An intelligent pointing flight control system and method for a vehicle-mounted unmanned aerial vehicle, a storage medium and a computer program product

Abstract

This invention proposes a vehicle-mounted drone intelligent point-and-fly control system, method, storage medium, and computer program product. The system includes: a map and interaction module, a geographic data service module, and a task planning and decision-making module. This invention enables deep integration of drones and vehicles, utilizing the vehicle's large screen and computing power to achieve intelligent point-and-fly flight; adaptively planning flight altitude based on cloud-based geographic data and task type; parallel processing of interactive commands such as zooming, gimbal control, and destination changes during flight, achieving a smooth "what you see is what you get" operation; and combining video clicks with cloud-based POI data to present building information in an augmented reality manner, achieving an intelligent "what you see is what you know" experience.

Description

Technical Field

[0001] This invention belongs to the field of automotive intelligent cockpit and drone collaborative control technology, specifically relating to an intelligent pointing flight control system, method, storage medium and computer program product for vehicle-mounted drones. Background Technology

[0002] With the increasing popularity of outdoor adventure, self-driving tours, emergency rescue, and geographic information collection, users' demand for beyond-line-of-sight perception of the vehicle's surroundings is growing. Traditional in-vehicle systems have a limited field of view, limited to the area around the vehicle, while consumer drones are complex to operate, requiring separate remote controllers and displays, thus being disconnected from the vehicle's systems and unable to achieve convenient, safe, and intelligent integrated operation.

[0003] Due to the lack of deep integration between drones and vehicle systems, existing technical solutions still have the following shortcomings in practical applications: 1) The interaction between drones and vehicle systems is weak, making it impossible to utilize the vehicle's large screen, computing power, and network for efficient control and information display; 2) Flight altitude planning is limited and lacks adaptability to the environment and tasks; 3) The process of changing targets and real-time interaction (such as viewpoint zooming and gimbal control) during flight is cumbersome; 4) The video information collected by drones is not deeply integrated with geographic information systems (GIS) and commercial data (POI), and the information value is not fully explored. Summary of the Invention

[0004] To achieve deep integration between drones and vehicles and realize an intelligent experience of "point and fly, what you see is what you get, what you see is what you know", this invention proposes a vehicle-mounted drone intelligent pointing flight control system, method, storage medium and computer program product.

[0005] One of the objectives of this invention is a vehicle-mounted unmanned aerial vehicle (UAV) intelligent pointing flight control system, comprising: The map and interaction module is used to receive the target point selected by the user on the map interface; The geographic data service module is used to provide geographic data of the flight area where the target point is located based on the target point; The mission planning and decision-making module uses static obstacle data from the geographic data of the flight area as obstacle avoidance constraints. It generates a two-dimensional planar path from the UAV's current position to the target point using a path generation algorithm. Each waypoint on the two-dimensional planar path is assigned a flight altitude to form a three-dimensional waypoint sequence. A flight mission is then generated based on this three-dimensional waypoint sequence and preset flight parameters. In this invention, the preset flight parameters refer to the necessary flight control parameters for the UAV to perform a flight mission, including but not limited to: takeoff point (i.e., the starting position of the UAV when it begins the mission), target speed at each waypoint (i.e., the speed the UAV should reach when it arrives at that waypoint), and hovering time (i.e., the time the UAV stays at the waypoint). These parameters can be preset according to factors such as the flight mission type, UAV performance, and environmental conditions, or can be dynamically adjusted during flight. The flight mission type refers to the operational purpose or application scenario corresponding to the UAV's flight mission, including but not limited to: sightseeing missions, mainly used for aerial photography of scenery and obtaining a wide field of view; detail observation missions, mainly used for close-up observation of the details of buildings or targets; and terrain survey missions, mainly used for large-scale scanning of ground terrain.

[0006] In this invention, static obstacles refer to obstacles whose positions remain fixed within the flight area, including but not limited to buildings, trees, utility poles, mountains, bridges, no-fly zone boundaries, etc. The location, outline, height, and other information of these static obstacles are pre-stored in the geographic data service module and provided in the form of digital elevation models (DEM) and building white model data. During the path planning process, the task planning and decision module uses these static obstacle data as obstacle avoidance constraints to ensure that the generated flight path does not cross or get too close to these fixed obstacles.

[0007] Furthermore, the map and interaction module is also used to receive the flight altitude value input by the user on the map interface; the mission planning and decision module generates the flight mission containing a three-dimensional waypoint sequence based on the target point and the flight altitude value.

[0008] Furthermore, the map and interaction module is also used to receive the flight mission type selected by the user; the mission planning and decision-making module generates a flight altitude value according to the preset rules corresponding to the flight mission type, and generates the flight mission containing a three-dimensional waypoint sequence based on the target point and the flight altitude value. The preset rules are determined according to the flight mission type. For example, when the flight mission type is sightseeing, the preset rule is: 50 meters above the tallest building; when the flight mission type is detailed observation, the preset rule is: 20 meters above the tallest building; when the flight mission type is terrain survey, the preset rule is: 100 meters above the ground.

[0009] Furthermore, the map and interaction module is also used to receive the user-selected altitude mode. When the altitude mode is automatic, the task planning and decision-making module sends the coordinates and query radius of the target point to the geographic data service module, receives the geographic data of the flight area returned by the geographic data service module, determines the ground elevation of the target point and the elevation of the highest building on the planned path based on the geographic data, and generates a flight altitude value based on the ground elevation of the target point, the elevation of the highest building on the planned path, and an additional offset determined according to the flight task type selected by the user; and generates the flight task containing a three-dimensional waypoint sequence based on the target point and the flight altitude value. The geographic data includes digital elevation model (DEM) data and building data. Specifically, the DEM data is used to obtain the ground elevation H_terrain of the target point, and the building data is used to obtain the elevation H_building of the highest building on the planned path. The additional offset H_task is determined according to the flight task type, for example: sightseeing task H_task = 20 meters, detailed observation task H_task = 0 meters, terrain survey task H_task = 50 meters. The safety redundancy ΔH_safe can be set according to the flight environment, with a typical value of 30 meters.

[0010] Furthermore, it also includes a visual processing and rendering module, used to receive and decode the real-time video stream uploaded by the drone, and to overlay the flight status information onto the decoded real-time video stream for display; it is also used to receive the target building clicked by the user on the decoded real-time video stream, estimate the geographic coordinates of the target building based on the pixel coordinates of the target building, and overlay the information of the target building onto the real-time video stream in an augmented reality manner for display.

[0011] Furthermore, the task planning and decision-making module is also used to convert various interactive commands input by the user into corresponding control commands and send them to the UAV during the autonomous flight of the UAV.

[0012] Furthermore, when the interaction command is a zoom command, the task planning and decision-making module is also used to recognize the user's two-finger pinch gesture in the video display area, calculate the zoom factor based on the change in the distance between the two finger touch points, map the zoom factor to a camera zoom ratio command, and send the zoom command to the drone.

[0013] Furthermore, it also includes using a first-order low-pass filtering algorithm to smooth the zoom ratio command, avoiding image jitter caused by sudden changes in the zoom command. The formula is: cmd_smooth=α×cmd_raw+1-α×cmd_smooth_prev, where α is the smoothing coefficient, which can be dynamically adjusted according to the gesture speed, with a typical value range of 0.2-0.5.

[0014] Furthermore, when the interaction command is a gimbal control command, the task planning and decision-making module is also used to identify the user's operation on the virtual joystick or directional buttons, convert the operation into the angular velocity of the gimbal's pitch axis and yaw axis; and continuously send the gimbal speed control command containing the angular velocities to the UAV. The operation refers to the control input value generated when the user operates the virtual joystick or directional buttons. For the virtual joystick, the operation is the offset of the joystick in the X and Y axis directions (usually expressed as a value between -1.0 and 1.0); for the directional buttons, the operation is the button's pressed state (e.g., "up button pressed" corresponds to a positive pitch angular velocity).

[0015] Furthermore, when the interaction command is a destination change command, the task planning and decision-making module is also used to: send a pause command to the drone after receiving the new target point selected by the user on the map interface; obtain the real-time position of the drone, and generate a new path with the real-time position as the starting point, the new target point as the ending point, and the static obstacle data in the geographic data and the local obstacle data transmitted back by the drone as obstacle avoidance constraints.

[0016] Furthermore, the task planning and decision-making module generates the new path using a fast random search tree algorithm or a dynamic window method, and evaluates the candidate paths using a cost function, the expression of which is: Cost=w1×PathLength+w2×HazardProximity+w3×EnergyConsumption Wherein, PathLength is the path length, HazardProximity is the reciprocal of the nearest distance between the path and the static obstacle in the geographic data, EnergyConsumption is the energy consumption estimated based on the path length and flight altitude, and w1, w2, and w3 are weighting coefficients preset according to mission requirements. The local obstacle data is obtained in real time by the UAV through onboard visual sensors, radar, or ultrasonic sensors, including the position and motion information of dynamic obstacles (such as birds, pedestrians, and vehicles). During path replanning, this local obstacle data, together with the static obstacle data from the geographic data service module, serves as the obstacle avoidance constraint.

[0017] Furthermore, the visual processing and rendering module is also used for: Receives user clicks on target buildings in the decoded real-time video stream; Based on the location of the target building, the flight status information, and the current attitude of the UAV's gimbal, the geographic coordinates of the target building are estimated by back projection calculation based on the pinhole camera model. The geographic coordinates of the target building are sent to the geographic data service module, and the information of the target building returned by the geographic data service module is received. The information of the target building is displayed by overlaying it onto the real-time video stream in an augmented reality manner.

[0018] The back projection calculation is based on a pinhole camera model, converting image pixel coordinates (u,v) into geographic coordinates. Specifically, based on the UAV's current position, gimbal attitude, camera intrinsic parameter matrix K, and depth estimate d, the calculation is performed using the formula [ X c , Y c , Z c ] T =R× ( K -1 × [ u , v ,1] T ×d )) +T pixel coordinates ( u , v Convert to geographic coordinate system X c , Y c , Z c The depth estimate can be calculated based on the relative altitude of the UAV.

[0019] Furthermore, the geographic data service module includes: Three-dimensional geographic information and airspace services are used to return digital elevation model data, building white model data and airspace restriction information of the flight area based on the coordinates of the target point and the query radius. The building information big data service has a built-in building information database. It receives the geographic coordinates of the target building, performs dual verification through coordinate matching and image feature matching, retrieves the information of the target building from the building information database, and returns it.

[0020] Furthermore, it also includes a security monitoring thread, which runs independently of the task planning and decision-making module or the drone, for: Monitor the signal strength and packet loss rate of the communication link, and trigger a signal loss response strategy when the signal strength is below a preset threshold. Monitor the remaining battery power of the drone, dynamically calculate the latest return time based on the distance required for return, and trigger the return process when the battery power is below the safety threshold; Monitor whether the drone exceeds the preset electronic fence, and trigger the fence boundary crossing response strategy when it does.

[0021] A second objective of this invention is a vehicle-mounted unmanned aerial vehicle (UAV) intelligent pointing-and-fly control method, comprising: Receive the target point selected by the user on the map interface; Provide geographical data of the flight area where the target point is located based on the target point; Using static obstacle data from the geographic data of the flight area as obstacle avoidance constraints, a path generation algorithm is used to generate a two-dimensional planar path from the current position of the UAV to the target point. Each waypoint on the two-dimensional planar path is assigned a flight altitude to form a three-dimensional waypoint sequence, and a flight mission is generated based on the three-dimensional waypoint sequence and preset flight parameters.

[0022] A non-transitory computer-readable storage medium storing a computer program is provided to achieve the third objective of the present invention. The computer program, when executed by a processor, implements the steps of the intelligent pointing flight control method for vehicle-mounted unmanned aerial vehicles.

[0023] A computer program product for achieving the fourth objective of the present invention includes a computer program / instruction, which, when executed by a processor, implements the steps of the intelligent pointing flight control method for the vehicle-mounted UAV.

[0024] The beneficial effects of this invention include: (1) By utilizing the large screen, high computing power and network resources of the vehicle, the deep integration of the drone and the vehicle was realized, which solved the problems of weak interaction and poor display in traditional solutions; (2) Intelligent altitude planning driven by cloud-based geographic data and mission type enables adaptive adjustment of flight altitude; (3) Through the parallel processing interaction mechanism, smooth operation of zooming, gimbal control and destination change during flight was realized; (4) By combining video clicks with cloud POI data, information is enhanced in the form of AR, realizing the intelligent experience of "what you see is what you know". Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the system described in this invention; Figure 2 This is a flowchart illustrating human-computer interaction. Detailed Implementation

[0026] The following detailed embodiments are provided to explain the technical solutions of the present invention, so that those skilled in the art can understand the present invention. The scope of protection of the present invention is not limited to the following specific embodiments. Any modifications or improvements made by those skilled in the art that incorporate the technical solutions of the present invention but differ from the following detailed embodiments are also within the scope of protection of the present invention.

[0027] Example 1 A vehicle-mounted drone intelligent pointing flight control system, such as Figure 1 As shown, the system includes: a map and interaction module, a geographic data service module, a task planning and decision-making module, and a visual processing and rendering module.

[0028] In this embodiment, the map and interaction module, the task planning and decision-making module, and the visual processing and rendering module are integrated into the vehicle's in-vehicle intelligent terminal host, interacting with the user through the in-vehicle display screen. The geographic data service module is deployed in a cloud server and connects to the in-vehicle intelligent terminal host via a cellular network (such as 5G). The drone establishes a communication connection with the in-vehicle intelligent terminal host via in-vehicle high-power Wi-Fi 6E or a private radio frequency link. In one embodiment, a 5G cellular network is used instead of Wi-Fi 6E, and both the in-vehicle terminal and the drone are equipped with 5G communication modules, forwarding signaling and video streams through the cloud server. This solution is suitable for beyond-visual-range flight scenarios, with no limitation on communication distance.

[0029] The map and interaction module is used to display a high-precision map interface, which integrates the vehicle's current location, the drone's real-time location, the planned path, and the geographic information of the flight area. Users can select target points on the map interface via touch screen, or input flight altitude values, select flight mission types, or select altitude modes.

[0030] The geographic data service module is deployed in the cloud and includes 3D geographic information and airspace services as well as building information big data services.

[0031] The three-dimensional geographic information and airspace service stores digital elevation model (DEM) data of the flight area, white model data of buildings, and real-time airspace restriction information (such as temporary no-fly zones and meteorological information). When the service receives the target point coordinates and query radius sent by the vehicle terminal, it returns DEM data and building data within a circular area centered on the target point and with the query radius as the range.

[0032] The building information big data service has a built-in building information database. This database establishes a correlation between visual features and building POI (Point of Interest) information, storing data such as the building's geographical location, name, function, user rating, and historical information. When the service receives the building's geographical coordinates or image from the vehicle terminal, it performs dual verification through coordinate matching and image feature matching to retrieve and return the building's detailed information.

[0033] The mission planning and decision-making module is the core control unit of the system, responsible for flight mission generation, path planning, flight altitude determination, and processing of user interaction commands.

[0034] The task planning and decision-making module uses the A* algorithm as the path generation algorithm. This algorithm uses static obstacle data (building outlines, no-fly zone boundaries) returned by the geographic data service module as an obstacle map. Starting from the UAV's current position and ending at the user-selected target point, it searches for an optimal path in a two-dimensional plane that avoids all static obstacles. During the search, the cost function of the first algorithm comprehensively considers the path length and the distance to obstacles to ensure that the generated path is both short and safe. After generating the two-dimensional plane path, the module uses Bézier curves to smooth the path, eliminating sharp turns and making the flight trajectory smoother.

[0035] In one embodiment, Dijkstra's algorithm is used instead of A* algorithm. Dijkstra's algorithm can also generate the optimal path from the starting point to the ending point based on a static obstacle map, and is suitable for scenarios with relatively simple terrain.

[0036] In one embodiment, the method for determining flight altitude includes: Users directly input the flight altitude value on the map interface, and the mission planning and decision-making module uses this altitude value as the flight altitude; In another embodiment, the method for determining flight altitude includes: the user selecting a flight mission type, including "sightseeing," "detailed observation," and "terrain survey," and the system having a preset mission type-altitude rule mapping table. The default rule for sightseeing missions is: the flight altitude must be 50 meters above the tallest building; The preset rule for the detailed observation task is: the flight altitude must be 20 meters above the tallest building; The preset rule for terrain survey missions is: the flight altitude is 100 meters above the ground; The mission planning and decision-making module generates the flight altitude by calling the corresponding preset rules based on the mission type selected by the user.

[0037] In another embodiment, the method for determining the flight altitude includes: after the user selects the automatic mode, the mission planning and decision-making module sends the coordinates of the target point and the query radius (e.g., 500 meters) to the geographic data service module, and receives the returned DEM data and building data. The module calculates the suggested altitude value according to the following formula: H_final=max(H_terrain,H_building+ΔH_safe)+H_task Where H_terrain is the ground elevation of the target point; H_building is the elevation of the tallest building on the planned path; ΔH_safe is the safety redundancy, which is 30 meters in this embodiment; and H_task is the offset added according to the task type. The system has a preset task type-offset mapping table, for example: When the task type is sightseeing: H_task = 20 meters, to obtain a better view; When the task type is a detailed observation task: H_task=0 meters, to facilitate close-range observation; When the task type is terrain survey task: H_task=50 meters, to expand the observation range.

[0038] The system will visualize the calculated altitude value H and present it to the user for confirmation or modification. The user can confirm the use of this value or make modifications accordingly. Once confirmed, this altitude value will be used as the flight altitude.

[0039] After determining the flight altitude, the mission planning and decision-making module assigns the flight altitude to each waypoint on the two-dimensional plane path, forming a waypoint sequence WP containing three-dimensional coordinates (latitude, longitude, and altitude). The module further packages the waypoint sequence WP with preset flight parameters to generate a complete flight mission package. The preset flight parameters include: takeoff point (the starting position of the UAV when it begins its mission), target speed for each waypoint (set to 10 m / s in this embodiment), hovering time (set to 3 seconds in this embodiment), etc.

[0040] In one embodiment, during the autonomous flight of the UAV, the mission planning and decision-making module processes various interactive commands input by the user in parallel, such as... Figure 2 As shown; When the interaction command is a zoom command, i.e., the user performs a two-finger pinch gesture in the video display area, the module recognizes the gesture and calculates the zoom factor `scale_factor` based on the ratio of the initial distance between the two finger touch points to the current distance (e.g., a change from 1.0 to 2.0 represents a 2x zoom). The module linearly maps the zoom factor to the camera zoom ratio command `zoom_target`, for example, when `scale_factor` = 2.0, `zoom_target` = 2.0x. In one embodiment, to avoid abrupt lens changes, the module uses a first-order low-pass filtering algorithm to smooth the zoom ratio command. cmd_smooth=α×cmd_raw+1-α×cmd_smooth_prev Here, α is a smoothing coefficient, which can be dynamically adjusted according to the gesture speed. In this embodiment, the value of α ranges from 0.2 to 0.5. After generating the smoothed zoom command, the module sends the command to the drone, and the gimbal camera performs continuous optical zoom.

[0041] When the interaction command is a gimbal control command, i.e., when the user operates on the virtual joystick or directional buttons, the module recognizes the operation and converts it into the angular velocities of the gimbal's pitch and yaw axes. The module continuously sends gimbal speed control commands containing angular velocities to the drone. As long as the user holds down the operation controls, the commands continue to be sent, and the gimbal rotates smoothly at the corresponding speed.

[0042] When the interaction command is a destination change command, i.e., when the user clicks on a new target point on the map interface, the module sends the MAV_CMD_DO_PAUSE command to the drone, causing the drone to decelerate and hover at a safe position in the current flight segment. The module obtains the drone's real-time position P_current, uses P_current as the starting point and the new target point P_new as the ending point, and uses static obstacle data from the geographic data and local obstacle data transmitted by the drone as obstacle avoidance constraints to call the path generation algorithm to generate a new path. After generating the new flight mission, the module sends it to the drone, enabling the drone to execute the new flight mission.

[0043] The drone includes an adaptive flight control unit, an intelligent gimbal camera unit, and an airborne communication relay unit.

[0044] The adaptive flight control unit integrates an inertial measurement unit (IMU), magnetometer, barometer, and onboard GNSS receiver to calculate the UAV's position, velocity, and attitude in real time. The module receives flight mission packages from the mission planning and decision-making module. After verifying mission feasibility, it controls the UAV to autonomously take off from its current location and fly along a designated path. During flight, the module transmits key status information back to the vehicle at a frequency of 5Hz, including: real-time position (latitude, longitude, and altitude), velocity, heading, remaining battery power, communication link strength, and onboard sensor status.

[0045] The intelligent gimbal camera unit integrates a three-axis stabilized gimbal and a high-resolution zoom camera. The camera is factory-calibrated, and its intrinsic parameter matrix K is known. The module receives zoom commands and gimbal speed control commands, executes corresponding optical zoom and gimbal rotation, and continuously transmits the acquired video stream to the vehicle's infotainment system via a high-bandwidth video channel after H.265 hardware encoding.

[0046] The airborne communication relay unit is responsible for establishing a communication link with the vehicle-mounted terminal, uploading video streams and flight status information, and receiving flight mission and control commands.

[0047] The visual processing and rendering module is used to receive and decode the real-time video stream uploaded by the drone, and to display the flight status information (flight speed, altitude, remaining battery power, distance from the Home point, current flight mode, etc.) by overlaying it onto the decoded real-time video stream in OSD format.

[0048] The visual processing and rendering module is also used to realize the intelligent information enhancement function of "what you see is what you know", specifically including: When a user clicks on a target building in the decoded real-time video stream, the module captures the current video frame and simultaneously records the corresponding precise drone pose (position, attitude angles) and gimbal attitude angles (pitch angle, yaw angle), as well as the camera's current optical zoom level. The module performs back-projection calculations based on a pinhole camera model to estimate the geographic coordinates of the clicked building. X c , Y c , Z c Specifically, based on the image pixel coordinates (u,v) and the camera intrinsic parameter matrix... K Depth estimate d (Estimated based on the drone's current relative altitude), drone attitude rotation matrix R Translation vector T It can be calculated using the following formula: [ X c , Y c ,Z c ] T =R× ( K -1 × [ u , v ,1] T ×d )) +T Calculated coordinates ( X c , Y c , Z c Using this as the initial value, the module iteratively registers it with the digital elevation model (DEM) of the flight area to obtain precise geodetic coordinates. The module then compares these coordinates with the current video frame's cutoff value. Figure 1 The information is uploaded to a cloud-based building information big data service. The cloud service uses both coordinate matching and image feature matching for dual verification to retrieve detailed information about the building (such as name, address, business hours, user ratings, and a brief description), and then sends this information to the vehicle's infotainment system. After receiving the information, the module overlays it onto the corresponding building location in the video feed as a semi-transparent information card or AR tag.

[0049] Example 2 A method for intelligent guided flight control of a vehicle-mounted unmanned aerial vehicle (UAV) includes: Receive the target point selected by the user on the map interface; Provide geographical data of the flight area where the target point is located based on the target point; Using static obstacle data from the geographic data of the flight area as obstacle avoidance constraints, a path generation algorithm is used to generate a two-dimensional planar path from the current position of the UAV to the target point. Each waypoint on the two-dimensional planar path is assigned a flight altitude to form a three-dimensional waypoint sequence, and a flight mission is generated based on the three-dimensional waypoint sequence and preset flight parameters.

[0050] In one embodiment, the method further includes: receiving a flight altitude value input by a user on the map interface; and generating the flight mission containing a three-dimensional waypoint sequence based on the target point and the flight altitude value.

[0051] In one embodiment, the method further includes: receiving a flight mission type selected by the user from the vehicle-mounted interface; generating a flight altitude value according to a preset rule corresponding to the flight mission type; and generating the flight mission containing a three-dimensional waypoint sequence based on the target point and the flight altitude value. The preset rule is determined according to the flight mission type. When the flight mission type is sightseeing, the preset rule is: the flight altitude value is 50 meters above the tallest building; when the flight mission is detailed observation, the preset rule is: the flight altitude value is 20 meters above the tallest building; and when the flight mission is terrain surveying, the preset rule is: the flight altitude value is 100 meters above the ground.

[0052] In one embodiment, the method further includes: receiving a user-selected altitude mode; when the altitude mode is automatic, determining the geographic data of the flight area based on the coordinates of the target point and the query radius; determining the ground elevation of the target point and the elevation of the highest building on the planned path based on the geographic data; generating a flight altitude value based on the ground elevation of the target point, the elevation of the highest building on the planned path, and an additional offset determined according to the user-selected flight mission type; and generating the flight mission containing a three-dimensional waypoint sequence based on the target point and the flight altitude value. The geographic data includes digital elevation model (DEM) data and building data. Specifically, the DEM data is used to obtain the ground elevation H_terrain of the target point, and the building data is used to obtain the elevation H_building of the highest building on the planned path. The additional offset H_task is determined according to the flight mission type. In one embodiment, the additional offset for a sightseeing mission is H_task = 20 meters, for a detailed observation mission it is 0 meters, and for a terrain survey mission it is 50 meters. The safety redundancy ΔH_safe can be set according to the flight environment, with a typical value of 30 meters.

[0053] In one embodiment, the method for calculating the flight altitude value H_final includes: H_final=max(H_terrain,H_building+ΔH_safe)+H_task.

[0054] In one embodiment, the method further includes: receiving and decoding a real-time video stream uploaded by a drone, and displaying the flight status information superimposed on the decoded real-time video stream; and also receiving a target building clicked by a user on the decoded real-time video stream, estimating the geographic coordinates of the target building based on the pixel coordinates of the target building, and displaying the information of the target building superimposed on the real-time video stream in an augmented reality manner.

[0055] In one embodiment, the method further includes: during the autonomous flight of the UAV, converting various interactive commands input by the user into corresponding control commands and sending them to the UAV.

[0056] In one embodiment, when the interaction command is a zoom command, the task planning and decision-making module is further configured to recognize the user's two-finger pinch gesture in the video display area, calculate the zoom factor based on the change in the distance between the two finger touch points, map the zoom factor to a camera zoom ratio command, and send the zoom command to the drone.

[0057] In one embodiment, the zoom ratio command is smoothed using a first-order low-pass filtering algorithm to avoid image jitter caused by sudden changes in the zoom command. The formula is: cmd_smooth = α × cmd_raw + 1 - α × cmd_smooth_prev, where α is a smoothing coefficient that can be dynamically adjusted according to the gesture speed, typically ranging from 0.2 to 0.5.

[0058] In one embodiment, when the interaction command is a gimbal control command, the method further includes: identifying the user's operation on the virtual joystick or directional buttons, converting the operation into the angular velocity of the gimbal's pitch axis and yaw axis; and continuously sending a gimbal speed control command containing the angular velocities to the UAV. The operation refers to the control input value generated when the user operates the virtual joystick or directional buttons. For a virtual joystick, the operation is the offset of the joystick in the X and Y axis directions (usually expressed as a value between -1.0 and 1.0); for directional buttons, the operation is the button's pressed state (e.g., "up button pressed" corresponds to a positive pitch angular velocity).

[0059] In one embodiment, when the interaction command is a destination change command, the method further includes: sending a pause command to the drone after receiving a new target point selected by the user on the map interface; obtaining the real-time position of the drone; using the real-time position as the starting point and the new target point as the ending point, and using static obstacle data in the geographic data and local obstacle data transmitted back by the drone as obstacle avoidance constraints to generate a new path.

[0060] In one embodiment, the new path is generated using a fast random search tree algorithm or a dynamic window method, and the candidate paths are evaluated using a cost function, the expression of which is: Cost=w1×PathLength+w2×HazardProximity+w3×EnergyConsumption Wherein, PathLength is the path length, HazardProximity is the reciprocal of the nearest distance between the path and the static obstacle in the geographic data, EnergyConsumption is the energy consumption estimated based on the path length and flight altitude, and w1, w2, and w3 are weighting coefficients preset according to mission requirements. The local obstacle data is obtained in real time by the UAV through onboard visual sensors, radar, or ultrasonic sensors, including the position and motion information of dynamic obstacles (such as birds, pedestrians, and vehicles). During path replanning, this local obstacle data, together with the static obstacle data from the geographic data service module, serves as the obstacle avoidance constraint.

[0061] In one embodiment, the method further includes: Receives user clicks on target buildings in the decoded real-time video stream; Based on the location of the target building, the flight status information, and the current attitude of the UAV's gimbal, the geographic coordinates of the target building are estimated by back projection calculation based on the pinhole camera model. Send the geographic coordinates of the target building to the cloud and receive the information of the target building returned by the cloud; The information of the target building is displayed by overlaying it onto the real-time video stream in an augmented reality manner.

[0062] The back projection calculation is based on a pinhole camera model, converting image pixel coordinates (u,v) into geographic coordinates. Specifically, it is based on the UAV's current position, gimbal attitude, camera intrinsic parameter matrix K, and depth estimate. d Through the formula [ X c , Y c , Z c ] T =R× ( K -1 × [ u , v ,1] T ×d )) +T pixel coordinates ( u , v Convert to geographic coordinate system X c , Y c , Z c The depth estimate can be calculated based on the relative altitude of the UAV.

[0063] In one embodiment, the services provided by the cloud include: Three-dimensional geographic information and airspace services are used to return digital elevation model data, building white model data and airspace restriction information of the flight area based on the coordinates of the target point and the query radius. The building information big data service has a built-in building information database. It receives the geographic coordinates of the target building, performs dual verification through coordinate matching and image feature matching, retrieves the information of the target building from the building information database, and returns it.

[0064] In one embodiment, a security monitoring thread is also included, which runs independently of the vehicle or drone, for: Monitor the signal strength and packet loss rate of the communication link, and trigger a signal loss response strategy when the signal strength is below a preset threshold. Monitor the remaining battery power of the drone, dynamically calculate the latest return time based on the distance required for return, and trigger the return process when the battery power is below the safety threshold; Monitor whether the drone exceeds the preset electronic fence, and trigger the fence boundary crossing response strategy when it does.

[0065] Example 3 A method for intelligent guided flight control of a vehicle-mounted unmanned aerial vehicle (UAV) includes: Step S100: System initialization and link establishment.

[0066] After the vehicle is powered on, the onboard system starts up and completes a self-test. Then, it automatically scans through the communication gateway and establishes a secure communication link with the drone placed in a dedicated compartment on the roof / trunk. After the link is established, the drone completes a self-test, and the onboard intelligent terminal subsystem shares its high-precision vehicle GNSS location with the drone. The drone records and stores this location as its initial return point (i.e., the Home point).

[0067] Step S200: Map point selection and intelligent height planning, specifically including: S210: The user selects the target point P_target for the drone's flight on the vehicle's map interface via touchscreen, with latitude Lat_t and longitude Lon_t.

[0068] S220: The vehicle infotainment system offers users the following three drone flight altitude modes to choose from: 1. Custom mode: The user directly inputs the height value H_final; 2. Preset Mode: The vehicle system generates an altitude value H_final by calling preset rules based on the task type selected by the user (such as sightseeing, detailed observation, terrain survey). The preset rules are as follows: when the task type is sightseeing, the preset rule is: the flight altitude value is 50 meters above the highest building; when the task type is detailed observation, the preset rule is: the flight altitude value is 20 meters above the highest building; when the task type is terrain survey, the preset rule is: the flight altitude value is 100 meters above the ground.

[0069] 3. Automatic Mode: The vehicle system sends the 3D coordinates of the target point to the 3D geographic information and airspace service in the cloud, along with a preset query radius R_query. In this embodiment, the preset radius is 500 meters, but the user can modify it within a range of 100 meters to 2000 meters. The system requests geographic data for this area. Based on the target point coordinates and the query radius, the cloud determines a circular geographic area and returns the following data within that circular area: Digital Elevation Model (DEM) data: used to obtain the ground elevation H_terrain of the target point and its surrounding area; Building data includes information such as the geographical location, outline, and altitude of each building in the area, which is used to calculate the altitude H_building of the highest building on the planned path.

[0070] The system calculates the suggested flight altitude value H_final based on the following formula: H_final=max(H_terrain,H_building+ΔH_safe)+H_task Where H_terrain is the ground elevation of the target point; H_building is the elevation of the tallest building on the planned path; ΔH_safe is the safety redundancy, which is 30 meters in this embodiment; and H_task is the offset added according to the task type. The system visualizes the calculated height value H and presents it to the user for confirmation or modification. The system has a preset task type-offset mapping table, for example: When the task type is sightseeing: H_task = 20 meters, to obtain a better view; When the task type is a detailed observation task: H_task=0 meters, to facilitate close-range observation; When the task type is terrain survey task: H_task=50 meters, to expand the observation range.

[0071] S230: The vehicle-mounted task planning and decision-making module obtains the UAV's current position P_current (provided by the UAV through status feedback), and based on the user-confirmed target point P_target and flight altitude H_final, and the altitude value H_final determined in step S220, the module calls a built-in path search algorithm (such as the A* algorithm), combined with static obstacle data (such as buildings, no-fly zones) obtained from the cloud, to generate a two-dimensional planar path from the UAV's current position P_current to the target point P_target, and smooths the path (e.g., using Bézier curves or B-spline curve interpolation) to avoid sharp turns. Then, the altitude of each waypoint on this two-dimensional path is assigned the value H_final, forming a waypoint sequence WP_initial containing three-dimensional coordinates (latitude, longitude, altitude), and this waypoint sequence is sent to the UAV. The task planning and decision-making module further packages the waypoint sequence WP_initial with preset flight parameters (including takeoff point, target speed for each waypoint, hovering time, etc.) to generate a complete flight mission package.

[0072] Step S300: Autonomous Flight and Real-Time Interaction The mission planning and decision-making module sends the flight mission to the onboard flight control module, and the UAV takes off autonomously and follows the trajectory. During flight, the onboard module transmits flight status information back at a frequency of 5Hz, and the vision processing and rendering module decodes the video stream and overlays it onto the OSD display.

[0073] This step is implemented by a main control loop running within the vehicle-mounted mission planning and decision-making module. This loop operates at a high frequency (e.g., 10Hz), monitoring flight status, communication links, and three independent user interaction channels (zoom, gimbal, target change) in parallel. The system employs a priority-based non-preemptive scheduling strategy, ensuring that flight safety commands (such as obstacle avoidance and emergency hovering) have the highest priority, followed by destination changes, and finally view control commands.

[0074] The vehicle-mounted aircraft sends the complete flight mission package generated in step S230, which includes the takeoff point, ordered waypoint sequence, target speed of each waypoint, hovering time, etc., to the UAV's adaptive flight control unit via a security protocol (such as MAVLink) through the communication gateway. The adaptive flight control unit is used to perform autonomous flight and obstacle avoidance. The UAV's adaptive flight control unit verifies the feasibility of the flight mission package, such as whether the waypoints are reachable, and sends back a confirmation signal. Subsequently, the UAV autonomously takes off from its current location (possibly in hand or in a vehicle's roof compartment) and flies along the designated path.

[0075] In one embodiment, the drone also transmits key status information back to the vehicle via an airborne communication relay at a high frequency (e.g., 5Hz). The key status information includes: real-time location (GPS / RTK), altitude, speed, and heading; remaining battery power and communication link strength; and the status of airborne sensors, such as the status of obstacle avoidance cameras and forward vision systems.

[0076] In one embodiment, the video stream captured by the drone's intelligent gimbal camera unit is efficiently hardware encoded (e.g., H.265) on the device side, and the encoded bitstream is continuously transmitted to the vehicle's infotainment system through an independent high-bandwidth video channel (e.g., the 5.8GHz band of Wi-Fi 6E). In one embodiment, the visual processing and rendering module receives and software decodes the video stream. Before rendering, the module overlays the UAV status information (OSD) onto the video screen in a graphical manner. Typical information includes: flight speed, altitude, distance from the Home point; battery level icon; current flight mode (autonomous route, hovering, controlled) indication; and a semi-transparent overlay of the planned route and the target point. This information is optional.

[0077] In one embodiment, during the primary task of autonomous flight performed by the UAV, the system must process various interactive commands from the user in parallel, in real time, and safely, achieving a flexible operational experience with controllable flight, adjustable viewing angle, and modifiable target. This includes: The system processes user interaction commands in parallel: (1) Scaling instruction processing Users perform two-finger gestures in the vehicle's video display area or dedicated interactive area. The touchscreen driver layer of the vehicle's operating system (such as QNX or Android Automotive) recognizes the gesture and calculates a dimensionless scaling factor based on the ratio of the initial distance between the two finger touch points to the current distance. This scaling factor is a ratio, such as 2, which represents magnification by 2 times. The system then generates a GESTURE_ZOOM event and passes the GESTURE_ZOOM event and the scaling factor to the visual processing and rendering module.

[0078] The module maps scale_factor to camera zoom ratio commands. To avoid abrupt changes in lens zoom, a first-order low-pass filter is used to smooth the commands, generating the final zoom_command.

[0079] In one embodiment, the first-order low-pass filtering algorithm is as follows: cmd_smooth = α × cmd_raw + (1-α) × cmd_smooth_prev, where α is the smoothing coefficient, which is dynamically adjusted according to the gesture speed.

[0080] The smoothed zoom command is sent to the drone gimbal via a low-latency control channel that is separate from and prioritized for transmission from the video channel. The camera of the drone gimbal performs continuous optical zoom or digital zoom and transmits the new footage back in real time.

[0081] (2) Gimbal control command processing Users interact with the virtual joystick or directional buttons on the vehicle's infotainment system. The system's high-precision map and interaction module converts these UI events into incremental control values ​​(Δpitch, Δyaw) for the gimbal's pitch and yaw axes. This incremental mode, relative to the current angle, is more intuitive and has higher fault tolerance than the absolute angle mode.

[0082] When a user operates the virtual joystick, the module identifies the joystick's offset in the vertical and horizontal directions. The vertical offset is mapped to the gimbal's pitch axis rotational angular velocity ω_pitch (° / s), and the horizontal offset is mapped to the gimbal's yaw axis rotational angular velocity ω_yaw (° / s). For example, when the joystick is offset upwards by 50%, ω_pitch = 30° / s (upward rotation); when the joystick is offset to the left by 30%, ω_yaw = 15° / s (leftward rotation). When the user operates in both directions simultaneously, the gimbal performs pitch and yaw movements simultaneously with both angular velocities. A small dead zone is also set at the center of the joystick to prevent accidental activation. The generated gimbal speed control commands are continuously sent through a low-latency control channel; as long as the user holds down the joystick, the gimbal rotates smoothly at the corresponding speed.

[0083] When a user clicks the directional buttons, the up or down button click event is converted into an incremental control value Δpitch for the gimbal's pitch axis (e.g., up button corresponds to +5°, down button corresponds to -5°); the left or right button click event is converted into an incremental control value Δyaw for the gimbal's yaw axis (e.g., left button corresponds to +5°, right button corresponds to -5°). The generated incremental control commands are sent through a low-latency control channel, and the gimbal performs a one-time angle adjustment based on its current position.

[0084] The generated gimbal speed control command GIMBAL_SPEED_CONTROL is continuously transmitted through the low-latency control channel, causing the UAV gimbal to rotate smoothly at the corresponding speed, achieving a responsive and precise control effect. The gimbal speed control command instructs the UAV gimbal to rotate continuously around the pitch and yaw axes at specified angular velocities.

[0085] (3) Dynamic Destination Change Processing The triggering action includes: the user clicking on a new target point P_new on the vehicle's map.

[0086] Upon receiving the user's selection of a new target point, the vehicle-mounted task planning and decision-making module immediately sends the MAV_CMD_DO_PAUSE command to the drone to pause it. If the drone is currently in a controllable flight state, it will decelerate and hover at a safe position in the current flight segment.

[0087] The vehicle-mounted task planning and decision-making module obtains the drone's precise real-time position P_current and velocity V_current. Starting from P_current and ending at the user's new point P_new, it invokes the online path replanning algorithm.

[0088] In one embodiment, an improved Fast Random Search Tree (RRT*) algorithm is used to generate a smooth and safe path from the current location to a new target point. The inputs to the RRT algorithm include: a starting point P_current, an ending point P_new, a known global obstacle map (from a cloud-based 3D geographic information service), and real-time local obstacle information perceived by the UAV (provided and transmitted back by the onboard obstacle avoidance system). The algorithm searches for a feasible path from the starting point to the ending point under constraints through random sampling and tree expansion. Then, a cost function is used to evaluate and optimize candidate paths, ultimately outputting a smooth and safe path. The cost function is used to evaluate the quality of the path during the path generation process, and its expression is: Cost=w1×PathLength+w2×HazardProximity+w3×EnergyConsumption Where PathLength is the path length; HazardProximity is the reciprocal of the closest distance between the path and a known obstacle (the closer the distance, the greater the risk); EnergyConsumption is the energy consumption estimated based on the path length and flight altitude; w1, w2, and w3 are weighting coefficients used to balance the importance of different optimization objectives in path planning, and their specific values ​​can be calibrated experimentally or preset according to the task type, for example: When the mission requires prioritizing flight safety, the value of w2 (obstacle avoidance weight) can be increased; When the mission requires prioritizing flight efficiency, the value of w1 (path length weight) can be increased; When the task requires prioritizing energy saving and battery life, the value of w3 (energy consumption weight) can be increased.

[0089] In one embodiment, for routine aerial photography missions, the values ​​of w1, w2, and w3 are 0.4, 0.4, and 0.2, respectively; for energy-saving cruise missions, the values ​​are 0.3, 0.3, and 0.4, respectively; and for high-risk area flight missions, the values ​​are 0.2, 0.6, and 0.2, respectively. Those skilled in the art can reasonably adjust the weighting coefficients according to the actual application scenario.

[0090] In one embodiment, the Dynamic Window Method (DWA) is used instead of the Fast Random Search Tree (RRT*) algorithm. The DWA algorithm has the advantages of fast computation speed and strong real-time performance in local path planning, and is suitable for scenarios with many dynamic obstacles. In this embodiment, the weight coefficients in the cost function are set as follows: w1=0.35, w2=0.45, w3=0.20, to prioritize flight safety.

[0091] According to the S220 rules, the original altitude strategy is recalculated or reused to generate a new waypoint sequence WP_new, and the WP_new flight mission package is sent to the UAV.

[0092] Sending the MAV_CMD_DO_RESUME command to the drone enables the drone's flight control module to smoothly connect to the new route, accelerate from the hovering state, and fly towards the new destination.

[0093] At this moment, the vehicle's map interface immediately updates, displaying the new planned route and real-time progress.

[0094] Step S400: Target Hovering and Intelligent Information Enhancement The drone hovers after reaching the target point. When the user clicks on a target building in the video frame, the vision processing module captures the current frame and the drone's pose. Based on a pinhole camera model, it performs back-projection calculations to estimate the geodetic coordinates (Lat_click, Lon_click) of the clicked object. Then, based on the following formula, it calculates the image pixel coordinates (u,v) using the camera intrinsic parameter K. inv The depth estimate d, the UAV attitude rotation matrix R, and the translation vector T are transformed into the geographic coordinate system [X]. c ,Y c Z c The depth estimate d can be calculated based on the drone's current altitude relative to the ground and the building height data. [ X c , Y c , Z c ] T =R× ( K -1 × [ u , v ,1] T ×d )) +T Compare the coordinates with the video screenshot Figure 1The data is uploaded to a cloud-based building information big data service. The cloud uses both coordinate matching and image feature matching for dual verification, returning detailed building information such as name, address, opening hours, user ratings, and a brief description, which is then sent to the vehicle-mounted system. The cloud has a built-in building information database that establishes a link between visual features and building point-of-purchase (POI) information, storing data such as the building's geographical location, name, function, user ratings, and historical information. The system receives images or coordinates of buildings uploaded by the vehicle-mounted system and returns their basic information.

[0095] The vehicle-mounted system displays building information on the video feed using information cards or AR overlays.

[0096] Step S500: Task completion and automatic homing.

[0097] When the user issues a return-to-home command or the system detects low battery, the mission planning and decision-making module plans a safe return-to-home path. The drone autonomously returns and lands precisely on the roof of the vehicle, connecting to the charging device inside the cabin via contact points to automatically begin charging.

[0098] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0099] Example 4 This embodiment adds a safety monitoring module to the basis of embodiment 1. It runs continuously on the task planning and decision-making module or the drone to realize communication link monitoring and / or power monitoring and / or electronic fence monitoring functions. The communication link monitoring function includes: continuous monitoring of signal strength (RSSI) and packet loss rate. When the signal strength is below -85dBm or the packet loss rate exceeds 20%, the signal loss response strategy is automatically triggered: the UAV increases its altitude to attempt to restore the link. If it cannot be restored within 30 seconds, the return-to-home procedure is automatically executed.

[0100] The power monitoring function includes: dynamically calculating the latest return point based on the distance between the drone's current location and the return point, current power consumption, and remaining power. When the remaining power is lower than 120% of the power required to complete the return, the system will display a warning message; when the remaining power is lower than 100% of the power required to complete the return, the system will automatically initiate the forced return process.

[0101] The electronic fence monitoring function includes: preset flight boundaries (electronic fence) for drones; continuous verification of the drone's position; and automatic triggering of a hovering command and sending an alarm to the vehicle's infotainment system if the drone detects that it is about to exceed the electronic fence boundary (less than 50 meters from the boundary).

[0102] Example 5 This embodiment is basically the same as Embodiment 1, except that the visual processing and rendering module supports offline mode. The vehicle-mounted system pre-caches offline POI map data within a certain range. When there is no network connection, it retrieves data from the local database through image feature matching or coordinate matching, thereby achieving information enhancement in the offline environment.

[0103] Example 6 A computer program product includes a computer program / instructions that, when executed by a processor, implement the various steps of the method described in this invention.

[0104] Example 7 This invention also provides a non-transitory computer-readable storage medium storing a computer program. The computer program includes program instructions that, when executed by a processor, implement the various steps of the method described in this invention, which will not be elaborated further here.

[0105] The computer-readable storage medium can be the data transmission apparatus or the internal storage unit of a computer device provided in any of the foregoing embodiments, such as the hard disk or memory of the computer device. The computer-readable storage medium can also be the external storage device of the computer device, such as the plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on the computer device.

[0106] Furthermore, the computer-readable storage medium may include both internal storage units and external storage devices of the computer device. The computer-readable storage medium is used to store the computer program and other programs and data required by the computer device. The computer-readable storage medium may also be used to temporarily store data that is to be output or has already been output.

[0107] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0108] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0109] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0110] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0111] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

Claims

1. A vehicle-mounted unmanned aerial vehicle (UAV) intelligent pointing flight control system, characterized in that, include: The map and interaction module is used to receive the target point selected by the user on the map interface; The geographic data service module is used to provide geographic data of the flight area where the target point is located based on the target point; The mission planning and decision-making module is used to use static obstacle data in the geographic data of the flight area as obstacle avoidance constraints, generate a two-dimensional plane path from the current position of the UAV to the target point using a path generation algorithm, assign a flight altitude to each waypoint on the two-dimensional plane path to form a three-dimensional waypoint sequence, and generate a flight mission based on the three-dimensional waypoint sequence and preset flight parameters.

2. The vehicle-mounted unmanned aerial vehicle (UAV) intelligent pointing flight control system as described in claim 1, characterized in that, The map and interaction module is also used to receive the flight altitude value input by the user on the map interface; the mission planning and decision module generates the flight mission containing a three-dimensional waypoint sequence based on the target point and the flight altitude value.

3. The vehicle-mounted unmanned aerial vehicle (UAV) intelligent pointing flight control system as described in claim 1 or 2, characterized in that, The map and interaction module is also used to receive the flight mission type selected by the user; the mission planning and decision-making module generates the flight altitude value according to the preset rules corresponding to the flight mission type.

4. The vehicle-mounted unmanned aerial vehicle intelligent pointing flight control system as described in claim 3, characterized in that, The map and interaction module is also used to receive the altitude mode selected by the user. When the altitude mode is automatic, the task planning and decision module sends the coordinates and query radius of the target point to the geographic data service module, receives the geographic data of the flight area returned by the geographic data service module, determines the ground elevation of the target point and the elevation of the highest building on the planned path based on the geographic data, and generates the flight altitude value based on the ground elevation of the target point, the elevation of the highest building on the planned path, and the additional offset determined according to the flight task type selected by the user.

5. The vehicle-mounted unmanned aerial vehicle (UAV) intelligent pointing flight control system as described in claim 1, characterized in that, It also includes a visual processing and rendering module, which is used to receive and decode the video stream uploaded by the drone, and overlay the flight status information onto the decoded video stream for display; it is also used to receive the target building clicked by the user on the decoded video stream, estimate the geographic coordinates of the target building based on the pixel coordinates of the target building, and overlay the information of the target building onto the video stream in an augmented reality manner for display.

6. The vehicle-mounted unmanned aerial vehicle (UAV) intelligent pointing flight control system as described in claim 5, characterized in that, The visual processing and rendering module is also used for: Receive the target building clicked by the user on the decoded video stream; Based on the location of the target building, the flight status information, and the current attitude of the UAV's gimbal, the geographic coordinates of the target building are estimated by back projection calculation based on the pinhole camera model. The geographic coordinates of the target building are sent to the geographic data service module, and the information of the target building returned by the geographic data service module is received. The information of the target building is displayed by overlaying it onto the real-time video stream in an augmented reality manner.

7. The vehicle-mounted unmanned aerial vehicle (UAV) intelligent pointing flight control system as described in claim 1, characterized in that, The task planning and decision-making module is also used to convert various interactive commands input by the user into corresponding control commands and send them to the drone during the autonomous flight of the drone.

8. The vehicle-mounted unmanned aerial vehicle intelligent pointing flight control system as described in claim 7, characterized in that, When the interaction command is a zoom command, the task planning and decision module is also used to recognize the user's two-finger pinch gesture in the video display area, calculate the zoom factor based on the change in the distance between the two finger touch points, map the zoom factor to a camera zoom ratio command, and send the zoom command to the drone.

9. The vehicle-mounted unmanned aerial vehicle intelligent pointing flight control system as described in claim 7, characterized in that, When the interaction command is a gimbal control command, the task planning and decision-making module is also used to identify the user's operation on the virtual joystick or directional buttons, convert the operation into the angular velocity of the gimbal's pitch axis and yaw axis, and continuously send the gimbal speed control command containing the angular velocity to the UAV.

10. The vehicle-mounted unmanned aerial vehicle intelligent pointing flight control system as described in any one of claims 7 to 9, characterized in that, When the interaction command is a destination change command, the task planning and decision-making module is further configured to: send a pause command to the drone after receiving the new target point selected by the user on the map interface; obtain the real-time position of the drone, and generate a new path with the real-time position as the starting point, the new target point as the ending point, and static obstacle data in the geographic data and local obstacle data transmitted back by the drone as obstacle avoidance constraints.

11. The vehicle-mounted unmanned aerial vehicle intelligent pointing flight control system as described in claim 10, characterized in that, The task planning and decision-making module uses the fast random search tree algorithm or the dynamic window method to generate new paths, and evaluates candidate paths using a cost function, the expression of which is: Cost = w1×PathLength+w2×HazardProximity+w3×EnergyConsumption Wherein, PathLength is the length of the new path, HazardProximity is the reciprocal of the nearest distance between the new path and the static obstacle in the geographic data, EnergyConsumption is the energy consumption estimated based on the new path length and flight altitude, and w1, w2, and w3 are weighting coefficients preset according to mission requirements.

12. A vehicle-mounted unmanned aerial vehicle (UAV) intelligent pointing-and-fly control method for the system as described in claim 1, characterized in that, include: Receive the target point selected by the user on the map interface; Provide geographical data of the flight area where the target point is located based on the target point; Using static obstacle data from the geographic data of the flight area as obstacle avoidance constraints, a path generation algorithm is used to generate a two-dimensional planar path from the current position of the UAV to the target point. Each waypoint on the two-dimensional planar path is assigned a flight altitude to form a three-dimensional waypoint sequence, and a flight mission is generated based on the three-dimensional waypoint sequence and preset flight parameters.

13. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent pointing flight control method for vehicle-mounted unmanned aerial vehicles as described in claim 12.

14. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the intelligent guided flight control method for vehicle-mounted unmanned aerial vehicles as described in claim 12.

Images