An intelligent navigation method, device and equipment of a UAV and a storage medium

Abstract

The application discloses an intelligent navigation method and device of a UAV, equipment and a storage medium. The method comprises the following steps: acquiring point cloud data and motion data, calibrating the relative position and time synchronization parameters; constructing a map by using the point cloud data and the motion data, and generating a three-dimensional point cloud map; analyzing flight area boundary coordinates, flight height and route step length, and generating route planning parameters; defining the target inspection area boundary in combination with the three-dimensional point cloud map, generating an optimized route and corresponding waypoint sequence data through route planning; acquiring real-time pose data of the UAV, comparing the current position with the waypoint sequence, and triggering image acquisition when the UAV is in the preset waypoint arrival tolerance range; obtaining inspection data; integrating the real-time pose data, the three-dimensional point cloud map, the flight state and the inspection data, and forming an intelligent navigation route. The application realizes high-precision positioning, linkage of waypoints and data acquisition, and improves the navigation accuracy, inspection efficiency and operation safety.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) navigation technology, and in particular to an intelligent navigation method, device, equipment, and storage medium for UAVs. Background Technology

[0002] With the continuous advancement of transportation infrastructure construction, bridges, as key hubs, are increasingly characterized by structural complexity and variability in their operating environment. Manual bridge inspection suffers from high risks, low efficiency, and incomplete coverage of details, making it difficult to meet the needs of modern operation and maintenance. Therefore, drone inspection is gradually becoming the mainstream choice for bridge inspection.

[0003] However, existing drones still face many bottlenecks in navigation and inspection in complex environments such as bridges. In terms of positioning, traditional drones generally rely on single navigation and positioning systems. When flying under bridge steel structures, across river sections, or in urban canyons, signals are easily blocked and interfered with by multipath effects, leading to decreased or even lost positioning accuracy, directly threatening flight safety. Regarding system coordination, collaborative operation efficiency is low. At the task execution level, the level of intelligence is insufficient. In waypoint planning, it largely relies on manual settings, which is not only inefficient but also prone to missing key inspection points, affecting the completeness and reliability of the inspection task.

[0004] Current UAV navigation methods struggle to simultaneously address issues such as positioning accuracy, intelligent route planning, automated task execution, and efficient system collaboration, resulting in compromised safety, efficiency, and data integrity in inspection operations. Therefore, achieving high-precision positioning, intelligent route generation, and automated task coordination for UAVs in complex environments, thereby improving navigation reliability and inspection operation quality, has become a pressing technical challenge. Summary of the Invention

[0005] In view of this, the intelligent navigation method, device, equipment, and storage medium for unmanned aerial vehicles (UAVs) provided in this application embodiment can achieve high-precision positioning, waypoint and data acquisition linkage, and improve navigation accuracy, inspection efficiency, and operational safety. The intelligent navigation method, device, equipment, and storage medium for UAVs provided in this application embodiment are implemented as follows:

[0006] This application provides an intelligent navigation method for unmanned aerial vehicles (UAVs), comprising:

[0007] Acquire point cloud data and motion data, perform calibration processing on the point cloud data and motion data, and obtain relative position and time synchronization parameters;

[0008] Based on the relative position and time synchronization parameters, map construction processing is performed on the point cloud data and the motion data to obtain a three-dimensional point cloud map.

[0009] The flight area boundary coordinates, flight altitude, and flight path step size are obtained. The flight area boundary coordinates, flight altitude, and flight path step size are analyzed to obtain the flight path planning parameters.

[0010] Based on the three-dimensional point cloud map, the boundary of the target inspection area is determined, and the route planning parameters and the boundary of the target inspection area are processed to obtain optimized route data and corresponding waypoint sequence data.

[0011] The real-time pose data of the UAV is acquired, and the current position and waypoint sequence data of the UAV are compared and processed based on the real-time pose data. When the current position of the UAV is within the tolerance range of the preset waypoint, image acquisition is performed to obtain image data. The real-time pose data includes the current waypoint position data and time data.

[0012] The current waypoint location data, the time data, and the image data are correlated to obtain inspection data;

[0013] The real-time pose data, the 3D point cloud map, the current flight status of the UAV, and the inspection data are integrated and processed to obtain an intelligent navigation route.

[0014] In some embodiments, the route planning parameters include path interval and flight altitude. The process of determining the target inspection area boundary based on the 3D point cloud map, and performing route planning processing on the route planning parameters and the target inspection area boundary to obtain optimized route data and corresponding waypoint sequence data includes:

[0015] Based on the three-dimensional point cloud map, the boundary of the initial inspection area is corrected to obtain the boundary of the target inspection area.

[0016] The path layout is obtained by processing the path interval, the flight altitude, and the boundary of the target inspection area.

[0017] The path layout is optimized to obtain an optimized path layout.

[0018] Cruise points are generated based on the path layout and the optimized direction of the path, and an initial waypoint sequence is constructed based on the cruise points;

[0019] The initial waypoint sequence and the optimized path layout are integrated to obtain optimized route data and corresponding waypoint sequence data.

[0020] In some embodiments, the real-time pose data of the UAV is acquired, and the current position and waypoint sequence data of the UAV are compared based on the real-time pose data. When the current position of the UAV is within a preset waypoint tolerance range, image acquisition is performed to obtain image data, including:

[0021] The point cloud data, the motion data, and the 3D point cloud map are fused to obtain the real-time pose data.

[0022] The current position in the real-time pose data is compared with the waypoint sequence data to obtain position deviation data;

[0023] The position deviation data is compared with a preset waypoint tolerance threshold. When the position deviation data is less than or equal to the preset waypoint tolerance threshold, a waypoint arrival confirmation signal is obtained.

[0024] Based on the waypoint arrival confirmation signal, the UAV is controlled to acquire images and obtain image data of the target area.

[0025] In some embodiments, the path layout processing of the path interval, the flight altitude, and the target inspection area boundary to obtain the path layout includes:

[0026] The boundary of the target inspection area is closed to obtain an effective inspection area;

[0027] The effective inspection area is subjected to benchmark fitting processing to obtain the circumscribed benchmark area;

[0028] Based on the path interval, the external reference region is processed to generate scan lines to obtain parallel scan lines;

[0029] Based on the boundary of the target inspection area, the parallel scan lines are subjected to orientation adaptation processing to obtain the processed parallel scan lines.

[0030] The parallel scan lines and the flight altitude are processed to obtain the path layout in three-dimensional space.

[0031] In some embodiments, generating cruise points based on the path layout and the optimized direction of the path, and constructing an initial waypoint sequence based on the cruise points, includes:

[0032] The intersection of the parallel scan line and the boundary of the target inspection area is calculated to obtain the intersection point data.

[0033] The intersection data is sorted to obtain ordered intersection data;

[0034] The ordered intersection data is subjected to adjacent pairing processing to obtain valid flight segment data;

[0035] Based on the path interval, the effective flight segment data is interpolated at equal intervals to obtain two-dimensional cruise point data;

[0036] Based on the flight altitude, the two-dimensional cruise point data is subjected to three-dimensional coordinate completion processing to obtain three-dimensional cruise point data;

[0037] The three-dimensional cruise point data is sequentially arranged to obtain an initial waypoint sequence.

[0038] In some embodiments, the correlation processing of the current waypoint location data, the time data, and the image data to obtain inspection data includes:

[0039] The image data is standardized to obtain the processed image data;

[0040] The current waypoint location data is extracted and processed to obtain standardized location identifier data;

[0041] The processed image data, standardized location identifier data, and time data are subjected to field association processing to obtain multi-dimensional data;

[0042] The multi-dimensional data is integrated and processed to obtain structured data;

[0043] The structured data is validated to obtain inspection data.

[0044] In some embodiments, the preset waypoint tolerance threshold ranges from 0.5m to 2m.

[0045] This application provides an intelligent navigation device for a drone, comprising:

[0046] The acquisition module is used to acquire point cloud data and motion data, perform calibration processing on the point cloud data and motion data, and obtain relative position and time synchronization parameters.

[0047] The processing module is used to perform map construction processing on the point cloud data and the motion data based on the relative position and time synchronization parameters to obtain a three-dimensional point cloud map.

[0048] The acquisition module is also used to acquire the flight area boundary coordinates, flight altitude and flight path step size, and to analyze the flight area boundary coordinates, flight altitude and flight path step size to obtain flight path planning parameters.

[0049] The processing module is also used to determine the boundary of the target inspection area based on the three-dimensional point cloud map, and to perform route planning processing on the route planning parameters and the boundary of the target inspection area to obtain optimized route data and corresponding waypoint sequence data.

[0050] The acquisition module is also used to acquire the real-time pose data of the UAV, and to compare the current position and waypoint sequence data of the UAV based on the real-time pose data. When the current position of the UAV is within the tolerance range of the preset waypoint, image acquisition is performed to obtain image data. The real-time pose data includes the current waypoint position data and time data.

[0051] The association module is used to perform association processing on the current waypoint location data, the time data, and the image data to obtain inspection data;

[0052] The processing module is also used to integrate and process the real-time pose data, the three-dimensional point cloud map, the current flight status of the UAV, and the inspection data to obtain an intelligent navigation route.

[0053] The computer device provided in this application includes a memory and a processor. The memory stores a computer program that can run on the processor. When the processor executes the program, it implements the method described in this application.

[0054] The computer-readable storage medium provided in this application embodiment stores a computer program thereon, which, when executed by a processor, implements the method described in this application embodiment.

[0055] This application provides an intelligent navigation method, device, equipment, and storage medium for unmanned aerial vehicles (UAVs). It acquires point cloud data and motion data, calibrates them to obtain relative position and time synchronization parameters, constructs a 3D point cloud map from the point cloud data and motion data, analyzes the flight area boundary coordinates, flight altitude, and flight path step length to generate flight path planning parameters, defines the target inspection area boundary using the 3D point cloud map, generates an optimized flight path and corresponding waypoint sequence data through flight path planning, acquires real-time UAV pose data, compares the current position with the waypoint sequence, and triggers image acquisition when the UAV is within the preset waypoint arrival tolerance range, obtaining inspection data, and integrating real-time pose data, the 3D point cloud map, flight status, and inspection data to form an intelligent navigation route. This enables high-precision positioning, waypoint and data acquisition linkage, improves navigation accuracy, inspection efficiency, and operational safety, and solves the technical problems mentioned in the background art. Attached Figure Description

[0056] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 A schematic diagram illustrating the implementation process of an intelligent navigation method for an unmanned aerial vehicle (UAV) provided in an embodiment of this application;

[0058] Figure 2 This application provides a schematic diagram of an implementation process for obtaining optimized route data and corresponding waypoint sequence data.

[0059] Figure 3 This is a schematic diagram of the structure of an intelligent navigation device for a drone provided in an embodiment of this application. Detailed Implementation

[0060] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0061] The following description of some technologies involved in the embodiments of this application is provided to aid understanding and should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this application. Similarly, for clarity and brevity, some descriptions of well-known functions and structures are omitted in the following description.

[0062] Figure 1 This is a schematic diagram illustrating the implementation flow of an intelligent navigation method for a drone provided in an embodiment of this application, including steps 101 to 107. Wherein, Figure 1 This is merely one execution order shown in the embodiments of this application and does not represent the only execution order of an intelligent navigation method for a drone. Where the final result can be achieved, Figure 1 The steps shown can be performed in parallel or in reverse order.

[0063] Step 101: Acquire point cloud data and motion data, perform calibration processing on the point cloud data and motion data, and obtain relative position and time synchronization parameters.

[0064] In this embodiment, the UAV is equipped with a lidar and an inertial measurement unit (IMU). Before the inspection mission begins, the lidar collects point cloud data of the target bridge area, and the inertial measurement unit collects motion data of the UAV. Subsequently, a multi-sensor joint calibration algorithm is executed to calibrate the collected point cloud data and motion data, accurately measure and correct the installation position deviation between the lidar and the inertial measurement unit, and simultaneously synchronize the data acquisition timestamps of both to obtain the relative position and time synchronization parameters.

[0065] Step 102: Based on the relative position and time synchronization parameters, perform map construction processing on the point cloud data and motion data to obtain a three-dimensional point cloud map.

[0066] In this embodiment, the onboard computer of the UAV initiates a self-starting script to automatically load the multi-sensor tightly coupled synchronous localization and mapping (SLAM) module. The obtained relative position and time synchronization parameters are used to perform spatial alignment and temporal matching processing on the point cloud data and motion data. The relative position is used to eliminate data spatial offset caused by installation deviations of the multi-source sensors, and the correspondence between multi-source data at the same moment is established based on the time synchronization parameters. The processed fused data is processed in real-time using the SLAM algorithm to output the real-time pose data of the UAV, while simultaneously constructing a 3D point cloud map of the target bridge area.

[0067] Step 103: Obtain the boundary coordinates of the flight area, the flight altitude, and the flight path step length. Perform analytical processing on the boundary coordinates of the flight area, the flight altitude, and the flight path step length to obtain the flight path planning parameters.

[0068] In this embodiment, the target bridge's flight area boundary coordinates, preset flight altitude, and route step length are input as task parameters. After receiving the parameters, the onboard computer performs integrity verification, removes missing or abnormal data, and then performs format parsing and standardization on the verified parameters, extracts core information, and classifies and integrates it to finally obtain route planning parameters that include path intervals, flight altitude, and area boundary parameters.

[0069] Step 104: Determine the boundary of the target inspection area based on the 3D point cloud map, perform route planning processing on the route planning parameters and the boundary of the target inspection area, and obtain optimized route data and corresponding waypoint sequence data.

[0070] In this embodiment, a generated 3D point cloud map is invoked to extract the bridge structure outline information. The flight area boundary coordinates in the obtained flight route planning parameters are corrected to determine the target inspection area boundary that conforms to the actual bridge structure. Subsequently, the vertex set of the target inspection area boundary is preprocessed, and a boundary polygon closure algorithm is used to form a continuous and closed effective inspection area. The circumscribed rectangle of the effective inspection area is calculated, and equally spaced parallel scan lines are generated based on the flight route step size in the flight route planning parameters. Intersection calculations are performed between each scan line and the effective inspection area boundary. After sorting the intersection points, adjacent intersection points form effective flight segments. Cruise waypoints are generated on each effective flight segment by interpolation according to the flight route step size, and the waypoint altitude is set to a preset flight altitude. A path direction alternation strategy is used to sort all waypoints (odd-numbered scan lines connect waypoints in ascending order, even-numbered lines connect in descending order) to form a zigzag optimized flight route, ultimately obtaining optimized flight route data and corresponding waypoint sequence data.

[0071] Step 105: Obtain the real-time pose data of the UAV. Based on the real-time pose data, compare and process the current position and waypoint sequence data of the UAV. When the current position of the UAV is within the tolerance range of the preset waypoint, perform image acquisition to obtain image data.

[0072] In this embodiment, during the UAV's flight, the SLAM module continuously outputs the UAV's real-time pose data (including current waypoint position data and time data). The flight control module acquires the real-time pose data in real time and compares it with the obtained waypoint sequence data to calculate the positional deviation between the UAV's current position and the target waypoint. A preset waypoint arrival tolerance range is a spherical region centered on the target waypoint. When the calculated positional deviation is less than or equal to this tolerance range, the UAV is determined to have arrived at the target waypoint. At this time, the flight control module sends a command to the flight control system to control the UAV to hover stably at the current waypoint for 2 seconds to eliminate body sway. Simultaneously, it sends a trigger signal to the mission payload control module to drive the camera to perform autofocus and image capture, acquiring image data of bridge structural details.

[0073] Step 106: Perform correlation processing on the current waypoint location data, time data, and image data to obtain inspection data.

[0074] In this embodiment, after acquiring image data, the mission payload control module immediately extracts the current waypoint position data and the time data of image acquisition from the real-time pose data. The image data is then standardized in format, and the standardized image data is correlated with the current waypoint position data and acquisition time data to establish a multi-dimensional data correspondence, generating structured data containing position and time identifiers. The structured data is then validated for integrity, and invalid correlated data is removed, ultimately yielding inspection data that meets the inspection requirements.

[0075] Step 107: Integrate and process the real-time pose data, 3D point cloud map, current flight status of the UAV, and inspection data to obtain the intelligent navigation route.

[0076] In this embodiment, the flight control module continuously collects current flight status data (including battery level, flight attitude, communication link status, etc.) from the flight control system and SLAM module. It integrates the flight status data with real-time pose data, 3D point cloud maps, and inspection data to form a complete navigation and inspection dataset. Based on real-time pose data, the flight trajectory is monitored, and the flight strategy is dynamically adjusted by combining flight status data and inspection data, ultimately forming an intelligent navigation route adapted to complex bridge environments.

[0077] This application's embodiments overcome the reliance on a single GNSS (Global Navigation Satellite System) positioning method through multi-source data fusion and calibration processing, achieving high-precision positioning in GNSS-denied environments such as under bridges and in urban canyons, thus improving flight safety in complex scenarios. A complete closed-loop process is constructed, encompassing data calibration, map building, flight route planning, pose comparison, data association, and navigation integration, achieving integrated navigation and mission execution and resolving the problems of fragmented modules and inefficient collaboration in existing systems. Optimizing flight route planning and waypoint generation with 3D point cloud maps replaces manual point setting, improving the completeness of inspection coverage, avoiding omissions of key points, reducing manual operation costs, and increasing operational efficiency. Automatic linkage between waypoint arrival and image acquisition, as well as multi-dimensional data association and integration, enhances the accuracy and traceability of inspection data, adapting to the high-precision inspection needs of complex structures such as bridges.

[0078] In the above Figure 1 Based on the above, this application also provides a schematic diagram of the implementation process for obtaining optimized route data and corresponding waypoint sequence data, as shown below. Figure 2 As shown, steps 201 to 205 are included:

[0079] Step 201: Correct the boundary of the initial inspection area based on the 3D point cloud map to obtain the boundary of the target inspection area.

[0080] In this embodiment, the flight path planning module calls a 3D point cloud map to extract the structural outline information of the target bridge, including the main beam orientation, pier distribution, and key structural node locations. The initial flight area boundary coordinates are compared with the actual bridge structure in the 3D point cloud map. Parts in the initial boundary that do not conform to the actual structure are corrected, invalid areas exceeding the bridge's range are removed, and missing key detection area boundary nodes are added. Finally, a target inspection area boundary that conforms to the actual bridge structure and covers the entire detection range is obtained. The vertex set of the target inspection area boundary is preprocessed using a boundary polygon closure algorithm to ensure that all vertices form a continuous, closed two-dimensional polygon region.

[0081] Step 202: Process the path layout by considering the path interval, flight altitude, and target inspection area boundary to obtain the path layout.

[0082] In this embodiment, the circumscribed rectangle of the two-dimensional polygon formed by the boundary of the target inspection area is first calculated. This circumscribed rectangle serves as the reference area for the path layout, ensuring that the path completely covers the target inspection area. Combining the path interval in the flight path planning parameters, a set of equally spaced parallel scan lines is generated based on one side of the circumscribed rectangle. The density of the scan lines is determined by the path interval, ensuring that the distance between adjacent scan lines is consistent with the preset path interval, achieving full coverage of the reference area. The direction of the scan lines is automatically set according to the direction of the main bridge beam identified in the 3D point cloud map. The preset flight altitude parameters are combined with the generated parallel scan lines to assign a fixed altitude to each scan line, forming a path layout in three-dimensional space and clearly defining the UAV's flight altitude and horizontal scan path distribution.

[0083] Step 203: Optimize the path layout to obtain the optimized path layout.

[0084] In this embodiment, the generated 3D path layout is optimized using an alternating path direction strategy to reduce redundant maneuvers during the UAV inspection process. The specific optimization rule is as follows: all parallel scan lines are numbered sequentially; for odd-numbered scan lines, their original extension direction remains unchanged; for even-numbered scan lines, their extension direction is reversed. Through this optimization strategy, after completing the inspection of one scan line, the UAV can switch to the starting position of the next scan line without performing a 180-degree turn, forming an efficient zigzag flight path. This significantly reduces flight energy consumption and time costs, while improving flight stability and avoiding positioning deviations caused by large maneuvers. After the above processing, an optimized path layout with smooth path connections and simplified maneuvers is obtained.

[0085] Step 204: Generate cruise points based on the path layout and the optimized direction of the path, and construct an initial waypoint sequence based on the cruise points.

[0086] In this embodiment, each scan line is processed sequentially according to the optimized path layout: Using a line segment and polygon intersection algorithm, all intersection points between each scan line and the target inspection area boundary are calculated. These intersection points are then sorted according to the extension direction of the scan line, with each pair of adjacent intersection points forming an effective flight segment within the target inspection area. On each effective flight segment, equal-interval interpolation is performed according to a preset path interval to generate multiple evenly distributed two-dimensional cruise points, ensuring that the distance between each cruise point is consistent with the path interval, achieving full coverage detection of the effective flight segment. Preset flight altitude parameters are assigned to each two-dimensional cruise point to complete the three-dimensional coordinate information, obtaining three-dimensional cruise point data. Following the optimized path layout order, all three-dimensional cruise points are arranged sequentially to form an initial waypoint sequence connected according to the flight path order, clarifying the flight sequence of the UAV.

[0087] Step 205: Integrate the initial waypoint sequence with the optimized path layout to obtain optimized route data and corresponding waypoint sequence data.

[0088] In this embodiment, the initial waypoint sequence is matched with the optimized path layout to verify whether the coordinates of each cruise point perfectly match the scan line position and extension direction in the path layout, correcting cruise points with coordinate deviations or unreasonable positions. The integrated waypoint sequence is structured by adding attribute information such as number, coordinates, flight altitude, and connection method between adjacent waypoints to each waypoint, forming standardized waypoint data. The structured waypoint sequence is integrated and packaged with the optimized path layout information to generate optimized route data containing complete flight path information and corresponding waypoint sequence data.

[0089] This application's embodiments correct the inspection area boundaries based on 3D point cloud maps, ensuring the target area definition closely matches the actual scenario and resolving the problem of incomplete coverage or invalid flights caused by deviations between the initial boundaries and the actual structure. Through path layout, optimization, and waypoint sequence integration, optimized flight routes adapted to mission requirements are generated. Strategies such as alternating path directions reduce redundant UAV maneuvers, lowering energy consumption and flight time, and improving inspection efficiency. A standardized flight route planning process automates and accurately generates waypoint sequences, avoiding the subjectivity and inefficiency of manual planning and ensuring the standardization and completeness of inspection paths.

[0090] In some embodiments, real-time pose data of the UAV is acquired, and the current position and waypoint sequence data of the UAV are compared and processed based on the real-time pose data. When the current position of the UAV is within the tolerance range of the preset waypoint, image acquisition is performed to obtain image data, including: fusing point cloud data, motion data and three-dimensional point cloud map to obtain real-time pose data.

[0091] Specifically, during the drone's flight, the lidar continuously collects point cloud data of the target area, while the inertial measurement unit (IMU) simultaneously collects the drone's motion data (including acceleration, angular velocity, etc.). The onboard computer's multi-sensor tightly coupled SLAM module receives the aforementioned point cloud data and motion data in real time, and simultaneously calls upon a 3D point cloud map to initiate the fusion processing flow. First, based on the environmental reference information provided by the 3D point cloud map, the newly collected point cloud data is spatially calibrated to eliminate local data deviations caused by environmental occlusion. Then, combined with the relative position and time synchronization parameters obtained from the previous calibration, the point cloud data and motion data are time-series aligned to ensure accurate correlation of multi-source data at the same moment. The SLAM algorithm performs real-time calculations on the fused multi-source data, outputting the drone's position, attitude, and other core information at a frequency of 10Hz, integrating them to form complete real-time pose data.

[0092] Furthermore, the current position in the real-time pose data is compared with the waypoint sequence data to obtain the position deviation data.

[0093] Specifically, the flight controller module acquires real-time pose data output by the SLAM module via the internal data link of the onboard computer, extracting the UAV's current 3D position information. Simultaneously, it reads the coordinates of the target waypoint from the waypoint sequence data generated by the route planning module in a preset order. The flight controller module uses a spatial distance calculation method to perform real-time comparison and calculation between the UAV's current 3D position and the target waypoint coordinates, accurately calculating the straight-line distance between the two in 3D space. This distance represents the position deviation data, characterizing the position difference. The entire comparison process is synchronized with the positioning output frequency of the SLAM module, ensuring the real-time nature and accuracy of the position deviation data.

[0094] Furthermore, the position deviation data is compared with the preset waypoint tolerance threshold. When the position deviation data is less than or equal to the preset waypoint tolerance threshold, a waypoint arrival confirmation signal is obtained.

[0095] Specifically, the system pre-sets a waypoint tolerance threshold, which corresponds to the radius of a spherical region centered on the target waypoint. In this application, the preset waypoint tolerance threshold ranges from 0.5m to 2m. The flight controller module compares the calculated position deviation data with the preset waypoint tolerance threshold in real time. When the position deviation data is less than or equal to the waypoint tolerance threshold, it indicates that the UAV has entered the effective coverage area of ​​the target waypoint. The flight controller module determines that the waypoint has been successfully reached and immediately generates a waypoint arrival confirmation signal.

[0096] Furthermore, based on the waypoint arrival confirmation signal, the drone is controlled to acquire images and obtain image data of the target area.

[0097] Specifically, after receiving the waypoint arrival confirmation signal, the flight control system, under the command of the flight controller module, controls the UAV to hover stably at its current position for 2 seconds to eliminate any potential aircraft sway during flight and ensure the clarity of the acquired images. Simultaneously, upon receiving the confirmation signal, the mission payload control module drives the onboard camera to start working via a hardware interface. After automatically completing the focusing operation, it performs image processing, accurately capturing detailed images of the bridge structure corresponding to the target waypoint. The raw image data acquired by the camera is directly transmitted to the mission payload control module for temporary storage, forming the image data of the target area.

[0098] This application embodiment acquires real-time pose data through a multi-source fusion method using point cloud data, motion data, and 3D point cloud maps, significantly improving positioning accuracy and stability and overcoming pose misjudgment caused by single sensor data deviation. By calculating position deviation and comparing it with waypoint tolerance thresholds, the system automatically determines the waypoint arrival status, replacing manual image acquisition and achieving automatic linkage between waypoint arrival and task execution, thus improving the level of intelligent inspection. Upon waypoint arrival, hovering and image acquisition are triggered simultaneously, eliminating interference from aircraft sway and ensuring the clarity and accuracy of the acquired images, providing high-quality basic data for subsequent inspection data association.

[0099] In some embodiments, path layout processing is performed on the path interval, flight altitude, and target inspection area boundary to obtain a path layout, including: closing the target inspection area boundary to obtain an effective inspection area.

[0100] Specifically, the route planning module receives the previously revised boundary data of the target inspection area. This boundary data contains a series of discrete vertex coordinates, which are processed using a boundary polygon closure algorithm. By sequentially connecting each vertex, missing boundary connection points are filled in, and duplicate or invalid vertex data is removed, ensuring that all vertices form a continuous, gap-free two-dimensional polygonal region. After closure processing, a valid inspection area with a clearly defined range and complete boundaries is formed.

[0101] Furthermore, the effective inspection area is subjected to benchmark fitting processing to obtain the circumscribed benchmark area.

[0102] Specifically, for the effective inspection area, a circumscribed reference region fitting operation is performed. By calculating the maximum circumscribed rectangle of the two-dimensional polygon, and using the extreme coordinates of the polygon's boundary vertices as a basis, the positions of the four sides of the rectangle are determined to ensure that it completely encloses the effective inspection area, and that each side of the rectangle is parallel to the coordinate axes. The circumscribed rectangle is the circumscribed reference region.

[0103] Furthermore, based on the path interval, scan lines are generated for the external reference area to obtain parallel scan lines.

[0104] Specifically, based on the circumscribed reference area and combined with the path interval in the flight path planning parameters, the scan line generation process is initiated. A horizontal or vertical edge of the circumscribed reference area is selected as the starting edge for scanning. Following the preset path interval, multiple equally spaced parallel scan lines are generated parallelly to the other side of the circumscribed reference area. The distance between adjacent scan lines strictly adheres to the path interval to ensure that the scan line density meets the inspection accuracy requirements, while achieving full coverage of the circumscribed reference area, thus ensuring no blind spots within the effective inspection area.

[0105] Furthermore, the parallel scan lines are subjected to orientation adaptation processing based on the boundary of the target inspection area to obtain the processed parallel scan lines.

[0106] Specifically, by combining the 3D point cloud map, the orientation information of the main beam in the target inspection area is extracted, which serves as the primary basis for adapting the scan line direction. If the orientation of the main beam identified in the 3D point cloud map is clear, the direction of the parallel scan line is adjusted to be consistent with the orientation of the main beam, allowing the scan line to extend along the bridge structure and optimizing the rationality of the inspection path. After receiving the instruction, the route planning module adjusts the direction of the parallel scan line to complete the direction adaptation. After the above processing, a parallel scan line with a direction that fits the requirements of the inspection scenario is obtained.

[0107] Furthermore, path layout processing is performed on the parallel scan lines and flight altitude to obtain the path layout in three-dimensional space.

[0108] Specifically, the preset flight altitude data is extracted from the flight path planning parameters, and this altitude parameter is uniformly assigned to each parallel scan line. By adding a fixed altitude attribute to each parallel scan line in the two-dimensional plane, the original two-dimensional scan path is transformed into a path segment in three-dimensional space. Each three-dimensional path segment clearly defines the UAV's flight altitude and horizontal trajectory on that path segment, and all three-dimensional path segments together constitute a complete three-dimensional spatial path layout.

[0109] This application embodiment clarifies the effective inspection range and benchmark framework by closing the boundary of the target inspection area and fitting a benchmark, avoiding path coverage omissions or invalid flights exceeding the inspection range due to blurred area boundaries. Parallel scan lines are generated at path intervals and adapted to the structural direction of the target area, ensuring full coverage without blind spots and conforming to the structure of the inspected object, thus improving the rationality and targeting of the path layout. A three-dimensional path layout is constructed by integrating flight altitude parameters, clarifying the UAV's spatial flight trajectory and providing a precise and standardized basic framework for subsequent path optimization and waypoint generation, ensuring the stability of subsequent route execution.

[0110] In some embodiments, cruise points are generated based on the path layout and the optimized direction of the path, and an initial waypoint sequence is constructed based on the cruise points, including: performing intersection processing on the parallel scan lines and the boundary of the target inspection area to obtain intersection point data.

[0111] Specifically, the route planning module extracts each parallel scan line after direction adaptation, treating it as a line segment to be processed. Intersection calculations are then performed between this line and the closed two-dimensional polygon formed by the boundary of the target inspection area. A line segment and polygon intersection algorithm is used to accurately calculate the coordinates of all intersection points between each scan line and the polygon boundary, including the entry point of the scan line into the polygon area and the exit point. The calculated intersection coordinates are then deduplicated, removing duplicates or redundant intersections with coordinate errors within acceptable limits. Finally, the intersection point data corresponding to each parallel scan line and the boundary of the target inspection area are obtained, clarifying the coverage area of ​​the scan line within the effective inspection area.

[0112] Furthermore, the intersection data is sorted to obtain ordered intersection data.

[0113] Specifically, the intersection data corresponding to each parallel scan line is sorted according to the extension direction of the scan line. Using the starting end of the scan line as a reference, all intersection points are arranged in ascending or descending order based on their position coordinates on the scan line, ensuring that the intersection points are arranged sequentially according to the scan line's progression. This sorting transforms the originally discrete intersection data into an ordered sequence, clearly defining the order in which the scan line enters and leaves the target inspection area.

[0114] Furthermore, adjacent pairing processing is performed on the ordered intersection data to obtain valid flight segment data.

[0115] Specifically, the sorted intersection data is paired according to the adjacency principle, that is, the first and second intersections, the third and fourth intersections, and so on, in the ordered intersection sequence are taken to form multiple pairs of intersections. The line segment corresponding to each pair of intersections is the valid flight segment located within the target inspection area. This flight segment is the path segment that the UAV actually needs to fly and perform the inspection task. If the number of intersection data for a certain scan line is odd, the last single intersection that cannot be paired is discarded to ensure that all valid flight segments are complete line segments and are completely located within the target inspection area, avoiding invalid paths for the UAV to fly outside the area.

[0116] Furthermore, based on the path interval, the effective flight segment data is interpolated at equal intervals to obtain two-dimensional cruise point data.

[0117] Specifically, for each valid flight segment, cruise points are generated using an equal-interval interpolation algorithm, taking the two endpoints of the valid flight segment as a reference and combining the path interval in the route planning parameters. During the interpolation process, the distance between two adjacent cruise points is strictly controlled to ensure it matches the preset path interval, resulting in a uniform distribution of cruise points across the valid flight segment. If the length of the valid flight segment cannot be divided evenly by the path interval, the distance between the last cruise point and the previous cruise point is appropriately adjusted at the end of the segment to ensure that all cruise points fall on the valid flight segment and do not exceed the segment boundary, ultimately yielding two-dimensional cruise point data for each valid flight segment.

[0118] Three-dimensional cruise point data is obtained by performing three-dimensional coordinate completion processing on the two-dimensional cruise point data based on the flight altitude.

[0119] Specifically, the preset flight altitude data is extracted from the route planning parameters. This altitude serves as the uniform flight altitude for the UAV to perform inspection missions. This flight altitude is used as the z-axis coordinate value and assigned to the corresponding two-dimensional cruise point for each valid flight segment. Three-dimensional spatial coordinate information is then completed for each two-dimensional cruise point. Through coordinate completion, the original two-dimensional cruise points, which only contained horizontal position, are transformed into three-dimensional cruise point data containing both horizontal position and flight altitude, thus clarifying the specific spatial position of the UAV at each cruise point.

[0120] Furthermore, the three-dimensional cruise point data is sequentially arranged to obtain the initial waypoint sequence.

[0121] Specifically, in conjunction with the path optimization direction, the 3D cruise point data corresponding to all valid flight segments are arranged sequentially. Cruise points corresponding to each scan line are processed in the order of their parallel scan line numbers: for scan lines with odd numbers, the corresponding 3D cruise points are arranged in ascending order; for scan lines with even numbers, the corresponding 3D cruise points are arranged in descending order in the opposite direction. This arrangement ensures smooth connection of cruise points between adjacent scan lines, forming a zigzag flight path sequence. All the 3D cruise points arranged according to the rules are then sequentially linked together to form a complete and ordered initial waypoint sequence.

[0122] This application's embodiments employ processing such as intersection, sorting, and pairing of scan lines with region boundaries to accurately divide effective flight segments, ensuring that cruise points are distributed only within the inspection area, avoiding the generation of invalid waypoints, and improving the effectiveness of the waypoint sequence. Cruise points are generated based on equal-interval interpolation of path intervals, and the three-dimensional coordinates are completed by combining flight altitude, ensuring uniform distribution and accurate spatial positioning of cruise points, providing support for stable UAV flight and accurate data collection. The three-dimensional cruise points are arranged according to the path optimization direction, forming a smooth zigzag initial waypoint sequence, reducing large maneuvers during UAV flight, lowering energy consumption and positioning deviation risks, and improving flight efficiency and stability.

[0123] In some embodiments, current waypoint location data, time data, and image data are correlated to obtain inspection data, including: standardizing the image data to obtain processed image data.

[0124] Specifically, after receiving the raw image data from the camera, the task payload control module initiates a format standardization process. Following preset inspection data storage specifications, the raw image data is uniformly converted into a common image format to ensure data compatibility and ease of subsequent processing. Simultaneously, the image resolution and compression ratio are standardized and adjusted, invalid pixel areas caused by shooting interference are removed, and valid image content with clear bridge structural details is retained, ultimately resulting in processed image data with a unified format and meeting quality standards.

[0125] Furthermore, the current waypoint location data is extracted and processed to obtain standardized location identifier data.

[0126] Specifically, the flight controller module accurately extracts the three-dimensional spatial coordinates of the current waypoint from the real-time pose data, and the coordinate data is consistent with the waypoint sequence data generated by the route planning module. Subsequently, the extracted three-dimensional coordinates are standardized according to the preset coordinate format specifications, unifying the coordinate system identifier and numerical precision, and converting them into standardized position identifier data in string form, clearly defining the specific spatial location corresponding to the image acquisition, and providing a unique position identifier for data association.

[0127] Furthermore, the processed image data, standardized location identifier data, and time data are subjected to field association processing to obtain multi-dimensional data.

[0128] Specifically, the mission payload control module obtains the time data of the image acquisition moment from the onboard computer system. This time data is accurate to the millisecond and recorded in the form of a timestamp. Subsequently, a field association mapping relationship is established, binding the processed image data, standardized location identification data, and timestamp data according to the image-location-time correspondence, forming a multi-dimensional data set containing image information, spatial location information, and time information, ensuring that each inspection image can accurately correspond to its acquisition location and acquisition time.

[0129] Furthermore, the multi-dimensional data is integrated and processed to obtain structured data.

[0130] Specifically, after receiving a multi-dimensional data set, the data processing module integrates it according to a preset structured format. Using key-value pairs or a tabular structure, fixed fields are defined for the integrated data, including core fields such as image storage path, standardized location identifier, acquisition timestamp, and waypoint number. Multi-dimensional data is then populated into the corresponding fields one by one, forming structured data in a unified format. This structured data supports rapid retrieval and parsing.

[0131] Furthermore, the structured data is validated to obtain inspection data.

[0132] Specifically, the data processing module performs integrity and validity checks on the structured data. For integrity checks, it checks for missing data in each field, ensuring that all fields in each data record are complete. For validity checks, it verifies whether the coordinates of the standardized location markers are within a reasonable range of the target inspection area, whether the data collection timestamp matches the task execution period, and whether the image storage path is accessible. Missing fields and invalid records with abnormal data discovered during the verification process are removed, retaining all valid structured data to ultimately form inspection data that meets the requirements of the inspection task.

[0133] This application's embodiments standardize image and location data to achieve unified formats for multi-source data, solving the problems of scattered, disorganized, and incompatible existing inspection data. It establishes multi-dimensional field associations between image, location, and time, ensuring each piece of inspection data has accurate spatial and temporal identifiers, thus improving data traceability. Through structured integration and integrity and validity checks, invalid data is eliminated, guaranteeing the quality and reliability of inspection data, preventing abnormal data from affecting the judgment of inspection results, and improving the overall effectiveness of inspection tasks.

[0134] While this application provides method operation steps as shown in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive labor. The order of steps listed in this embodiment is merely one possible execution order among many and does not represent the only execution order. In actual device or client product execution, the method can be executed sequentially according to this embodiment or the accompanying drawings, or in parallel (e.g., in a parallel processor or multi-threaded processing environment).

[0135] like Figure 3 As shown in the illustration, this application also provides an intelligent navigation system 300 for a drone. The device includes:

[0136] The acquisition module 301 is used to acquire point cloud data and motion data, perform calibration processing on the point cloud data and motion data, and obtain relative position and time synchronization parameters.

[0137] The processing module 302 is used to perform map construction processing on point cloud data and motion data based on relative position and time synchronization parameters to obtain a three-dimensional point cloud map.

[0138] The acquisition module 301 is also used to acquire the flight area boundary coordinates, flight altitude and route step size, and to analyze and process the flight area boundary coordinates, flight altitude and route step size to obtain route planning parameters.

[0139] The processing module 302 is also used to determine the boundary of the target inspection area based on the three-dimensional point cloud map, perform route planning processing on the route planning parameters and the boundary of the target inspection area, and obtain optimized route data and corresponding waypoint sequence data.

[0140] The acquisition module 301 is also used to acquire the real-time pose data of the UAV, and to compare and process the current position and waypoint sequence data of the UAV based on the real-time pose data. When the current position of the UAV is within the tolerance range of the preset waypoint, image acquisition is performed to obtain image data. The real-time pose data includes the current waypoint position data and time data.

[0141] The association module 303 is used to perform association processing on the current waypoint location data, time data, and image data to obtain inspection data.

[0142] The processing module 302 is also used to integrate and process real-time pose data, 3D point cloud map, current flight status of UAV and inspection data to obtain intelligent navigation route.

[0143] In some embodiments, the processing module 302 is further configured to correct the boundary of the initial inspection area based on the three-dimensional point cloud map to obtain the boundary of the target inspection area.

[0144] The processing module 302 is also used to process the path layout of the path interval, flight altitude and target inspection area boundary to obtain the path layout.

[0145] The processing module 302 is also used to perform path optimization processing on the path layout to obtain an optimized path layout.

[0146] The processing module 302 is also used to generate cruise points based on the path layout and the optimized direction of the path, and to construct an initial waypoint sequence based on the cruise points.

[0147] The processing module 302 is also used to integrate the initial waypoint sequence with the optimized path layout to obtain optimized route data and corresponding waypoint sequence data.

[0148] In some embodiments, the processing module 302 is further configured to perform fusion processing on point cloud data, motion data, and a 3D point cloud map to obtain real-time pose data.

[0149] The processing module 302 is also used to compare the current position in the real-time pose data with the waypoint sequence data to obtain position deviation data.

[0150] The processing module 302 is also used to compare the position deviation data with the preset waypoint tolerance threshold, and to obtain a waypoint arrival confirmation signal when the position deviation data is less than or equal to the preset waypoint tolerance threshold.

[0151] The acquisition module 301 is also used to control the UAV to acquire images based on the waypoint arrival confirmation signal, so as to obtain image data of the target area.

[0152] In some embodiments, the processing module 302 is further configured to close the boundary of the target inspection area to obtain an effective inspection area.

[0153] The processing module 302 is also used to perform benchmark fitting processing on the effective inspection area to obtain the circumscribed benchmark area.

[0154] The processing module 302 is also used to perform scan line generation processing on the external reference area according to the path interval to obtain parallel scan lines.

[0155] The processing module 302 is also used to perform orientation adaptation processing on the parallel scan lines based on the boundary of the target inspection area to obtain the processed parallel scan lines.

[0156] The processing module 302 is also used to perform path layout processing on the parallel scan lines and flight altitude to obtain the path layout in three-dimensional space.

[0157] In some embodiments, the processing module 302 is further configured to perform intersection processing on the parallel scan lines and the boundary of the target inspection area to obtain intersection point data.

[0158] The processing module 302 is also used to sort the intersection data to obtain ordered intersection data.

[0159] The processing module 302 is also used to perform adjacent pairing processing on the ordered intersection data to obtain valid flight segment data.

[0160] The processing module 302 is also used to perform equal-interval interpolation processing on the effective flight segment data based on the path interval to obtain two-dimensional cruise point data.

[0161] The processing module 302 is also used to perform three-dimensional coordinate completion processing on the two-dimensional cruise point data based on the flight altitude to obtain three-dimensional cruise point data.

[0162] The processing module 302 is also used to sequentially arrange the three-dimensional cruise point data to obtain an initial waypoint sequence.

[0163] In some embodiments, the processing module 302 is further configured to perform format standardization processing on the image data to obtain processed image data.

[0164] The processing module 302 is also used to extract and process the current waypoint position data to obtain standardized position identification data.

[0165] The association module 303 is also used to perform field association processing on the processed image data, standardized location identification data and time data to obtain multi-dimensional data.

[0166] The processing module 302 is also used to integrate and process multi-dimensional data to obtain structured data.

[0167] The processing module 302 is also used to perform verification processing on the structured data to obtain inspection data.

[0168] Some modules in the apparatus described in this application can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, classes, etc., that perform a specific task or implement a specific abstract data type. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0169] The apparatus or module described in the above embodiments can be implemented by a computer chip or physical entity, or by a product with a certain function. For ease of description, the above apparatus is described by dividing it into various modules according to their functions. When implementing the embodiments of this application, the functions of each module can be implemented in one or more software and / or hardware. Of course, a module that implements a certain function can also be implemented by combining multiple sub-modules or sub-units.

[0170] The methods, apparatus, or modules described in this application can be implemented in a computer-readable program code manner. The controller can be implemented in any suitable manner, such as a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicon Labs C8051F320. A memory controller can also be implemented as part of the control logic of a memory. Those skilled in the art will also recognize that, in addition to implementing the controller in purely computer-readable program code manner, the same functionality can be achieved by logically programming the method steps to make the controller take the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the means included within it for implementing various functions can also be considered as structures within the hardware component. Alternatively, the device used to implement various functions can be viewed as either a software module that implements the method or a structure within a hardware component.

[0171] This application also provides an apparatus, the apparatus comprising: a processor; a memory for storing processor-executable instructions; wherein, when the processor executes the executable instructions, it implements the method described in this application.

[0172] This application also provides a non-volatile computer-readable storage medium storing a computer program or instructions thereon, which, when executed, enables the method described in this application embodiment to be implemented.

[0173] Furthermore, in the various embodiments of the present invention, each functional module can be integrated into a processing module, or each module can exist independently, or two or more modules can be integrated into a single module.

[0174] The aforementioned storage media include, but are not limited to, Random Access Memory (RAM), Read-Only Memory (ROM), Cache, Hard Disk Drive (HDD), or Memory Card. The memory can be used to store computer program instructions.

[0175] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary hardware. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product, or it can be embodied in the process of data migration. The computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, mobile terminal, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application.

[0176] The various embodiments described in this specification are presented in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. All or part of this application can be used in numerous general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, mobile communication terminals, multiprocessor systems, microprocessor-based systems, programmable electronic devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices, etc.

[0177] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of this application.

Claims

1. An intelligent navigation method for unmanned aerial vehicles (UAVs), characterized in that, include: Acquire point cloud data of the target inspection area and motion data of the UAV, perform calibration processing on the point cloud data and motion data to obtain relative position and time synchronization parameters; Based on the relative position and time synchronization parameters, map construction processing is performed on the point cloud data and the motion data to obtain a three-dimensional point cloud map. The flight area boundary coordinates, flight altitude, and flight path step size are obtained. The flight area boundary coordinates, flight altitude, and flight path step size are analyzed to obtain the flight path planning parameters. Based on the three-dimensional point cloud map, the boundary of the target inspection area is determined, and the route planning parameters and the boundary of the target inspection area are processed to obtain optimized route data and corresponding waypoint sequence data. The real-time pose data of the UAV is acquired, and the current position and waypoint sequence data of the UAV are compared and processed based on the real-time pose data. When the current position of the UAV is within the tolerance range of the preset waypoint, image acquisition is performed to obtain image data. The real-time pose data includes the current waypoint position data and time data. The current waypoint location data, the time data, and the image data are correlated to obtain inspection data; The real-time pose data, the 3D point cloud map, the current flight status of the UAV, and the inspection data are integrated and processed to obtain an intelligent navigation route; The route planning parameters include path interval and flight altitude. The process involves determining the target inspection area boundary based on the 3D point cloud map, performing route planning processing on the route planning parameters and the target inspection area boundary to obtain optimized route data and corresponding waypoint sequence data, including: Based on the three-dimensional point cloud map, the boundary of the initial inspection area is corrected to obtain the boundary of the target inspection area. The path layout is obtained by processing the path interval, the flight altitude, and the boundary of the target inspection area. The path layout is optimized to obtain an optimized path layout. Cruise points are generated based on the optimized path layout and the optimized path direction, and an initial waypoint sequence is constructed based on the cruise points. The initial waypoint sequence and the optimized path layout are integrated to obtain optimized route data and corresponding waypoint sequence data.

2. The method according to claim 1, characterized in that, The process involves acquiring the real-time pose data of the UAV, comparing the UAV's current position and waypoint sequence data based on the real-time pose data, and performing image acquisition when the UAV's current position is within a preset waypoint tolerance range to obtain image data, including: The point cloud data, the motion data, and the 3D point cloud map are fused to obtain the real-time pose data. The current position in the real-time pose data is compared with the waypoint sequence data to obtain position deviation data; The position deviation data is compared with a preset waypoint tolerance threshold. When the position deviation data is less than or equal to the preset waypoint tolerance threshold, a waypoint arrival confirmation signal is obtained. Based on the waypoint arrival confirmation signal, the UAV is controlled to acquire images and obtain image data of the target area.

3. The method according to claim 1, characterized in that, The process of performing path layout processing on the path interval, the flight altitude, and the target inspection area boundary to obtain the path layout includes: The boundary of the target inspection area is closed to obtain an effective inspection area; The effective inspection area is subjected to benchmark fitting processing to obtain the circumscribed benchmark area; Based on the path interval, the external reference region is processed to generate scan lines to obtain parallel scan lines; Based on the boundary of the target inspection area, the parallel scan lines are subjected to orientation adaptation processing to obtain the processed parallel scan lines. The parallel scan lines and the flight altitude are processed to obtain the path layout in three-dimensional space.

4. The method according to claim 3, characterized in that, The step of generating cruise points based on the path layout and the optimized direction of the path, and constructing an initial waypoint sequence based on the cruise points, includes: The intersection of the parallel scan line and the boundary of the target inspection area is calculated to obtain the intersection point data. The intersection data is sorted to obtain ordered intersection data; The ordered intersection data is subjected to adjacent pairing processing to obtain valid flight segment data; Based on the path interval, the effective flight segment data is interpolated at equal intervals to obtain two-dimensional cruise point data; Based on the flight altitude, the two-dimensional cruise point data is subjected to three-dimensional coordinate completion processing to obtain three-dimensional cruise point data; The three-dimensional cruise point data is sequentially arranged to obtain an initial waypoint sequence.

5. The method according to claim 1, characterized in that, The process of correlating the current waypoint location data, the time data, and the image data to obtain inspection data includes: The image data is standardized to obtain the processed image data; The current waypoint location data is extracted and processed to obtain standardized location identifier data; The processed image data, standardized location identifier data, and time data are subjected to field correlation processing to obtain multi-dimensional data; The multi-dimensional data is integrated and processed to obtain structured data; The structured data is validated to obtain inspection data.

6. The method according to claim 2, characterized in that, The preset waypoint tolerance threshold ranges from 0.5m to 2m.

7. An intelligent navigation device for an unmanned aerial vehicle (UAV), characterized in that, include: The acquisition module is used to acquire point cloud data of the target inspection area and motion data of the UAV, and to perform calibration processing on the point cloud data and motion data to obtain relative position and time synchronization parameters. The processing module is used to perform map construction processing on the point cloud data and the motion data based on the relative position and time synchronization parameters to obtain a three-dimensional point cloud map. The acquisition module is also used to acquire the flight area boundary coordinates, flight altitude and flight path step size, and to analyze the flight area boundary coordinates, flight altitude and flight path step size to obtain flight path planning parameters. The processing module is also used to determine the boundary of the target inspection area based on the three-dimensional point cloud map, and to perform route planning processing on the route planning parameters and the boundary of the target inspection area to obtain optimized route data and corresponding waypoint sequence data. The acquisition module is also used to acquire the real-time pose data of the UAV, and to compare the current position and waypoint sequence data of the UAV based on the real-time pose data. When the current position of the UAV is within the tolerance range of the preset waypoint, image acquisition is performed to obtain image data. The real-time pose data includes the current waypoint position data and time data. The association module is used to perform association processing on the current waypoint location data, the time data, and the image data to obtain inspection data; The processing module is also used to integrate and process the real-time pose data, the three-dimensional point cloud map, the current flight status of the UAV, and the inspection data to obtain an intelligent navigation route. The route planning parameters include path interval and flight altitude. The processing module is further configured to determine the target inspection area boundary based on the 3D point cloud map, and perform route planning processing on the route planning parameters and the target inspection area boundary to obtain optimized route data and corresponding waypoint sequence data, wherein: Based on the three-dimensional point cloud map, the boundary of the initial inspection area is corrected to obtain the boundary of the target inspection area. The path layout is obtained by processing the path interval, the flight altitude, and the boundary of the target inspection area. The path layout is optimized to obtain an optimized path layout. Cruise points are generated based on the optimized path layout and the optimized path direction, and an initial waypoint sequence is constructed based on the cruise points. The initial waypoint sequence and the optimized path layout are integrated to obtain optimized route data and corresponding waypoint sequence data.

8. A computer device comprising a memory and a processor, the memory storing a computer program executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 6.

Images