each line segment, calculate the vertical distance from the target error data to the line segment in a two-dimensional plane, as well as the projection ratio of the target error data onto the line segment; For each line segment, the estimated actual data value corresponding to the line segment is calculated based on the projection ratio and the data values ​​corresponding to the two endpoints of the line segment. The correction value of the target error data is calculated based on the estimated values ​​of the actual data corresponding to all line segments and the weights of the estimated values; wherein, the weights of the estimated values ​​are determined based on the vertical distance.

10. A low-error mapping UAV, applied to the low-error mapping method as described in any one of claims 1-9, characterized in that, The low-error mapping UAV includes: a data acquisition module, a candidate error data filtering module, a target error data identification module, a data correction module, and an image output module; The data acquisition module is used to acquire a ground height dataset obtained by a surveying drone measuring a target area; wherein, the ground height dataset includes ground height data from multiple measurement points; The alternative error data filtering module is used to filter out multiple alternative error data from the ground height dataset; wherein, the alternative error data are measurement point data that deviate from the normal fluctuation range; The target error data identification module identifies multiple target error data from the multiple candidate error data; wherein, the target error data is the error data caused by the turbulence of the mapping UAV among the multiple candidate error data; The data correction module is used to correct the target error data to obtain a corrected ground height dataset. The image output module is used to determine a ground height map based on the corrected ground height dataset.

Images