Agricultural unmanned aerial vehicle first view picture semantic augmented reality guidance method and software program product
Patent Information
- Application Number
- CN202611130918.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-28
- Publication Date
- 2026-10-30
AI Technical Summary
[0005]鉴于上述的分析，本发明实施例旨在解决现有增强现实方案无法在农业作业阶段主动在无人机第一视角画面中提供决策信息的问题
1、本发明方案不再将农业无人机任务对象以普通点、线或航点形式直接叠加到第一视角画面，而是根据农业无人机当前所处的作业阶段，将底层任务对象动态转换为具有明确作业语义的目标语义化引导构件，并将目标语义化引导构件在当前第一视角画面中显示，从而使飞手直接感知“当前该如何飞”“飞到哪里算有效”“怎样切入或回接下一条作业带”。
Smart Images
Figure CN123015913A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of drones, and in particular to a semantic augmented reality guidance method and software product for first-person view of agricultural drones. Background Technology
[0002] When agricultural quadcopter drones are used for plant protection spraying, sowing, field inspection and plot boundary operations, the pilots usually rely on first-person view, ground station map, mission route and experience to complete operations such as entering the field, flying along the zone, turning around, bypassing obstacles and returning to the field.
[0003] Most existing augmented reality (AR) solutions simply overlay points, lines, waypoints, paths, or boundaries onto the current image. While they can provide some auxiliary information, they are still significantly insufficient in agricultural operation scenarios. The information displayed on the current image is usually not the information that the drone pilot is currently concerned with. For example, during the spray pattern operation phase, the drone pilot is concerned with whether the drone has deviated from the center of the spray pattern. If only ordinary flight paths or ordinary waypoints are displayed, no decision-making information can be provided.
[0004] Therefore, there is an urgent need for a technical solution that can proactively provide decision-making information in the first-person view of drones during agricultural operations. Summary of the Invention
[0005] Based on the above analysis, the embodiments of the present invention aim to solve the problem that existing augmented reality solutions cannot proactively provide decision-making information in the first-person view of the drone during agricultural operations.
[0006] In a first aspect, embodiments of the present invention provide a semantic augmented reality guidance method for first-person view footage of agricultural drones, characterized by comprising the following steps: Step S100: Determine the current operational stage of the UAV based on its current first-person view, current pose, and task information; Step S200: Based on the current operational stage of the UAV, select a semantic guidance component of the target type from a preset set of semantic guidance components; Step S300: Generate a target semantic guidance component based on the target type of the selected semantic guidance component and the current task parameters of the UAV, wherein the current task parameters of the UAV are determined by the current pose of the UAV and the task information; and Step S400: Display the target semantic guidance component in the current first-view view of the UAV.
[0007] Based on a further improvement of the above method, in response to the current operation stage of the UAV being the operation along a belt, the target type of the selected semantic guidance component is an operation corridor. The space of the operation corridor is represented by two solid lines parallel to the operation direction and a central dashed line equidistant from the two solid lines. The central dashed line is the center line of the target operation belt.
[0008] Based on a further improvement of the above method, the work corridor is determined in the following way: The effective area of the work corridor is determined according to the following formula. :
[0009] in, This is the current location of the drone. Let the center line of the i-th target work zone be _____. for arrive lateral distance, This refers to the spray width during drone spraying operations. This is the wind deflection correction amount during drone spraying operations. , , Obtained from the drone operation task information.
[0010] Based on a further improvement of the above method, in response to the current operation stage of the UAV being the approach stage or the turn-around transition stage, the target type of the selected semantic guidance component is the entry frame, the space of the entry frame is represented by a quadrilateral composed of four vertices, and the space of the entry frame represents the spatial entrance of the corresponding target operation zone.
[0011] Based on a further improvement of the above method, the entry frame is determined in the following way: The four vertices of the entry frame are determined according to the following formula. , , , :
[0012]
[0013]
[0014]
[0015] in, The location of the center point of the entrance to the target work zone. Let be the normal vector of the target assignment. The width of the door frame. The height of the door frame. , Obtained from the drone operation task information. and It is determined by the drone operation task information.
[0016] Based on a further improvement to the above method, in response to the current operating phase of the drone being a turnaround transition phase, the entry frame is displayed when the following equation holds true:
[0017] in, The angle between the drone's return direction and the target work zone direction. To allow for the included angle threshold, The position and mission information of the UAV are determined.
[0018] Based on a further improvement of the above method, in response to the current operation phase of the UAV being the obstacle avoidance phase, the target type of the selected semantic guidance component is an obstacle avoidance reconnection channel. The obstacle avoidance reconnection channel space is represented as a feasible area that can still reconnect to the target operation zone after obstacle avoidance, and the obstacle avoidance reconnection channel is displayed when the following formula is true:
[0019] in, To cover the cost of reconnecting to the feasibility, To set the reconnection cost threshold, Determined as follows:
[0020] in, The lateral deviation of the obstacle bypass reconnection point relative to the target work zone. To account for the error in the drone's heading angle during obstacle avoidance and reconnection. This refers to the distance of missed spraying or respraying caused after bypassing an obstacle and reconnecting. These are the weighting coefficients. , , The position and mission information of the UAV are determined.
[0021] Based on a further improvement of the above method, in response to the current operation stage of the UAV being the return stage, the target type of the selected semantic guidance component is a return funnel. The space of the return funnel is represented as a three-dimensional funnel-shaped spatial region that gradually converges from far to near. The boundaries on both sides of the return funnel are jointly determined by the return target point and the extended boundary. The extended boundary is determined by the current pose of the UAV and the operation task information.
[0022] Based on a further improvement of the above method, the semantic guidance component in step S200 includes multiple types, and step S300 further includes: The respective target semantic guidance components are prioritized according to at least one of the following dimensions: Semantic targeting guides the task relevance of components; Risk relevance of target semantic guidance components; The timing urgency of semantically guided components; The visual display of semantically guided components incurs congestion costs. Step S400 includes: The top K semantic guidance components with the highest priority are displayed in the current first-view view of the UAV, where K is an integer greater than or equal to 1.
[0023] Secondly, embodiments of the present invention provide a software program product including program code instructions. When the program product is executed by a computing device, the program code instructions cause the computing device to perform the method described in any of the first aspects of the present invention.
[0024] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects: 1. The present invention no longer directly overlays the agricultural drone task object onto the first-person view as a regular point, line or waypoint. Instead, based on the current operation stage of the agricultural drone, it dynamically converts the underlying task object into a target semantic guidance component with clear operational semantics and displays the target semantic guidance component in the current first-person view, so that the pilot can directly perceive "how to fly now", "where to fly to is effective" and "how to switch to or rejoin the next operation zone".
[0025] 2. In the present invention, since ordinary task objects are converted into target semantic guidance components, the pilot no longer sees only abstract points and lines, but directly sees "task corridor", "entry frame" or "obstacle bypass and reconnection channel" with operational significance.
[0026] 3. In the present invention, since the target semantic guidance component type is dynamically switched according to the operation stage, AR information can adapt to different scenarios such as entering the site, following the belt, turning around, bypassing obstacles, and returning to the site, thereby improving the targeting of auxiliary display.
[0027] 4. In the present invention, since an effective working corridor is generated based on the spray width parameters, the drone operator can directly determine whether the drone has deviated from the effective spraying area, thereby reducing the risk of missed spraying and overspraying.
[0028] 5. In the present invention, since a reconnection gate or obstacle bypass reconnection channel is generated based on the reconnection cost, the pilot can more intuitively understand the recovery path after turning around or bypassing obstacles, thereby improving operational efficiency and safety.
[0029] 6. In the present invention, due to the introduction of priority compression display, only the semantic guidance components with the most decision-making value are retained in the current screen, reducing information congestion and cognitive burden.
[0030] In this invention, the above-described technical solutions can be combined with each other to achieve more preferred combinations. Other features and advantages of this invention will be set forth in the following description, and some advantages may become apparent from the description or be learned by practicing the invention. The objects and other advantages of this invention can be realized and obtained from what is particularly pointed out in the description and drawings. Attached Figure Description
[0031] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. Figure 1 A flowchart illustrating a semantic augmented reality guidance method for first-person view of agricultural drones according to an embodiment of the present invention is shown.
[0032] Figure 2 An example of a work corridor according to an embodiment of the present invention is shown.
[0033] Figure 3 An example of a cut-in door frame / reconnection door according to an embodiment of the present invention is shown.
[0034] Figure 4 An example of a bypass reconnection channel according to an embodiment of the present invention is shown. Detailed Implementation
[0035] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.
[0036] Agricultural quadcopter drone operations are essentially "strip-covering operations." The pilot's focus is not on a single target point, but rather on whether they are currently within an effective work zone, whether there will be missed spraying or overspraying, and whether they can smoothly rejoin the next work zone. Existing AR (Augmented Reality) systems typically only display the original task objects directly, without converting them into "semantic guidance components" that the pilot can directly see. The information the pilot truly needs to focus on differs at different stages of the operation. For example, during the approach phase, the focus is on where to enter the next work zone; during the zone alignment phase, the focus is on whether they are deviating from the center of the spray pattern; during the turnaround phase, the focus is on whether they can safely enter the next work zone; and during the obstacle avoidance phase, the focus is on whether they can rejoin the original work zone after detouring. Displaying all objects simultaneously would lead to information overload; displaying only ordinary flight paths or waypoints fails to convey crucial decision-making information such as the effective area of the work zone, the entry threshold, the turnaround envelope, and the rejoinability after obstacle avoidance.
[0037] To address these issues, this invention provides a semantic augmented reality guidance method for first-person view footage from agricultural drones. Figure 1 A flowchart illustrating a semantic augmented reality guidance method for first-person view of agricultural drones according to an embodiment of the present invention is shown. Figure 1 As shown, this semantic augmented reality guidance method for first-person view of agricultural drones includes: Step S100: Determine the current operation stage of the drone based on the drone's current first-person view, current pose, and task information.
[0038] In this embodiment, the current first-person view of the UAV, its current attitude, current position, mission flight path, spray parameters, boundary information of the target object, and obstacle information can be acquired first. In this embodiment, the current attitude includes the current position; the mission information includes the mission flight path, spray parameters, boundary information of the target object, and obstacle information. Then, based on the UAV's current first-person view, current attitude, and mission information, the current flight state and mission state of the UAV can be determined, and the current mission stage can be identified based on the current flight state and mission state. Specifically, the current mission stage can be determined based on the UAV's current speed direction, its relative position to the current target work zone, its heading angle, and the mission state.
[0039] In this embodiment, the current operational phase of the drone Selected from the following set of states:
[0040] in, This indicates the entry phase. Indicates the phase of operation along the belt. This indicates a transitional phase where the vehicle turns around. Indicates the obstacle avoidance phase. This indicates a return or resupply phase.
[0041] Step S200: Based on the current operational stage of the UAV, select a semantic guidance component of the target type from a preset set of semantic guidance components.
[0042] In this embodiment, the preset set of semantic guidance components includes the following types of semantic guidance components: Work corridor, entry frame, obstacle bypass and reconnection channel, return funnel.
[0043] When the current operation phase of the drone is along a corridor, the target type of the selected semantic guidance component is "Operation Corridor". When the current operation phase of the drone is approaching, the target type of the selected semantic guidance component is "Entry Box". When the current operation phase of the drone is turning around, the target type of the selected semantic guidance component is "Entry Box". When the current operation phase of the drone is obstacle avoidance, the target type of the selected semantic guidance component is "Obstacle Avoidance Reconnection Channel". When the current operation phase of the drone is returning or resupplying, the target type of the selected semantic guidance component is "Return Funnel".
[0044] Step S300: Generate a target semantic guidance component based on the target type of the selected semantic guidance component and the current task parameters of the UAV.
[0045] In this embodiment, the current task parameters of the UAV are determined by the UAV's current pose and task information. In this embodiment, the current task parameters of the UAV are related to the target type of the selected semantic guidance component, and these parameters are necessary for generating the target semantic guidance component. For example, if the target type of the selected semantic guidance component is a task corridor, then the current task parameters of the UAV are necessary for generating the target task corridor, including the current position, current attitude, spray width parameters, boundary information of the task object, etc.; if the target type of the selected semantic guidance component is an obstacle avoidance and reconnection channel, then the current task parameters of the UAV are necessary for generating the target obstacle avoidance and reconnection channel, including the current position, current attitude, obstacle information, etc.
[0046] In some embodiments, if the selected semantic guidance component is a task corridor, a target task corridor can be generated based on the current task parameters of the UAV. The spatial representation of the target task corridor is two solid lines parallel to the task direction and a central dashed line equidistant from the two solid lines, wherein the central dashed line is the center line of the target task zone. Figure 2 An example of a target work corridor according to an embodiment of the present invention is shown. Figure 2As shown, the two parallel solid lines on the left and right represent the effective spray width corridor, the dashed line in the center represents the center line of the target operation zone, the rounded rectangle surrounding the target operation corridor represents the boundary of the drone's first-person view, and the solid circle represents the drone's current position.
[0047] In some embodiments, the target work corridor can be determined in the following ways: The effective area of the target work corridor is determined using the following formula. :
[0048] in, This is the current location of the drone. Let the center line of the i-th target work zone be _____. for arrive lateral distance, This refers to the spray width during drone spraying operations. This is the wind deflection correction amount during drone spraying operations. , , It can be obtained from drone operation mission information.
[0049] In some embodiments, the operational deviation can be further determined based on the current operational task parameters of the UAV. Specifically, the operational deviation can be determined according to the following formula. :
[0050] When the following formula is met, a deviation warning will be displayed or the target work corridor display status will be switched to warning status:
[0051] In some embodiments, the current operation phase of the UAV is the approach phase or the turn-around transition phase, and the target type of the selected semantic guidance component is the entry frame. The space of the target entry frame is represented as a quadrilateral composed of four vertices, and the space of the target entry frame represents the spatial entrance of the corresponding target operation zone. Figure 3 An example of a target cut-in door frame / reconnection door according to an embodiment of the present invention is shown. Figure 3 As shown, the quadrilateral formed by the four vertices is the target entry frame (corresponding to the entry phase) or the target return gate (corresponding to the turn-around transition phase). The solid circle represents the current position of the drone. The direction of the arrow indicates the deviation between the drone's current actual working direction and the pre-planned working direction on the target working zone. The rounded rectangle surrounding the entry frame / return gate represents the boundary of the drone's first-person view.
[0052] In some embodiments, the target entry frame is determined in the following manner: The four vertices of the target entry frame are determined using the following formula. , , , :
[0053]
[0054]
[0055]
[0056] in, The location of the center point of the entrance to the target work zone. Let be the normal vector of the target assignment. The width of the door frame. The height of the door frame. , Obtained from drone operation mission information. and The width and height of the gate frame are determined by the drone operation mission information. Specifically, the width and height of the gate frame can be generated based on the spray width, drone body size, wind deflection correction, and the drone's current flight status.
[0057] In some embodiments, the current operational phase of the drone is a turnaround transition phase, and the target entry box is displayed when the following expression holds true:
[0058] in, The angle between the drone's return direction and the target work zone direction. To allow for the included angle threshold, The drone's current pose and mission information are used to determine its trajectory. The drone's return direction is its current actual operational direction.
[0059] In some embodiments, the current operation phase of the UAV is the obstacle avoidance phase, the target type of the selected semantic guidance component is the obstacle avoidance reconnection channel, and the target obstacle avoidance reconnection channel space is represented as the feasible area that can still reconnect to the target operation zone after bypassing the obstacle. Figure 4 An example of a target obstacle bypass reconnection channel according to an embodiment of the present invention is shown. Figure 4 The channel in the diagram represents a feasible area where the target work zone can still be reconnected after bypassing the obstacle. The solid circle represents the obstacle area, and the rounded rectangle surrounding the obstacle reconnection channel represents the boundary of the UAV's first-person view.
[0060] In some embodiments, the target obstacle bypass reconnection channel is displayed when the following expression holds:
[0061] in, To cover the cost of reconnecting to the feasibility, To set the reconnection cost threshold, Determined as follows:
[0062] in, The lateral deviation of the obstacle bypass reconnection point relative to the target work zone. To account for the error in the drone's heading angle during obstacle avoidance and reconnection. This refers to the distance of missed spraying or respraying caused after bypassing an obstacle and reconnecting. These are the weighting coefficients. , , The position and mission information of the UAV are determined.
[0063] The significance of the above conditions lies in not only considering whether it is possible to return in terms of spatial geometry, but also comprehensively taking into account reconnection deviation, heading error, and operational quality loss.
[0064] In some embodiments, the current operation phase of the UAV is the return phase or the resupply phase, the target type of the selected semantic guidance component is the return funnel, the space of the target return funnel is represented as a three-dimensional funnel-shaped spatial region that gradually converges from far to near, the boundaries on both sides of the target return funnel are jointly determined by the return target point and the extended boundary, and the extended boundary is determined by the current pose of the UAV and the operation task information.
[0065] In this embodiment, the center point of the return target is set as... The direction of the encore guidance is The entrance width is The remote extension width is The guide length is Then the two sides of the target return funnel can be determined by the target point and the extended boundary.
[0066] In this embodiment, the extended boundary is not a fixed value, but is generated jointly based on the guidance path output by the return planning algorithm, the passable space within the current sensing area, and the safe convergence range near the return target. Specifically, the return planning algorithm first generates a return guidance path or return direction based on the current UAV position, the return target position, boundary constraints, and obstacle information; then, within the forward region of the return guidance path, it expands laterally on both sides of the path, combining the passable free space range within the current sensing area, forming the outward boundary of the funnel; simultaneously, a safe circular convergence range is set near the return target point or resupply point, so that the extended boundary gradually converges to the width of the return entrance as it approaches the target. This creates a funnel-shaped guiding area that gradually converges from far to near.
[0067] In this embodiment, the far-end extended width The range of the funnel can be adaptively determined based on the current sensing area width, obstacle distribution, boundary margin, and return speed. When the surrounding passable area is wide, the funnel extension range is appropriately increased. When the boundary constraints are strong or the obstacles are dense, the extension width is reduced accordingly to ensure that the funnel boundary is located within the safe passable area.
[0068] In this embodiment, the safe circular convergence range is used to describe the spatial region near the return target that allows stable entry. The far boundary of the return funnel continuously converges toward this safe convergence range, so that what the pilot sees from the first-person perspective is not a single return point, but a return guidance area that is "accessible, convergent, and safely aligned".
[0069] Step S400: Display the target semantic guidance component in the current first-view view of the UAV.
[0070] In this embodiment, the spatial representation of the target semantic guidance component can be displayed on the image plane through a projection function:
[0071] in, Represents the spatial geometric description of the target semantic guiding component. Represents the coordinates of the projected image. This represents the current position vector of the drone. Represents the attitude quaternion of the drone. This represents the projection function of a spatial component onto the image plane.
[0072] In this embodiment, after projection, normalization and display clipping are further performed to ensure that the display position of the semantic component in the first-view view is consistent with the current flight view. Figures 2 to 4 An example is shown where the corresponding target semantic guidance component is displayed in the drone's current first-person view. Figure 2 In the image, the rounded rectangle surrounding the work corridor represents the boundary of the drone's first-person view. Figure 3 In the image, the rounded rectangle surrounding the cut-in gate / reconnection gate represents the boundary of the drone's first-person view. Figure 4 In the image, the rounded rectangle surrounding the obstacle return channel represents the boundary of the drone's first-person view.
[0073] In some embodiments, the semantic guidance component in step S200 includes multiple types, such as a work corridor, a cut-in gate, and an obstacle bypass reconnection channel, and step S300 further includes: The respective target semantic guidance components are prioritized according to at least one of the following dimensions: Semantic targeting guides the task relevance of components; Risk relevance of target semantic guidance components; The timing urgency of semantically guided components; The visual display of the target semantic guidance component incurs congestion costs.
[0074] Task relevance, risk relevance, time urgency, and display congestion cost can all be calculated by normalizing the current UAV status, task planning results, environmental perception results, and image projection results. Task relevance characterizes the degree of consistency between the semantic component of the target waiting to be displayed and the current operation stage and the current task objective. Its value is calculated based on the matching relationship between the current operation stage and the applicable stage of the component, the distance between the UAV and the corresponding operation object of the component (e.g., farmland), and whether the target belongs to the current operation objective. The higher the matching degree and the closer the distance, the greater the task relevance.
[0075] Risk relevance characterizes the extent to which failing to prioritize displaying the semantic component of the target could lead to increased pilot operational deviations, operational errors, or increased traffic risks. Its value is determined by considering the relative positional deviation between the UAV and the corresponding target area, the deviation between the current heading and the target guidance direction, and the presence of boundary constraints, obstacles, or return restrictions within the current area. A greater deviation, stronger constraints, and higher risk result in a higher risk relevance.
[0076] The timing urgency is used to characterize the immediacy requirement of the action corresponding to the semantic component of the target. Its value is calculated based on the remaining distance or estimated remaining arrival time from the current position of the UAV to the trigger position such as the entry point, reconnection point, or return entry point. The shorter the remaining time or the smaller the remaining distance, the greater the timing urgency.
[0077] Display congestion cost is used to characterize the degree of information occlusion caused by the projection of the target semantic component onto the current first-view image. Its value is calculated based on the area of the component's projection region, the overlap ratio with other displayed components, and the degree of occlusion of the key viewing area in the center of the image. The more severe the occlusion, the greater the display congestion cost.
[0078] And step S400 includes: The top K semantic guidance components with the highest priority are displayed in the current first-view view of the UAV, where K is an integer greater than or equal to 1.
[0079] For multiple candidate semantic components, priority scoring can be performed as follows:
[0080] in, For task relevance, For risk relevance, For time urgency, To illustrate the cost of congestion, This represents the weighting coefficient. This design avoids overlaying too much irrelevant information on the same screen, highlighting the most valuable semantic components of the task at the current stage.
[0081] The following are some example numerical ranges for the relevant parameters in this invention: 1. Spray width 3m to 12m; 2. Wind deflection correction amount : 0.1m to 1.0m; 3. Door frame width : 2m to 8m; 4. Door frame height : 1m to 5m.
[0082] Secondly, embodiments of the present invention provide a software program product including program code instructions. When the program product is executed by a computing device, the program code instructions cause the computing device to perform the method described in any of the first aspects of the present invention.
[0083] Compared with the prior art, the embodiments of the present invention can achieve at least one of the following beneficial effects: 1. The present invention no longer directly overlays the agricultural drone task object onto the first-person view as a regular point, line or waypoint. Instead, based on the current operation stage of the agricultural drone, it dynamically converts the underlying task object into a target semantic guidance component with clear operational semantics and displays the target semantic guidance component in the current first-person view, so that the pilot can directly perceive "how to fly now", "where to fly to is effective" and "how to switch to or rejoin the next operation zone".
[0084] 2. In the present invention, since ordinary task objects are converted into target semantic guidance components, the pilot no longer sees only abstract points and lines, but directly sees "task corridor", "entry frame" or "obstacle bypass and reconnection channel" with operational significance.
[0085] 3. In the present invention, since the target semantic guidance component type is dynamically switched according to the operation stage, AR information can adapt to different scenarios such as entering the site, following the belt, turning around, bypassing obstacles, and returning to the site, thereby improving the targeting of auxiliary display.
[0086] 4. In the present invention, since an effective working corridor is generated based on the spray width parameters, the drone operator can directly determine whether the drone has deviated from the effective spraying area, thereby reducing the risk of missed spraying and overspraying.
[0087] 5. In the present invention, since a reconnection gate or obstacle bypass reconnection channel is generated based on the reconnection cost, the pilot can more intuitively understand the recovery path after turning around or bypassing obstacles, thereby improving operational efficiency and safety.
[0088] 6. In the present invention, due to the introduction of priority compression display, only the semantic guidance components with the most decision-making value are retained in the current screen, reducing information congestion and cognitive burden.
[0089] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.
Claims
1. A semantic augmented reality guidance method for first-person view footage of agricultural drones, characterized in that, Includes the following steps: Step S100: Determine the current operational stage of the UAV based on its current first-person view, current pose, and task information; Step S200: Based on the current operational stage of the UAV, select a semantic guidance component of the target type from a preset set of semantic guidance components; Step S300: Generate a target semantic guidance component based on the target type of the selected semantic guidance component and the current task parameters of the UAV, wherein the current task parameters of the UAV are determined by the current pose and task information of the UAV. as well as Step S400: Display the target semantic guidance component in the current first-view view of the UAV.
2. The method according to claim 1, characterized in that, In response to the current operation phase of the UAV being the operation phase along the belt, the target type of the selected semantic guidance component is the operation corridor. The space of the operation corridor is represented by two solid lines parallel to the operation direction and a central dashed line equidistant from the two solid lines. The central dashed line is the center line of the target operation belt.
3. The method according to claim 2, characterized in that, The work corridor is determined in the following way: The effective area of the work corridor is determined according to the following formula. : in, This is the current location of the drone. Let the center line of the i-th target work zone be _____. for arrive lateral distance, This refers to the spray width during drone spraying operations. This is the wind deflection correction amount during drone spraying operations. , , Obtained from the drone operation task information.
4. The method according to claim 1, characterized in that, In response to the current operation phase of the UAV being either the approach phase or the turnaround transition phase, the target type of the selected semantic guidance component is the entry frame, and the space of the entry frame is represented by a quadrilateral composed of four vertices, and the space of the entry frame represents the spatial entrance of the corresponding target operation zone.
5. The method according to claim 4, characterized in that, The entry frame is determined in the following way: The four vertices of the entry frame are determined according to the following formula. , , , : in, The location of the center point of the entrance to the target work zone. Let be the normal vector of the target assignment. The width of the door frame. The height of the door frame. , Obtained from the drone operation task information. and It is determined by the drone operation task information.
6. The method according to claim 5, characterized in that, The entry frame is displayed in response to the drone's current operational phase being a turnaround transition phase, and when the following expression holds true: in, The angle between the drone's return direction and the target work zone direction. To allow for the included angle threshold, The position and mission information of the UAV are determined.
7. The method according to claim 1, characterized in that, In response to the fact that the current operation phase of the UAV is the obstacle avoidance phase, the target type of the selected semantic guidance component is the obstacle avoidance reconnection channel. The obstacle avoidance reconnection channel space is represented as a feasible area that can still reconnect to the target operation zone after bypassing the obstacle, and the obstacle avoidance reconnection channel is displayed when the following formula is true: in, To cover the cost of reconnecting to the feasibility, To set the reconnection cost threshold, Determined as follows: in, The lateral deviation of the obstacle bypass reconnection point relative to the target work zone. To account for the error in the drone's heading angle during obstacle avoidance and reconnection. This refers to the distance of missed spraying or respraying caused after bypassing an obstacle and reconnecting. These are the weighting coefficients. , , The position and mission information of the UAV are determined.
8. The method according to claim 1, characterized in that, In response to the current operation phase of the UAV being the return phase, the target type of the selected semantic guidance component is a return funnel. The space of the return funnel is represented as a three-dimensional funnel-shaped spatial region that gradually converges from far to near. The boundaries on both sides of the return funnel are jointly determined by the return target point and the extended boundary. The extended boundary is determined by the current pose of the UAV and the operation task information.
9. The method according to any one of claims 1-8, characterized in that, The semantic guidance component in step S200 includes multiple types, and step S300 further includes: The respective target semantic guidance components are prioritized according to at least one of the following dimensions: Semantic targeting guides the task relevance of components; Risk relevance of target semantic guidance components; The timing urgency of semantically guided components; The visual display of semantically guided components incurs congestion costs. Step S400 includes: The top K semantic guidance components with the highest priority are displayed in the current first-view view of the UAV, where K is an integer greater than or equal to 1.
10. A software program product comprising program code instructions that, when executed by a computing device, cause the computing device to perform the method of any one of claims 1 to 9.