WO2020024134A1 - 轨迹切换的方法和装置 - Google Patents

轨迹切换的方法和装置 Download PDF

Info

Publication number
WO2020024134A1
WO2020024134A1 PCT/CN2018/097944 CN2018097944W WO2020024134A1 WO 2020024134 A1 WO2020024134 A1 WO 2020024134A1 CN 2018097944 W CN2018097944 W CN 2018097944W WO 2020024134 A1 WO2020024134 A1 WO 2020024134A1
Authority
WO
WIPO (PCT)
Prior art keywords
trajectory
control point
drone
point
circle
Prior art date
Application number
PCT/CN2018/097944
Other languages
English (en)
French (fr)
Inventor
李劲松
周游
严嘉祺
Original Assignee
深圳市大疆创新科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳市大疆创新科技有限公司 filed Critical 深圳市大疆创新科技有限公司
Priority to PCT/CN2018/097944 priority Critical patent/WO2020024134A1/zh
Priority to CN201880037381.7A priority patent/CN110730934A/zh
Publication of WO2020024134A1 publication Critical patent/WO2020024134A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
    • G05D1/12—Target-seeking control
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
    • G05D1/02—Control of position or course in two dimensions
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules

Definitions

  • the present application relates to the field of drones, and in particular, to a method and device for trajectory switching of a drone.
  • Point of interest is important information for enriching navigation maps.
  • a POI can be a house, a shop, a post box, or a bus stop.
  • drones are being used in more and more scenarios. Making autonomous drones to perform flight operations is an important goal in the development of drones.
  • POI surround is one of the most interesting flight modes for users.
  • a drone After obtaining the accurate coordinates of a POI on a map, a drone usually first flies near the POI with the highest efficiency, for example, and then surrounds the POI with a certain radius.
  • this POI orbiting method requires knowing the precise coordinates of the POI, solidifying the flight trajectory, and poor autonomy of the drone.
  • the existing trajectories of the unmanned aerial vehicles are not smoothly connected and switched, and the user experience is poor.
  • the present application provides a method and device for trajectory switching, which enables smooth connection of various trajectories of an unmanned aerial vehicle during autonomous flight, which can improve the autonomous operation of the unmanned aerial vehicle and improve the user experience.
  • a trajectory switching method which includes: a drone moves along a first trajectory, the first trajectory is a circular trajectory, a center of a circle where the first trajectory is located is a first center, and a radius is A first radius, the center of the first circle is at a first position; the drone determines a second position, the second position is the position of the point of interest updated relative to the first position; the drone is based on The second position determines a second trajectory, the second trajectory is also an arc trajectory, the first trajectory is tangent to the second trajectory, and the center of the circle where the second trajectory is located is the second center and the radius Is the second radius, and the center of the second circle is located at the second position; the drone is switched from the first trajectory to running along the second trajectory.
  • an apparatus for trajectory switching including: a processor and a memory, where the memory is configured to store computer-executable instructions, and the processor is configured to execute the computer-executable instructions to implement the following operations:
  • the drone moves along a first trajectory, where the first trajectory is an arc trajectory, the center of the circle where the first trajectory is located is the first center, the radius is the first radius, and the first center is at the first position;
  • a second position where the second position is an updated point of interest relative to the first position;
  • a second trajectory is determined based on the second position, the second trajectory is also an arc trajectory, and the first
  • the trajectory is tangent to the second trajectory, the center of the circle in which the second trajectory is located is the second center, the radius is the second radius, and the second center is located at the second position;
  • One track is switched to run along the second track.
  • an unmanned aerial vehicle including the trajectory switching device of the second aspect.
  • a computer-readable storage medium on which instructions are stored, and when the instructions are run on the computer, the computer is caused to execute the method of the first aspect.
  • a computer program that causes a computer to execute the method of the first aspect.
  • the two arc trajectories are tangent, so that the trajectories of the drone during autonomous flight can be smoothly connected, which can Improve autonomous operation of drones and improve user experience.
  • a trajectory switching method including: the UAV divides at least part of the trajectory on the second trajectory into N equal parts to obtain N + 1 waypoints, and the second trajectory is a drone waiting The trajectory to switch to; the drone determines, from the N + 1 waypoints, the waypoint closest to the current position of the drone; the drone determines the current position of the drone The foot from the position to the tangent of the second trajectory made by the nearest waypoint is the target point; the drone switches to run along the second trajectory via the target point.
  • a trajectory switching device including: a processor and a memory, where the memory is configured to store computer-executable instructions, and the processor is configured to execute the computer-executable instructions to implement the following operations: At least part of the trajectory on the second trajectory is divided into N equal parts to obtain N + 1 waypoints, and the second trajectory is the trajectory to which the drone is to be switched; the distance is determined from the N + 1 waypoints without people The nearest waypoint where the aircraft is currently located; determining the foot from the current location of the drone to the tangent of the second trajectory made by the nearest waypoint as the target point; The human machine switches to run along the second trajectory via the target point.
  • an unmanned aerial vehicle including the trajectory switching device of the seventh aspect.
  • a computer-readable storage medium on which instructions are stored, and when the instructions are run on the computer, the computer is caused to execute the method of the sixth aspect.
  • a computer program is provided that causes a computer to execute the method of the sixth aspect.
  • the trajectory to be switched to is discretized into a plurality of waypoints, a waypoint nearest to the current position of the drone is determined, and a tangent of the trajectory made by the nearest waypoint Using the vertical foot as a target point and switching the drone to run along the trajectory through the target point can make the drone smoothly perform trajectory switching and improve the user experience.
  • FIG. 1 is a schematic flowchart of a trajectory switching method according to an embodiment of the present application.
  • FIG. 2 is a schematic diagram of an application scenario provided by an embodiment of the present application.
  • FIG. 3 is a schematic diagram of a reference image according to an embodiment of the present application.
  • FIG. 4 is a schematic flowchart of a drone control method according to an embodiment of the present application.
  • FIG. 5 is a schematic diagram of a drone flying around a reference object according to an embodiment of the present application.
  • FIG. 6 is a schematic flowchart of a drone control method according to another embodiment of the present application.
  • FIG. 7 is a schematic diagram of feature point tracking provided by an embodiment of the present application.
  • FIG. 8 is a schematic diagram of a correspondence relationship between three-dimensional coordinates and pixel coordinates provided by an embodiment of the present application.
  • FIG. 9 is a schematic diagram of a drone flying around a reference object according to an embodiment of the present application.
  • FIG. 10 is a schematic diagram of a drone flying around a target object provided by an embodiment of the present application.
  • FIG. 11 is a schematic flowchart of a drone control method according to another embodiment of the present application.
  • FIG. 12 is a schematic diagram of feature point tracking provided by an embodiment of the present application.
  • FIG. 13 is a schematic diagram of a principle of generating a circular trajectory by third-order Bezier curve fitting according to an embodiment of the present application.
  • FIG. 14 is a schematic diagram of a first circle center and four control points according to an embodiment of the present application.
  • FIG. 15 is a schematic diagram of a first trajectory and a second trajectory according to an embodiment of the present application.
  • FIG. 16 is a schematic diagram of a principle of determining a target point according to an embodiment of the present application.
  • FIG. 17 is a schematic flowchart of a trajectory switching method according to another embodiment of the present application.
  • FIG. 18 is a schematic block diagram of a trajectory switching apparatus according to an embodiment of the present application.
  • FIG. 19 is a schematic block diagram of a trajectory switching apparatus according to another embodiment of the present application.
  • a component when a component is called “fixed to” another component, it may be directly on another component or a centered component may exist. When a component is considered to be “connected” to another component, it can be directly connected to another component or a centered component may exist at the same time.
  • the method and device for trajectory switching in the embodiments of the present application can be applied to a drone or other vehicle with a single or multiple cameras, especially a vehicle with autonomous operation functions, such as an unmanned car, an automatic flight Drones, unmanned ground robots, and unmanned ships.
  • a vehicle with autonomous operation functions such as an unmanned car, an automatic flight Drones, unmanned ground robots, and unmanned ships.
  • this article uses a drone as a schematic description. It can be understood that the drones in this article can be equivalently replaced by the aforementioned vehicles.
  • the drone gradually and accurately calculates the position of the POI (which may be a three-dimensional position or a two-dimensional position omitting the height), and may automatically shoot around the POI. Due to the initial stage, the drone may not know the position of the POI, and generally surrounds it according to a preset radius (for example, 300m). In subsequent calculations, one measurement result about the position of the POI can be obtained each time, and the measurement result will be gradually accurate as the number of measurement increases, until the final measurement result is given when a certain accuracy condition is met. In this process, because the measurement of the position of the POI may not be accurate at one time, the circle center and radius may vary. This brings certain challenges to route planning and trajectory switching and control.
  • FIG. 1 is a schematic flowchart of a trajectory switching method 100 according to an embodiment of the present application. As shown in FIG. 1, the method 100 may include the following steps.
  • the drone moves along a first trajectory.
  • the first trajectory is an arc trajectory.
  • the center of the circle where the first trajectory is located is the first center.
  • the radius is the first radius. .
  • the drone determines a second position, where the second position is a position of the point of interest updated relative to the first position.
  • the drone determines a second trajectory according to the second position.
  • the second trajectory is also an arc trajectory.
  • the first trajectory is tangent to the second trajectory.
  • the center of the circle where the second trajectory is located is the second center.
  • the radius is the second radius, and the second circle center is located at the second position.
  • the drone is switched from the first trajectory to running along a second trajectory.
  • the method of trajectory switching in the embodiment of the present application is to switch between two arc trajectories by updating the position of the point of interest, and the two arc trajectories are tangent, so that the trajectories of the drone are smoothly connected during autonomous flight , Can improve the autonomous operation of the drone and improve the user experience.
  • the drone first moves along the trajectory of the center of the circle as the first position (the initial position or the position of the first measured POI), the arc with the radius of the first radius, or a circle or more than one complete circle.
  • the drone calculates the position of the POI again, and the measurement result is the second position, that is, the second position is the position of the updated POI relative to the first position.
  • the drone determines a new trajectory, that is, a second trajectory, which is tangent to the first trajectory (eg, inward or outward), based on the new second position.
  • the drone moves along a trajectory with the center of the circle as the second position, the arc with the radius as the second radius, or a circle or more than one complete circle.
  • the tangent point of the first trajectory and the second trajectory may be the current position of the drone, that is, the drone switches the trajectory immediately.
  • the tangent point of the first trajectory and the second trajectory may also be a position after the position where the drone is currently located on the first trajectory, that is, the drone continues to run along the first trajectory for a period before switching trajectories. The embodiment does not limit this.
  • points of interest may also be referred to herein as target objects.
  • the first position and / or the second position may be calculated by a binocular camera or a monocular camera provided on the drone according to a vision measurement algorithm. That is, the first position and / or the second position may be measured in a process of gradually and accurately measuring the position of the POI.
  • the first position and / or the second position may be obtained through the following steps S1 and S2.
  • step S1 the drone obtains the indication information of the target object, wherein the indication information includes position information of the target object in the reference image output by the photographing device.
  • FIG. 2 is a schematic diagram of an application scenario provided by an embodiment of the present application.
  • the drone 20 is equipped with a photographing device 21, and the photographing device 21 may specifically be a camera, a video camera, or the like.
  • the photographing device 21 may be mounted on the drone 20 through the gimbal 22, or the photographing device 21 may be fixed on the drone 20 through other fixing devices.
  • the shooting device 21 can capture video data or image data in real time, and send the video data or image data to the control terminal 24 through the wireless communication interface 23 of the drone 20.
  • the control terminal 24 may specifically correspond to the drone 20
  • the remote control may also be a user terminal such as a smart phone, a tablet computer, or the like.
  • the drone 20 may further include a control device, and the control device may include a general-purpose or special-purpose processor, which is only schematically illustrated here and does not limit the specific structure of the drone.
  • the image captured by the shooting device 21 includes a target object 31 as shown in FIG. 2.
  • a certain frame image output by the shooting device 21 is recorded as a reference image, and the processor of the drone 20 may obtain Instruction information of the target object, where the instruction information includes position information of the target object in the reference image.
  • the obtaining the indication information of the target object includes: receiving the indication information sent by the control terminal, wherein the indication information is detected by the control terminal on the interactive interface displaying the reference image The target object selection operation is determined.
  • FIG. 3 is a schematic diagram of a reference image according to an embodiment of the present application.
  • 30 represents a reference image displayed in the interactive interface, and the reference image 30 includes a target object 31.
  • One possible way for the user to select the target object 31 in the interactive interface is: the user selects the point 32, It starts to slide from point 32 to point 33. This is only a schematic description here, and the specific selection operation is not limited in this embodiment.
  • the control terminal 24 may determine a region 34 selected by the user in the interactive interface according to a user's selection operation in the interactive interface, and determine position information of the region 34 in the reference image 30. For example, the control terminal 24 may It is determined that the upper left corner of the region 34 is the position information of the point 32 in the reference image 30 and the size of the region 34, such as length and width, or the control terminal 24 may determine that the upper left corner of the region 34 is the point 32 in the reference. The position information in the image 30 and the position information of the point 33 in the reference image 30 in the lower right corner of the region 34. Further, the control terminal 24 may send the position information of the area 34 in the reference image 30 to the drone 20 as the position information of the target object 31 in the reference image 30, that is, the indication information of the target object 31.
  • the acquiring the indication information of the target object includes: identifying the target object in the reference image to acquire the indication information of the target object.
  • the processor of the drone may identify the target object in the reference image output by the photographing device 21, and obtain the indication information of the target object through recognition. Further, the processor of the drone may input the reference image into a trained neural network model, and obtain indication information of the target object output by the neural network model.
  • Step S2 Determine position information of the target object according to the instruction information.
  • the drone may determine the position information of the target object according to the indication information, where the position information of the target object may be three-dimensional position information or two-dimensional position information;
  • the position information of the target object may be position information based on the world coordinate system; in addition, the position information of the target object may be position information based on the global coordinate system, and the position information may include at least longitude and latitude;
  • the position information of the target object may also be position information based on a body coordinate system of the drone.
  • determining the position information of the target object according to the instruction information includes: determining the orientation of the target object with respect to the drone according to the instruction information, and determining the orientation of the target object with respect to the drone according to the orientation information.
  • the horizontal distance between the objects or the ground-height value of the drone determines the position information of the target object.
  • the orientation of the target object 31 with respect to the drone 20 may be determined according to the position information of the target object 31 in the reference image 30 and the attitude of the gimbal that carries the photographing device 21;
  • the horizontal distance between the object 31 and the drone 20 determines the position information of the target object 31.
  • the FOV of the shooting device 21 is known, and the angle of the target object 31 relative to the optical axis of the shooting device 21 can be determined according to the position information of the target object 31 in the reference image. For example, if the target object 31 is at the center of the reference image, the angle of the target object 31 with respect to the optical axis of the shooting device is 0.
  • the attitude of the head 22 of the photographing device 21 also determines the light of the photographing device 21
  • the orientation of the axis combined with the angle of the target object 31 relative to the optical axis of the photographing device 21 and the orientation of the optical axis, can obtain the orientation of the target object 31 with respect to the drone 20.
  • the position information of the target object 31 is determined according to the orientation of the target object 31 with respect to the drone 20 and the horizontal distance between the target object 31 and the drone 20. In some embodiments, the position information of the target object 31 is determined according to the orientation of the target object 31 with respect to the drone 20, the horizontal distance between the target object 31 and the drone 20, or the ground-height value of the drone 20 .
  • the angle of the target object 31 relative to the drone 20 in the pitch direction can be determined according to the orientation of the target object 31 with respect to the drone 20, such as the angle ⁇ shown in FIG. 2, and then, an unmanned person can be obtained.
  • the ground height value of the drone measured by the distance sensor configured on the aircraft 20, such as h shown in FIG. 2, according to the ⁇ angle and the ground height value, the target object can be determined in a vertical direction relative to the drone.
  • the angle of the target object 31 with respect to the drone 20 in the yaw direction can be determined according to the ⁇ angle and the target object 31 and no
  • the horizontal distance L between the human machine 20 can determine the position information of the target object relative to the drone in the horizontal direction; according to the position information of the target object relative to the drone in the vertical direction and the target object relative to the drone
  • the position information in the horizontal direction can determine the position information of the target object relative to the drone, and further, the target can be determined based on the position information of the target object relative to the drone and the position information of the drone.
  • Location information of the object, location information of the target object may be the position of the target object in the world coordinate system, may be the position of the target object in the global coordinate system.
  • the position information of the target object relative to the drone in the vertical direction may also be determined according to the horizontal distance L between the target object 31 and the drone 20 and the ⁇ angle.
  • the indication information of the target object may indicate the size of the image area corresponding to the target object in the reference image, and the horizontal distance between the target object 31 and the drone 20 may be determined according to the size of the image area.
  • the drone may be controlled to move around the target object according to the position information of the target object.
  • the target object 31 is used as the center, and the surrounding trajectory is generated according to the position relationship between the drone 20 and the target object 31, and the The man-machine 20 moves on the orbit, that is, the drone 20 is controlled to fly on the orbit, and the orbiting target object 31 is realized. While the drone 20 is flying around the target object 31, the shooting device 21 can shoot the target object 31 in real time, and send the image data or video data obtained through the shooting to the control terminal 24 through the wireless communication interface 23, so that the user can browse and watch .
  • This embodiment determines position information of a target object by acquiring position information of a target object captured by the shooting device in a reference image output by the shooting device, and controls the drone to move around the target object according to the position information of the target object.
  • the drone does not need to move to the orbiting center to record the position of the orbiting center, the drone can move around the target object, which simplifies the process of the drone's orbiting movement of the target object, and improves the drone. Operational safety.
  • FIG. 4 is a flowchart of a method for controlling a drone provided by another embodiment of the present application. As shown in FIG. 4, based on the embodiment shown in FIG. 1, this embodiment provides another implementable manner for determining the position information of the target object according to the instruction information. Specifically, according to the instruction The information determining the location information of the target object may include the following steps:
  • Step S401 Control the UAV to orbit the reference object.
  • a point may be taken as a reference object at a preset distance directly in front of the drone, and the reference object is specifically a virtual target point, and the drone is controlled to orbit the reference object.
  • FIG. 5 is a schematic diagram of orbiting a reference object by a drone according to an embodiment of the present application. As shown in FIG. 5, 50 indicates a reference object at a preset distance in front of the drone, and 51 indicates the nose of the drone. The processor in the drone can specifically control the drone to surround the reference object 50. flight.
  • the controlling the drone to orbit the reference object includes: determining the reference object according to a preset orbit radius, and controlling the drone to orbit the reference object.
  • the UAV is controlled with the reference object 50 as the orbiting center, and a preset orbiting radius such as 500 meters as a radius to generate a circular trajectory, such as the circular trajectory 53 shown in FIG. 5, and controlling the drone at The circular track 53 orbits the reference object 50.
  • the drone may fly on the circular trajectory 53 in a counterclockwise direction, and may also fly on the circular trajectory 53 in a clockwise direction.
  • the processor in the drone receives the instruction information of the target object sent by the control terminal, it can determine the reference object according to the preset orbit radius, and control the drone to orbit the reference object. . In other words, after the user selects the target object framed in the reference image, the drone can orbit the reference object.
  • controlling the drone to orbit the reference object includes: after receiving the start control instruction sent by the control terminal, controlling the drone to orbit the reference object mobile.
  • the activation control button 35 may be displayed in the interaction interface.
  • the activation control button 35 may be an icon in the interaction interface, that is, when After the user selects the target object 31, the drone does not immediately orbit the reference object, but waits for the user to click the start control button 35 in the interactive interface before the drone begins to orbit the reference object. .
  • the control terminal when the user clicks the start control button 35 in the interactive interface, the control terminal generates a start control instruction according to the user's click operation, and sends the start control instruction to the drone, and when the processor in the drone After receiving the start control instruction, the UAV is controlled to orbit the reference object.
  • the specific orbit control method may be the method shown in FIG. 5, which is not repeated here.
  • Step S402 In the process of moving the drone around a reference object, obtain a plurality of frames of a first target image output by the photographing device, where the first target image includes the target object.
  • the shooting device of the drone may also be captured and a target image including the target object 31 may be output.
  • the target image captured by the drone during the orbital flight of the reference object 50 is recorded as the first target image.
  • the first target image output by the shooting device of the drone may be multiple frames.
  • the processor of the drone may obtain a plurality of frames of the first target image output by the shooting device, and the first target image includes the target object 31 .
  • the angle at which the target object 31 is offset from the optical axis of the shooting device is not limited here, as long as the target object 31 is ensured in the shooting screen of the shooting device.
  • Step S403 Determine the position information of the target object according to the indication information of the target object and the multi-frame first target image.
  • the processor of the drone may determine the position information of the target object 31 according to the indication information of the target object 31 obtained in the foregoing embodiment and the first target image of the multiple frames obtained in the foregoing steps.
  • determining the position information of the target object according to the indication information of the target object and the multi-frame first target image may include the following steps as shown in FIG. 6:
  • Step S601 Acquire a feature point in a target area of the reference image, where the target area is an image area indicated by the instruction information in the reference image.
  • the drone After the drone receives the instruction information of the target object sent by the control terminal, it can determine the target area of the reference image according to the instruction information of the target object, and the target area is specifically the image area indicated by the instruction information. For example, as shown in FIG. 3, after the drone receives the position information of the area 34 in the reference image 30 sent by the control terminal or obtains the position information of the area 34 in the reference image 30 by recognition, the drone's processor A target area may be determined in the reference image 30, and the target area may specifically be the area 34, that is, the drone may use the area framed by the user in the interactive interface as the target area. Further, the processor of the drone may obtain the feature points in the target area.
  • the processor may determine the feature points in the target area according to a preset feature point extraction algorithm.
  • the feature point extraction algorithm Including at least one of the following: Harris corner detection algorithm, Scale-invariant feature transform (SIFT), Speeded Up Robust Features (SURT) algorithm, Fast feature point extraction and description algorithm (Oriented FAST and Rotated Brief (ORB), etc.
  • Harris corner detection algorithm is used to extract feature points in the target area.
  • Step S602 Use a tracking algorithm to obtain the feature points of the first target image of each frame based on the feature points in the target area of the reference image.
  • the tracking algorithm is used to track the feature points in the target area, that is, the position of the feature points in the target area in the first target image of each frame is determined using the tracking algorithm.
  • the tracking algorithm may specifically be a (Kanade-Lucas-Tomasi Feature Tracker, KLT) feature tracking algorithm.
  • FIG. 7 is a schematic diagram of feature point tracking provided by an embodiment of the present application.
  • A, B, C, D, E, F, and G respectively represent the feature points in the target region of the reference image 30, that is, the feature points in the region 34, and the feature points A, B, C, D, E, F, and G. It is also a characteristic point of the target object 31.
  • 71, 72, and 73 respectively represent the first target images sequentially output by the shooting device during the process of the drone flying around the reference object.
  • feature points of the target object 31 in the reference image 30, such as A, B, C, D, E, F, and G, can be determined in the first target image 71, the first target image 72, and the first target image, respectively.
  • Position in 73 the photographing device first outputs the reference image 30, and then sequentially outputs the first target image 71, the first target image 72, and the first target image 73; the reference image 30, the first target image 71, the first target image 72, and the first target.
  • the image 73 may be an adjacent image or a non-adjacent image.
  • the position of the target object 31 relative to the drone constantly changes, resulting in the target object 31 in the first target image sequentially output by the photographing device.
  • the positions are constantly changed, so that the positions of the feature points corresponding to the target object 31 in the first target image 71, the first target image 72, and the first target image 73 in the corresponding first target image are continuously changed.
  • This is only a schematic description, and does not limit the number of feature points in the area 34, nor the number of first target images, and the position of the feature points in the area 34 in the first target image in each frame.
  • Step S603 Determine the position information of the target object according to the position information of the feature points of the first target image in each frame in the corresponding first target image.
  • the position information of the target object 31 is determined according to the position information of the feature points corresponding to the target object 31 in the corresponding first target image in the first target image 71, the first target image 72, and the first target image 73.
  • the position information of the target object 31 is specifically three-dimensional coordinates of the target object 31 in a three-dimensional space.
  • the position information of the target object 31 is determined according to the position information of the feature points corresponding to the target object 31 in the first target image 71, the first target image 72, and the first target image 73 in the corresponding first target image. Recorded as the first position information.
  • a new first target image is also output, and the position of the feature point of the target object 31 in the new first target image can be determined according to the KLT feature tracking algorithm; further According to the first target image 71, the first target image 72, the first target image 73, and the position information of the feature points corresponding to the target object 31 in the new first target image in the corresponding first target image, another target can be determined The position information of the object 31.
  • the position information of the target object 31 is recorded as the second position information.
  • the above-mentioned first position information and the second position information may be the same or different, but it can be understood that as the shooting device continuously outputs new first target images, according to the first target image 71, the first target The accuracy of the position information of the target object 31 determined by the position information of the feature point corresponding to the target object 31 in the corresponding first target image in the image 72, the first target image 73, and the first target image continuously output by the photographing device subsequently. Degree continues to increase.
  • the processor of the drone can determine a new position information of the target object 31.
  • determining the position information of the target object based on the position information of the feature points of the first target image in each frame in the corresponding first target image includes: The position information in the first target image is determined by using a fitting algorithm.
  • FIG. 8 is a schematic diagram of a correspondence relationship between three-dimensional coordinates and pixel coordinates provided by an embodiment of the present application.
  • 80 represents a target object
  • 81, 82, and 83 represent the first target images successively output by the shooting device during the movement of the shooting device around the target object 80 in the direction shown by the arrow.
  • the three-dimensional points on the first target image 81, 82, 83 can be mapped, and the mapping points of the three-dimensional points in the first target image 81, 82, 83 can be features in the first target image 81, 82, 83.
  • the number of feature points that can be tracked is decreasing.
  • point A, point B, and point C are three-dimensional points on the target object 80 respectively.
  • Point a1, point b1, and point c1 represent feature points in the first target image 81.
  • Point a1 corresponds to point A
  • point b1 corresponds to point.
  • B corresponds
  • point c1 corresponds to point C;
  • points a2, b2, and c2 represent feature points in the first target image 82, point a2 corresponds to point A, point b2 corresponds to point B, and point c2 corresponds to point C;
  • Points a3 and b3 represent feature points in the first target image 83, point a3 corresponds to point A, and point b3 corresponds to point B.
  • the three-dimensional coordinates (x w , y w , z w ) of the three-dimensional point on the target object 80 in the world coordinate system and the three-dimensional point in the first target image can be obtained.
  • the position information of the mapping point in the first target image is, for example, a relationship of pixel coordinates ( ⁇ , ⁇ ), and the relationship is specifically shown in the following formula (1):
  • z c represents the coordinate of the three-dimensional point on the Z axis of the camera coordinate system
  • K represents an internal parameter of the camera
  • R represents a rotation matrix of the camera
  • T represents a translation matrix of the camera.
  • ( ⁇ , ⁇ ), K, R, and T are known quantities
  • z c and (x w , y w , z w ) are unknown quantities.
  • an equation shown in formula (1) can be established, and according to the point a2
  • the pixel coordinates in the first target image 82 and the R and T corresponding to the first target image 82 captured by the photographing device can be used to establish another equation as shown in formula (1), according to the point a3 in the first target image
  • the pixel coordinates in 83, and the R and T corresponding to the first target image taken by the shooting device 83 can establish another equation as shown in formula (1), and as the shooting device continuously outputs new first target images
  • the equation shown in formula (1) is gradually increased.
  • the corresponding unknowns can be solved. That is, by solving these equations using a fitting algorithm, the three-dimensional coordinates of the three-dimensional point A in the world coordinate system can be calculated. Similarly, the three-dimensional coordinates of the three-dimensional point B and the three-dimensional point C in the world coordinate system can be calculated, and details are not described herein again. It can be understood that the more the first target image output by the photographing device is, the more accurate the three-dimensional coordinates of the three-dimensional point obtained in the world coordinate system based on the pixel coordinates of the feature points in the multi-frame first target image using a fitting algorithm.
  • the target After determining the three-dimensional coordinates of the three-dimensional points on the target object 80, such as the three-dimensional coordinates of the three-dimensional points A, B, and C in the world coordinate system, the target can be determined according to the three-dimensional coordinates of the three-dimensional points A, B, and C in the world coordinate system.
  • the UAV may obtain the position information of the target object according to the three-dimensional coordinates of the target object 80 in the world coordinate system.
  • the position information of the target object 31 when the position information of the target object 31 is based on the position in the global coordinate system, the position information of the target object 31 may be determined according to the position information of the drone and the three-dimensional coordinates of the target object 80 in the world coordinate system.
  • the position information of the target object 31 is based on the position in the body coordinate system of the drone, the three-dimensional coordinates of the target object 80 in the world coordinate system can be converted to the body coordinate system to obtain the position information based
  • the method further includes: after acquiring the feature points of the first target image of each frame, determining, from the feature points of the first target image of each frame, the target feature points that meet the preset requirements; accordingly, the according to The determination of the position information of the target object by the position information of the feature points of the first target image in each frame in the corresponding first target image includes: according to the target feature points of the first target image in each frame in the corresponding first target image The position information of the target object determines position information of the target object.
  • target feature points that meet preset requirements are determined, for example, an offset amount of each feature point between the first target image 71 and the reference image 30 It may be different, assuming that the offset amount of the feature point A between the first target image 71 and the reference image 30 is recorded as h1, and the offset amount of the feature point B between the first target image 71 and the reference image 30 is recorded as h2, and so on.
  • the offset of the feature point G between the first target image 71 and the reference image 30 is recorded as h7, and the average and variance of h1, h2, ..., h7 are calculated.
  • ⁇ 2 according to the Gaussian distribution, select the feature points with the offset within [u-3 ⁇ , u + 3 ⁇ ] as the target feature points. Assuming that h1 is outside [u-3 ⁇ , u + 3 ⁇ ], the first target image 71 The feature points A in the first target image 71 are deleted, the feature points B, C, D, E, F, and G in the first target image 71 are retained, and the feature points B, C, and D, E, F, and G serve as target feature points of the first target image 71. Similarly, target feature points in the first target image 72 and the first target image 73 can be calculated, and details are not described herein again.
  • the average and variance of h1, h2, ... h7 are calculated according to the offset between each feature point between the first target image 71 and the reference image 30, such as h1, h2, ... h7, Feature points with offsets within [u-3 ⁇ , u + 3 ⁇ ] are selected as valid points according to the Gaussian distribution. For example, if h1 is outside [u-3 ⁇ , u + 3 ⁇ ], the features in the first target image 71 The point A is deleted, and the feature points B, C, D, E, F, and G in the first target image 71 are used as valid points. The target feature points are further determined from the valid points, and the target features are determined from the valid points.
  • the target feature points in the first target image 72 and the first target image 73 can be used as the target feature points of the first target image 71.
  • the target object 31 is determined in the world according to the position information of the target feature points in the corresponding first target image.
  • the specific principle of the three-dimensional coordinates in the coordinate system is consistent with the principle shown in FIG. 8, and is not repeated here.
  • the drone by controlling the drone to orbit the reference object, in the process of the drone orbiting the reference object, multiple frames of the first target image output by the photographing device are acquired, and according to the instruction information of the target object and the multiple frames of the first target.
  • the position information of the target object is determined by the image.
  • the shooting device continuously outputs the first target image
  • the position information of the target object can be continuously determined according to the instruction information of the target object and the first target image continuously output by the shooting device.
  • the accuracy of the position information of the target object is continuously improved; in addition, after obtaining the feature points of the first target image of each frame output by the shooting device, a feature point that meets the preset requirements is determined from the feature points of the first target image of each frame.
  • the target feature point when determining the position information of the target object according to the position information of the target feature point of each frame of the first target image in the corresponding first target image, can improve the accuracy of the position information of the target object, and Removing the feature points that do not meet the preset requirements can also reduce the corresponding calculation amount.
  • the embodiment of the present application provides a method for controlling an unmanned aerial vehicle.
  • the method further includes: determining, according to the position information of the feature point of the first target image in each frame in the corresponding first target image, the position of the drone during the movement of the reference object.
  • the parallax of the shooting device relative to the target object; correspondingly, controlling the drone to move around the target object according to the position information of the target object includes: when the parallax is greater than a first preset parallax threshold And, according to the determined position information of the target object, determining an orbit track of the drone to orbit the target object, and controlling the drone to move on the orbit track.
  • the first target image 72, and the first target image 73 such as A, B, C, D, E, F, and G
  • the feature points A , B, C, D, E, F, G respectively in the first target image 71, the first target image 72, the first target image 73 position information can be determined as shown in Figure 5 During the flight of the reference object 50, the parallax of the shooting device of the drone relative to the target object 31.
  • the first target image 71 is an image taken by the shooting device when the drone is at the m1 position
  • the first target image 72 is The image captured by the device when the human machine is at the m2 position
  • the first target image 73 is an image captured by the device when the drone is at the m3 position. According to the position information of the feature points A, B, C, D, E, F, and G in the first target image 71 and the first target image 72, respectively, the process of the drone from the m1 position to the m2 position can be determined.
  • the parallax of the shooting device of the drone relative to the target object 31 specifically, the pixel position of the feature point A in the first target image 71 is labeled as ( ⁇ 1 , ⁇ 1 ), and the feature point A is at the first target
  • the pixel mark in the image 72 is ( ⁇ 2 , ⁇ 2 ).
  • the parallax of the feature point A can be calculated according to the following formula (2), and the parallax of the feature point A is recorded as parallaxA:
  • R 21 represents a change in the rotation direction of the attitude of the camera when shooting the first target image 72 relative to the attitude of the camera when shooting the first target image 71.
  • c x and c y represent the positions of the camera optical centers. It can be understood that the positions of the camera optical centers in the first target image 71 and the first target image 72 are the same.
  • f represents the focal length of the camera.
  • the disparity of feature points B, C, D, E, F, and G can be calculated. The disparity of feature points A, B, C, D, E, F, and G is averaged, and the average is the first.
  • the parallax of the first target image 72 is the parallax of the shooting device of the drone with respect to the target object 31 during the process of the drone from the m1 position to the m2 position.
  • the parallax of the first target image 73 can be determined.
  • the parallax of a target image 73 is the parallax of the shooting device of the drone with respect to the target object 31 during the process of the drone from the m1 position to the m3 position. It can be understood that as the drone flies along the circular trajectory 53, the parallax of the shooting device of the drone with respect to the target object 31 is continuously increasing, and the three-dimensional coordinates of the target object 31 are continuously determined by using a fitting algorithm.
  • the fitting algorithm is stopped to obtain The newly determined three-dimensional coordinates of the target object 31, that is, the precise three-dimensional coordinates of the target object 31, and according to the newly determined three-dimensional coordinates of the target object 31, determine the orbit of the drone to orbit the target object 31.
  • the trajectory is different from the surrounding trajectory 53 of the reference object 50 by the drone.
  • FIG. 9 is a schematic diagram of a drone flying around a reference object according to an embodiment of the present application.
  • the parallax of the shooting device of the drone relative to the target object 31 is greater than the first preset parallax threshold, and the target object is determined according to the latest determination.
  • the three-dimensional coordinates of 31 and a preset surround parameter, such as a surround radius determine a target trajectory 91 for the drone to orbit the target object 31, and control the drone to fly along the target trajectory 91.
  • the method further includes: determining a change speed of the parallax; and adjusting a speed at which the drone moves around the reference object according to the change speed of the parallax.
  • the determining the change speed of the parallax includes determining the position information of the feature points of two adjacent first target images in the first target image in the corresponding first target image in multiple frames of the first target image. The speed at which parallax changes.
  • parallax_speed (PA i -PA i-1 ), that is, when the image frequency is fixed, measure (PA i
  • the magnitude of -PA i-1 ) / t is consistent with the significance of measuring the magnitude of PA i -PA i-1 .
  • the method further includes: when the parallax is greater than a second preset parallax threshold, adjusting a radius of the drone to orbit the reference object according to the determined position information of the target object, where , The first preset parallax threshold is greater than the second preset parallax threshold.
  • the parallax of the shooting device of the drone relative to the target object 31 is greater than the first preset parallax threshold, and the target determined according to the latest The three-dimensional coordinates of the object 31 determine the target trajectory 91 for the drone to orbit the target object 31.
  • the drone may be far away from the target trajectory 91, and the drone needs to fly from the current position such as the m3 position to the target. A point on the trajectory 91 starts to fly along the target trajectory 91 again.
  • the parallax of the shooting device of the drone relative to the target object 31 is greater than the second preset parallax threshold.
  • the second preset parallax threshold is smaller than the first preset parallax threshold.
  • the target trajectory 91 of the target object 31 for orbital flight starts from the m2 position, and can continuously adjust the radius of the orbiting of the reference object 50 by the drone, such as continuously reducing the reference of the drone to the reference.
  • the radius of the subject 50 performing orbital flight.
  • the parallax of the shooting device of the drone relative to the target object 31 is constantly changing.
  • the drone may reach a point on the target trajectory 91 (accurate target trajectory), such as m4, or the drone may Reaching a point closer to the target trajectory 91 makes the UAV smoothly transition from this point to the target trajectory 91.
  • This embodiment determines the parallax of the shooting device relative to the target object during the flight around the reference object by using the position information of the feature points of the first target image in each frame in the corresponding first target image. Change the speed to adjust the flying speed of the drone flying around the reference object, so that the drone can determine the 3D coordinates of the target object in a short time, especially when the target object is far away from the drone and the drone orbits around the reference object.
  • the speed of change of the parallax can increase the flying speed of the drone and the efficiency of calculating the three-dimensional coordinates of the target object.
  • the first preset parallax threshold is greater than the second preset parallax threshold, and when the parallax is greater than the second preset parallax threshold, by adjusting the radius of the drone to orbit the reference object, So that when the parallax is greater than the first preset parallax threshold, the drone arrives at the orbit of the orbit flight of the target object , The arrival distance or trajectory position close surrounding, so that a smooth transition from the UAVs reference object to fly around the track around the object to fly around the target trajectory around.
  • FIG. 11 is a flowchart of a method for controlling a drone according to another embodiment of the present application. As shown in FIG. 11, on the basis of the above embodiment, the method further includes: after acquiring the instruction information, controlling a shooting posture of the photographing device according to the instruction information so that the target object is in the shooting state Center of the device's shooting screen.
  • the target object 31 may not be at the center of the shooting screen of the shooting device.
  • the drone obtains the instruction of the target object 31 Information, for example, after receiving position information of the area 34 in the reference image 30 sent by the control terminal 24, based on the position information of the area 34 in the reference image 30, it is possible to determine the position of the target object 31 relative to the optical axis of the photographing device 21. Angle, according to which the attitude of the drone and / or the attitude of the gimbal can be adjusted to control the shooting attitude of the shooting device, so that the angle of the target object 31 relative to the optical axis of the shooting device is 0, that is, the target object 31 is in the desired position. The center of the shooting frame of the shooting device is described.
  • the drone when the user selects the target object 31, the drone can orbit the reference object; therefore, when the drone obtains the instruction information of the target object 31, it can adjust the attitude of the drone. And / or the attitude of the gimbal, so that the target object 31 is at the center of the shooting frame of the shooting device, that is, adjusting the attitude and / or the cloud of the drone during the drone's orbiting of the reference object.
  • the attitude of the stage is such that the target object 31 is at the center of the shooting frame of the shooting device until the UAV determines the three-dimensional coordinates of the target object 31.
  • the drone when the user selects the target object 31, the drone does not immediately orbit the reference object, but waits until the user clicks the activation control button 35 in the interactive interface. Only then began to orbit the reference object. For example, the drone obtains the instruction information of the target object at time t1. The user clicks the start control button 35 at time t2 after time t1, that is, the drone orbits the reference object from time t2. At time t3 after time t2, the three-dimensional coordinates of the target object 31 are determined.
  • the drone may adjust the attitude of the drone and / or the attitude of the gimbal between time t1 and time t2, so that the target object 31 is at the center of the shooting screen of the shooting device, because from time t1 to t2
  • the drone may not move between moments, but the target object 31 has moved, resulting in a change in the position of the target object 31 in the shooting screen of the shooting device.
  • the drone may also adjust the attitude of the drone and / or the attitude of the gimbal between time t2 and time t3, so that the target object 31 is at the center of the shooting screen of the shooting device.
  • the drone may also adjust the attitude of the drone and / or the attitude of the gimbal between time t1 and time t3, so that the target object 31 is at the center of the shooting frame of the shooting device.
  • the method further includes: after acquiring the instruction information, acquiring a plurality of frames of the second target image output by the photographing device, wherein the second target image includes a target object.
  • the drone when the user selects the target object 31, that is, after the drone obtains the instruction information of the target object 31, the drone can orbit the reference object, and obtain the output of the shooting device when the drone object orbits.
  • the multi-frame second target image then the multi-frame second target image includes the multi-frame first target image.
  • the drone when the user selects the target object 31, the drone does not immediately orbit the reference object, but waits for the user to click the start control button 35 in the interactive interface before the drone starts to reference the object.
  • the multi-frame second target image output by the shooting device may be taken by the shooting device between time t1 and time t2, or it may be at time t2 It was taken between time t3, and it may be taken between time t1 and t3. That is, the multiple target second images include at least multiple first target images.
  • controlling the shooting posture of the shooting device according to the instruction information includes the following steps:
  • Step S1101 Use a tracking algorithm to obtain the feature points of the second target image of each frame based on the feature points in the target area of the reference image.
  • a tracking algorithm is used to calculate an offset between each feature point in the target area between adjacent target images, such as a second target image. If the feature point is in a previous frame target image with respect to a subsequent frame target image The offset of the target point and the target point in the next frame are the same as the offset of the target image in the previous frame, and the directions are opposite, and it can be determined that the characteristic point is the correct tracking point.
  • A, B, C, D, E, F, and G respectively indicate the feature points in the target region of the reference image 30, that is, the feature points in the region 34, and the feature points A, B, C, D, E, F, and G. It is also a characteristic point of the target object 31.
  • 121 indicates that after the UAV acquires the instruction information, one of the second target images in the plurality of frames of the second target image output by the shooting device is only schematically described here.
  • the positions of the feature points of the target object 31 in the reference image 30 such as A, B, C, D, E, F, and G in the second target image 121 can be determined according to the KLT feature tracking algorithm.
  • Step S1102 Determine position information of the target object corresponding to the second target image according to the feature points of the second target image of each frame.
  • position information of the target object 31 in the second target image 121 can be determined, such as Position information of the center point N1 in the second target image 121.
  • Step S1103 Control the shooting posture of the shooting device according to the position information of the target object corresponding to the second target image.
  • the position of the center point N1 of the target object 31 relative to the second target image 121 may be determined.
  • ⁇ and the horizontal FOV can determine the angle of the target object 31 with respect to the optical axis of the camera in the horizontal direction
  • ⁇ and FOV of the camera in the vertical direction can determine the target 31 with respect to the optical axis of the camera Offset angle in the vertical direction.
  • Adjust the shooting attitude of the shooting device by adjusting the attitude of the target object 31 relative to the optical axis of the shooting device in the horizontal and vertical directions, so that the optical axis of the shooting device
  • the target object 31 is aligned, and the target object 31 is located at the center of the screen of the second target image 121.
  • the target object 31 may not be adjusted to the center of the screen of the first target image or the second target image, and the target object 31 may also be adjusted to a preset area in the first target image or the second target image. That is, by adjusting the attitudes of the drone and the gimbal, the angles of the target object 31 with respect to the optical axis of the camera in the horizontal and vertical directions are both non-zero preset angles.
  • the target object by controlling the shooting attitude of the shooting device so that the target object is at the center of the shooting screen of the shooting device, the target object can be prevented from moving outside the shooting screen of the shooting device when the drone is flying around the reference object.
  • the three-dimensional coordinates of the target object cannot be determined normally; in addition, the target object can be prevented from disappearing from the shooting screen of the shooting device during the movement.
  • circular or circular trajectories such as the first trajectory and / or the second trajectory
  • the Bezier curve may be a third-order Bezier curve.
  • one part of the M of the circle where the first trajectory is generated is fitted by a third-order Bezier curve, and one of the parts of the circle where the first trajectory includes four control points; and / or, where the second trajectory is located.
  • the M-fraction of the circle is generated by a third-order Bezier curve fitting.
  • the M-fraction of the circle where the second trajectory includes four control points.
  • t is a curve parameter
  • A, B, C, and D are four control points when a circle is fitted by a M-stage third-order Bezier curve.
  • FIG. 13 is a schematic diagram of a principle of generating a circular trajectory by third-order Bezier curve fitting according to an embodiment of the present application.
  • M 4 that is, a complete circle is fitted by 4 segments of a third-order Bezier curve, and the center of the circle is O.
  • P (0) and P (1) pass through control point A and control point D, respectively, that is, the two endpoints of the curve.
  • control point A is on the x axis and the coordinates are (1,0)
  • the control point D is on the y axis and the coordinates are (0,1).
  • the coordinates of control point B are (h, 1) and the coordinates of control point C are (1, h).
  • the starting position of the M-th circle that is, the control point A may be the current position of the drone.
  • the center of the M-th circle can be the first position of the currently measured interest point. Whether it is a circular orbit or a circular orbit, you only need to specify the two points of the center of the circle and the starting position of the curve, and then you can use four 1/4 circles to splice into a complete circle.
  • the following uses the first trajectory as an example to describe in detail the process of generating a trajectory by third-order Bezier curve fitting.
  • the four control points include a first control point A 1 , a second control point B 1 , a third control point C 1, and a fourth control point D 1 .
  • the first control point A 1 is the current position of the drone, and the first control point A 1 and the fourth control point D 1 are two endpoints of a quarter of the circle where the first trajectory is located.
  • the second control point B 1 is on a tangent to the circle where the first trajectory passes through the first control point A 1 , and the second control point B 1 and the fourth control point D 1 are located on the same side of O 1 A 1 , and the second control point
  • the distance between the point B 1 and the first control point A 1 is h.
  • the third control point C 1 is on a tangent to the circle where the first trajectory passes through the fourth control point D 1 , and the third control point C 1 and the first control point A 1 are located on the same side of O 1 D 1 , and the third control point The distance between the point C 1 and the fourth control point D 1 is h. among them
  • the first control point A 1 may not be exactly on the x axis, and the first circle center O 1 may not be exactly at the coordinate (0,0). May be (i.e., two-dimensional or three-dimensional position of the position of the point of interest) and the first control point A 1 (i.e., the current position of the UAV is located) to the second control point estimated point B 1, depending on the position of the center O 1 of the first Three control points C 1 and a fourth control point D 1 .
  • FIG. 14 is a schematic diagram of a first circle center and four control points according to an embodiment of the present application. The embodiment of the present application only cares about the two-dimensional plane, so the height (ie, z-axis) direction is ignored.
  • the first control point A 1 , the second control point B 1 , the third control point C 1, and the fourth control point D 1 satisfy the following relationship:
  • ⁇ DA is the angle between the vector O 1 D 1 and the vector O 1 A 1.
  • C 1 h (A 1 -O 1 ) + D 1 , where
  • T 0 (t) a 0 + a 1 ⁇ t + a 2 ⁇ t 2 + a 3 ⁇ t 3 .
  • the embodiments of the present application may also obtain the first trajectory by other methods, such as least square fitting. Knowing the center, radius, and points of the arc of the first trajectory, the expression of the first trajectory can be obtained in various ways, which are not listed in the embodiments of this application.
  • the tangent point since the tangent point where the first trajectory and the second trajectory can be obtained, the tangent point is used as the fifth control point A 2 .
  • the second circle center is labeled O 2 and is located at the second position, that is, the position of the updated point of interest relative to the first position.
  • M 4
  • the second circle center is marked as O 2
  • the four control points include a fifth control point A 2 , a sixth control point B 2 , a seventh control point C 2, and an eighth control point D 2
  • the fifth control point A 2 is the current position of the drone.
  • the fifth control point A 2 and the eighth control point D 2 are two endpoints of a quarter of the circle where the second trajectory is located.
  • the sixth control point B 2 The second trajectory of the five control points A 2 is on a tangent to the circle, and the sixth control point B 2 and the eighth control point D 2 are on the same side of O 2 A 2 , and the sixth control point B 2 and the fifth control point A The distance between 2 is h, the seventh control point C 2 is on a tangent to the circle where the second trajectory passes through the eighth control point D 2 , and the seventh control point C 2 and the fifth control point A 2 are located at O 2 D On the same side of 2 , the distance between the seventh control point C 2 and the eighth control point D 2 is h, where The specific process is not repeated here.
  • the second trajectory may be determined according to the first circle center, the first radius, the second circle center, the second radius, and the first trajectory. Assume that the first circle center is O 1 , the first radius is R 1 , the second circle center is O 2 , and the second radius is R 2 .
  • FIG. 15 is a schematic diagram of a first trajectory and a second trajectory according to an embodiment of the present application.
  • the expression of the second trajectory can be obtained in various ways. Since the second track and the first starting track are tangent, the two tracks are smooth in position.
  • first trajectory, the second trajectory, and the like in the embodiments of the present application are only ideal running trajectories planned by the drone.
  • the actual position of the drone is usually deviated from the ideal running trajectory.
  • path following technology which mainly constrains the search for the nearest point, uses the last position to set the initial value, iteratively finds the nearest point, and performs follow-up control.
  • path following technology is now used to correct deviations in operation, such deviations are still difficult to avoid.
  • drones usually have high flight speeds, at the time of real switching, the drone will rush out of the trajectory for a distance. This makes drones usually not smooth when switching.
  • switching the S140 drone from the first trajectory to running along a second trajectory may include: determining, on the second trajectory, the distance from the drone to the current The closest point is the target point; the UAV switches from the first trajectory to run along the second trajectory through the target point. Based on the existing method, determining the point on the second trajectory closest to the position where the drone is currently located can be achieved by obtaining the second derivative of the second trajectory, which will not be repeated here.
  • the switching of the S140 drone from the first trajectory to running along the second trajectory may include: the drone performs at least a part of the trajectory on the second trajectory by N, etc. Points to obtain N + 1 waypoints; the drone determines, from the N + 1 waypoints, the waypoint closest to the current position of the drone; the drone determines the waypoint The target point is the foot from the current position of the drone to the tangent of the second trajectory made by the nearest waypoint; the drone is switched by the first trajectory via the target point Until running along the second track.
  • FIG. 16 is a schematic diagram of a principle of determining a target point according to an embodiment of the present application.
  • N can take a value of 20, and then the partial trajectory can be divided into 20 equal parts.
  • the specific solution process can be based on the following steps.
  • the vertical foot P 3 (x 3 , y 3 ) can be obtained and used as a target point. Through the target point, the drone is switched from the first trajectory to the second trajectory.
  • the waypoint closest to the current position of the drone may also be directly used as the target point.
  • using the vertical foot as the target point does not cause the problem of unevenness caused by the drone's backward movement.
  • switching the S140 drone from the first trajectory to running along the second trajectory may include: the drone performs at least a part of the trajectory on the second trajectory by N, etc. Points to obtain N + 1 waypoints; the drone determines from the N + 1 waypoints that is closest to the position where the drone is currently located and is located where the drone is currently located The waypoint in front of the movement of the position is used as the target point; the drone switches from the first trajectory to run along the second trajectory via the target point. Using the waypoint located in front of the drone's motion and closest to the current position of the drone as the target point can also avoid the problem of the drone retreating when the trajectory is switched.
  • switching the UAV from the first trajectory to running along the second trajectory through the target point may include: by proportion-integral-derivative (PID) control, the drone switches from the first trajectory to running along the second trajectory via the target point.
  • PID control is a technique used to correct the system response and perform adjustment control.
  • FIG. 17 is a schematic flowchart of a trajectory switching method 1700 according to another embodiment of the present application.
  • the method 1700 may include the following steps.
  • S1710 The UAV divides at least part of the trajectory on the second trajectory into N equal parts to obtain N + 1 waypoints, and the second trajectory is the trajectory to which the UAV is to be switched.
  • the drone determines, from the N + 1 waypoints, the waypoint closest to the current position of the drone.
  • the drone determines that the foot from the current position of the drone to the tangent of the second trajectory made by the nearest waypoint is the target point.
  • the drone switches to run along the second trajectory via the target point.
  • a plurality of waypoints are discretized into the trajectory to be switched to, a waypoint nearest to the current position of the drone is determined, and a trajectory made by the nearest waypoint
  • the tangent line uses the vertical foot as a target point, and the drone is switched to run along the trajectory through the target point, so that the drone can smoothly switch the trajectory and improve the user experience.
  • the drone switches to run along the second trajectory via the target point, which may include: through PID control, the drone passes the target point and is controlled by the first A track is switched to run along the second track.
  • the second trajectory may be an arc trajectory, a center of a circle where the second trajectory is located is a second center, a radius is a second radius, and the second center is located at a second position .
  • the method 1700 may further include: the drone runs along the first trajectory,
  • the first trajectory is an arc trajectory, the center of the circle where the first trajectory is located is the first center, the radius is the first radius, the first center is located at the first position, the first trajectory is tangent to the second trajectory, the The second position is an updated position of the point of interest relative to the first position.
  • the second position may be calculated by using a binocular camera or a monocular camera provided on the drone according to a vision measurement algorithm.
  • the second trajectory may be generated by Bezier curve fitting.
  • the Bezier curve may be a third-order Bezier curve.
  • one-mth of the circle where the second trajectory is located may be generated by the third-order Bezier curve fitting and one-mth of the circle where the second trajectory is located It includes four control points.
  • M 4
  • the second center point is marked as O 2
  • the four control points include a first control point A 2 , a second control point B 2 , and a third control point C 2 and a fourth control point D 2
  • the fifth control point A 2 is the current position of the drone
  • the fifth control point A 2 and the eighth control point D 2 are the circles where the second trajectory is located.
  • the sixth control point B 2 is on a tangent to the circle where the second trajectory passes through the fifth control point A 2
  • the sixth control point B 2 and the eighth control Point D 2 is located on the same side of O 2 A 2
  • the distance between the sixth control point B 2 and the fifth control point A 2 is h
  • the seventh control point C 2 passes through the eighth control point D 2
  • the seventh control point C 2 and the fifth control point A 2 are located on the same side of O 2 D 2
  • the seventh control point C 2 and the eighth control point D The distance between 2 is h, where
  • the second trajectory may be determined according to the first circle center, the first radius, the second circle center, the second radius, and the first trajectory.
  • the second trajectory is obtained by performing a circle center translation transformation and a radius scale transformation on the first trajectory.
  • FIG. 18 is a schematic block diagram of a trajectory switching apparatus 1800 according to an embodiment of the present application.
  • the trajectory switching device 1800 includes a processor 1810 and a memory 1820, where the memory 1810 is configured to store computer-executable instructions, and the processor 1820 is configured to execute the computer-executable instructions to implement the following Operation: Make the drone run along a first trajectory, the first trajectory is an arc trajectory, the center of the circle where the first trajectory is located is the first center, the radius is the first radius, and the first center is located at the first Determine the second position, the second position is the position of the point of interest updated relative to the first position; determine the second trajectory according to the second position, the second trajectory is also an arc trajectory, so The first trajectory is tangent to the second trajectory, the center of the circle where the second trajectory is located is the second center, the radius is the second radius, and the second center is located at the second position; The first track is switched to run along the second track.
  • the first position and / or the second position are calculated by a binocular camera or a monocular camera provided on the drone according to a vision measurement algorithm.
  • the first trajectory and / or the second trajectory are generated by Bezier curve fitting.
  • the Bezier curve is a third-order Bezier curve.
  • a circle in which the first trajectory and / or a circle in which the second trajectory is located is generated by the third-order Bezier curve fitting, and the first A circle on which one trajectory is located and / or one-Mth of a circle on which the second trajectory includes four control points.
  • M 4, the first center point is marked as O 1 , and the four control points include a first control point A 1 , a second control point B 1 , and a third control point C 1 and fourth control point D 1 , the first control point A 1 is the current position of the drone, and the first control point A 1 and the fourth control point D 1 are the first Two ends of a quarter of a circle where a trajectory is located, the second control point B 1 is on a tangent to the circle where the first trajectory passes through the first control point A 1 , and the second control Point B 1 and the fourth control point D 1 are located on the same side of O 1 A 1 , the distance between the second control point B 1 and the first control point A 1 is h, and the third control point Point C 1 is on a tangent to the circle where the first trajectory passes through the fourth control point D 1 , and the third control point C 1 and the first control point A 1 are located at the same position of O 1 D 1 Side, the distance between the third control point C 1
  • M 4
  • the second center point is marked as O 2
  • the four control points include a fifth control point A 2 , a sixth control point B 2 , and a seventh control point C 2 and the eighth control point D 2
  • the fifth control point A 2 is the current position of the drone
  • the fifth control point A 2 and the eighth control point D 2 are the first Two endpoints of a quarter of a circle where two trajectories are located
  • the sixth control point B 2 is on a tangent to the circle where the second trajectory passes through the fifth control point A 2
  • the sixth control Point B 2 and the eighth control point D 2 are located on the same side of O 2 A 2.
  • the distance between the sixth control point B 2 and the fifth control point A 2 is h, and the seventh control point Point C 2 is on a tangent to the circle through which the second trajectory of the eighth control point D 2 is located, and the seventh control point C 2 and the fifth control point A 2 are located at the same position of O 2 D 2 On the other hand, the distance between the seventh control point C 2 and the eighth control point D 2 is h,
  • the processor 1810 is specifically configured to determine according to the first circle center, the first radius, the second circle center, the second radius, and the first trajectory.
  • the second track is specifically configured to determine according to the first circle center, the first radius, the second circle center, the second radius, and the first trajectory. The second track.
  • the processor 1810 is specifically configured to perform a circle center translation transformation and a radius scale transformation on the first trajectory to obtain the second trajectory.
  • the processing instrument 1810 is configured to: determine a point on the second trajectory closest to a position where the drone is currently located as a target point; The human machine switches from the first trajectory to running along the second trajectory through the target point.
  • the processor 1810 is specifically configured to: divide at least part of the trajectory on the second trajectory into N equal parts to obtain N + 1 waypoints; and from the N + 1 Among the waypoints, a waypoint closest to a position where the drone is currently located is determined; and a perpendicularity of a tangent to the second trajectory made by the nearest waypoint to the drone is determined.
  • a foot is a target point; the UAV is switched from the first trajectory to run along the second trajectory through the target point.
  • the processor 1810 is specifically configured to: divide at least part of the trajectory on the second trajectory into N equal parts to obtain N + 1 waypoints; and from the N + 1 Among the waypoints, a waypoint that is closest to the current position of the drone and is located in front of the movement of the current position of the drone is determined as a target point; the drone passes through the The target point is switched from the first trajectory to running along the second trajectory.
  • the processor 1810 is specifically configured to enable the drone to switch from the first trajectory to the target position via the target point through proportional-integral-derivative PID control.
  • the second track operation is described.
  • the trajectory switching device in each embodiment of the present application may be implemented based on a module.
  • the trajectory switching device 1800 may include a control module for causing the drone to run along the first trajectory, and to switch the drone from the first trajectory to run along the second trajectory, and the like.
  • the trajectory switching device 1800 may include a calculation module for determining a second position, and determining a second trajectory according to the second position, where the second trajectory is also an arc trajectory, and the like.
  • Each module in the trajectory switching device may be used to execute a method in a corresponding embodiment of the present application, and details are not described herein again.
  • An embodiment of the present application further provides an unmanned aerial vehicle, which includes the foregoing trajectory switching device 1800.
  • FIG. 19 is a schematic block diagram of a trajectory switching apparatus 1900 according to an embodiment of the present application.
  • the trajectory switching device 1900 includes a processor 1910 and a memory 1920, where the memory 1920 is used to store computer-executable instructions, and the processor 1910 is used to execute the computer-executable instructions to implement the following Operation: N equal division of at least part of the trajectory on the second trajectory to obtain N + 1 waypoints, the second trajectory is the trajectory to be switched to by the drone; determined from the N + 1 waypoints Determine the waypoint closest to the position where the drone is currently located; determine the foot point of the tangent to the second trajectory made by the nearest waypoint as the target point; The drone switches to run along the second trajectory via the target point.
  • the processor 1910 is specifically configured to enable the drone to switch from the first trajectory to the destination via the target point through proportional-integral-derivative PID control.
  • the second track operation is described.
  • the second trajectory is an arc trajectory
  • the center of the circle where the second trajectory is located is the second center
  • the radius is the second radius
  • the second center is at the second position
  • the processor 1910 is further configured to: cause the drone to follow The first trajectory runs, the first trajectory is an arc trajectory, the center of the circle where the first trajectory is located is the first center, the radius is the first radius, the first center is at the first position, and the first Tangent to the second trajectory, the second position is a position of the point of interest updated relative to the first position.
  • the second position is calculated by using a binocular camera or a monocular camera provided on the drone according to a visual measurement algorithm.
  • the second trajectory is generated by Bezier curve fitting.
  • the Bezier curve is a third-order Bezier curve.
  • one-Mth of the circle in which the second trajectory is located is generated by the third-order Bezier curve fitting, and one-in-one M of the circle in which the second trajectory is located Includes four control points.
  • M 4
  • the second center point is marked as O 2
  • the four control points include a first control point A 2 , a second control point B 2 , and a third control point C 2 and a fourth control point D 2
  • the fifth control point A 2 is the current position of the drone
  • the fifth control point A 2 and the eighth control point D 2 are the first Two endpoints of a quarter of a circle where two trajectories are located
  • the sixth control point B 2 is on a tangent to the circle where the second trajectory passes through the fifth control point A 2
  • the sixth control Point B 2 and the eighth control point D 2 are located on the same side of O 2 A 2.
  • the distance between the sixth control point B 2 and the fifth control point A 2 is h, and the seventh control point Point C 2 is on a tangent to the circle through which the second trajectory of the eighth control point D 2 is located, and the seventh control point C 2 and the fifth control point A 2 are located at the same position of O 2 D 2 On the other hand, the distance between the seventh control point C 2 and the eighth control point D 2 is h,
  • the processor 1910 is specifically configured to determine according to the first circle center, the first radius, the second circle center, the second radius, and the first trajectory.
  • the second track is specifically configured to determine according to the first circle center, the first radius, the second circle center, the second radius, and the first trajectory. The second track.
  • the processor 1910 is specifically configured to perform a circle center translation transformation and a radius scale transformation on the first trajectory to obtain the second trajectory.
  • the trajectory switching device in each embodiment of the present application may be implemented based on a module.
  • the trajectory switching device 1900 may include a calculation module for dividing at least part of the trajectory on the second trajectory into N equal parts to obtain N + 1 waypoints, and determining a distance from the N + 1 waypoints to the drone.
  • the current waypoint is the closest waypoint, and the foot from which the drone is currently located to the tangent of the second trajectory made by the nearest waypoint is the target point.
  • the trajectory switching device 1900 may include a control module for causing the drone to switch to run along the second trajectory, etc., via the target point.
  • Each module in the trajectory switching device may be used to execute a method in a corresponding embodiment of the present application, and details are not described herein again.
  • An embodiment of the present application further provides an unmanned aerial vehicle, including the above-mentioned trajectory switching device 1900.
  • processors mentioned in the embodiments of the present application may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), and application-specific integrated circuits (DSPs).
  • DSPs digital signal processors
  • DSPs application-specific integrated circuits
  • ASIC application-specific integrated circuit
  • FPGA off-the-shelf programmable gate array
  • a general-purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
  • the memory mentioned in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memory.
  • the non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (erasable PROM, EPROM), electrical memory Erase programmable read-only memory (EPROM, EEPROM) or flash memory.
  • the volatile memory may be a random access memory (RAM), which is used as an external cache.
  • RAM random access memory
  • DRAM dynamic random access memory
  • SDRAM synchronous dynamic random access memory
  • double SDRAM double SDRAM
  • DDR SDRAM double data rate synchronous dynamic random access memory
  • enhanced SDRAM enhanced SDRAM
  • SLDRAM synchronous connection dynamic random access memory
  • direct RAMbus RAM direct RAMbus RAM
  • the processor is a general-purpose processor, a DSP, an ASIC, an FPGA, or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component
  • the memory memory module
  • memory described herein is intended to include, but is not limited to, these and any other suitable types of memory.
  • An embodiment of the present application further provides a computer-readable storage medium having instructions stored thereon.
  • the computer is caused to execute the methods of the foregoing method embodiments.
  • An embodiment of the present application further provides a computer program, which causes a computer to execute the methods of the foregoing method embodiments.
  • An embodiment of the present application further provides a computing device, where the computing device includes the computer-readable storage medium described above.
  • circuits, sub-circuits, and sub-units in the embodiments of the present application is merely schematic. Those of ordinary skill in the art may realize that the circuits, sub-circuits, and sub-units of the various examples described in the embodiments disclosed herein can be further split or combined.
  • a computer program product includes one or more computer instructions.
  • the computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable device.
  • the computer instructions may be stored in or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from a website site, computer, server, or data center via a wired (e.g., Coaxial cable, optical fiber, Digital Subscriber Line (DSL) or wireless (such as infrared, wireless, microwave, etc.) transmission to another website site, computer, server or data center.
  • the computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, a data center, or the like that includes one or more available medium integrations.
  • Usable media may be magnetic media (for example, floppy disks, hard disks, magnetic tapes), optical media (for example, high-density digital video discs (DVDs)), or semiconductor media (for example, solid state drives (Solid State Disks, SSDs) )Wait.
  • magnetic media for example, floppy disks, hard disks, magnetic tapes
  • optical media for example, high-density digital video discs (DVDs)
  • DVDs digital video discs
  • semiconductor media for example, solid state drives (Solid State Disks, SSDs)
  • an embodiment or “an embodiment” mentioned throughout the specification means that a particular feature, structure, or characteristic related to the embodiment is included in at least one embodiment of the present application.
  • the appearances of "in one embodiment” or “in an embodiment” appearing throughout the specification are not necessarily referring to the same embodiment.
  • the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
  • B corresponding to A means that B is associated with A, and B can be determined according to A.
  • determining B based on A does not mean determining B based on A alone, but also determining B based on A and / or other information.
  • the disclosed systems, devices, and methods may be implemented in other ways.
  • the device embodiments described above are only schematic.
  • the division of the unit is only a logical function division.
  • multiple units or components may be combined or Can be integrated into another system, or some features can be ignored or not implemented.
  • the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, which may be electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, may be located in one place, or may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the objective of the solution of this embodiment.
  • the functional units in the embodiments of the present application may be integrated into one processing unit, or each of the units may exist separately physically, or two or more units may be integrated into one unit.

Landscapes

  • Engineering & Computer Science (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

一种轨迹切换的方法和装置，该方法包括：（S110）无人机沿第一轨迹运行，该第一轨迹为圆弧轨迹，该第一轨迹所在圆的圆心为第一圆心，半径为第一半径，该第一圆心位于第一位置；（S120）该无人机确定第二位置，该第二位置是相对该第一位置更新后的兴趣点的位置；（S130）该无人机根据该第二位置确定第二轨迹，该第二轨迹也为圆弧轨迹，该第一轨迹与该第二轨迹相切，该第二轨迹所在圆的圆心为第二圆心，半径为第二半径，该第二圆心位于第二位置；（S140）该无人机由该第一轨迹切换至沿第二轨迹运行。该轨迹切换的方法通过更新兴趣点的位置，在两段圆弧轨迹间进行切换，两段圆弧轨迹相切，使得无人机在自主性飞行时的各段轨迹平滑连接，能够提升无人机的自主性操作，提高用户体验。

Description

轨迹切换的方法和装置
版权申明
本专利文件披露的内容包含受版权保护的材料。该版权为版权所有人所有。版权所有人不反对任何人复制专利与商标局的官方记录和档案中所存在的该专利文件或者该专利披露。
技术领域
本申请涉及无人机领域，尤其涉及一种无人机轨迹切换的方法和装置。
背景技术
兴趣点(point of interest，POI)是丰富导航地图的重要资讯。在地理信息系统中，POI可以是一栋房子、一个商铺、一个邮筒或一个公交站等。随着科技的发展，无人机被应用在越来越多的场景中，使无人机自主进行飞行操作是无人机领域发展的一个重要目标。在无人机领域，POI环绕是用户最感兴趣的飞行模式之一。现有的技术中，无人机通常是在获得地图上某个POI的准确坐标后，首先以最高效率例如直线飞行至POI附近，然后以一定的半径进行POI环绕。然而，这种POI环绕方式要求知道POI的准确坐标，飞行轨迹固化，无人机自主性操作差。此外，在POI环绕中，现有的无人机飞行时的各段轨迹联接和切换不平滑，用户体验较差。
发明内容
本申请提供了一种轨迹切换的方法和装置，使得无人机在自主性飞行时的各段轨迹平滑联接，能够提升无人机的自主性操作，提高用户体验。
第一方面，提供了一种轨迹切换的方法，包括：无人机沿第一轨迹运行，所述第一轨迹为圆弧轨迹，所述第一轨迹所在圆的圆心为第一圆心，半径为第一半径，所述第一圆心位于第一位置；所述无人机确定第二位置，所述第二位置是相对所述第一位置更新后的兴趣点的位置；所述无人机根据所述第二位置确定第二轨迹，所述第二轨迹也为圆弧轨迹，所述第一轨迹与所述第二轨迹相切，所述第二轨迹所在圆的圆心为第二圆心，半径为第二半径，所 述第二圆心位于第二位置；所述无人机由所述第一轨迹切换至沿第二轨迹运行。
第二方面，提供了一种轨迹切换的装置，包括：处理器和存储器，所述存储器用于存储计算机可执行指令，所述处理器用于执行所述计算机可执行指令，以实施以下操作：使无人机沿第一轨迹运行，所述第一轨迹为圆弧轨迹，所述第一轨迹所在圆的圆心为第一圆心，半径为第一半径，所述第一圆心位于第一位置；确定第二位置，所述第二位置是相对所述第一位置更新后的兴趣点的位置；根据所述第二位置确定第二轨迹，所述第二轨迹也为圆弧轨迹，所述第一轨迹与所述第二轨迹相切，所述第二轨迹所在圆的圆心为第二圆心，半径为第二半径，所述第二圆心位于第二位置；使所述无人机由所述第一轨迹切换至沿第二轨迹运行。
第三方面，提供了一种无人机，包括第二方面的轨迹切换的装置。
第四方面，提供了一种计算机可读存储介质，其上存储有指令，当指令在计算机上运行时，使得计算机执行第一方面的方法。
第五方面，提供了一种计算机程序，该计算机程序使得计算机执行第一方面的方法。
第一方面至第五方面，通过更新兴趣点的位置，在两段圆弧轨迹间进行切换，两段圆弧轨迹相切，使得无人机在自主性飞行时的各段轨迹平滑联接，能够提升无人机的自主性操作，提高用户体验。
第六方面，提供了一种轨迹切换的方法，包括：无人机将第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点，所述第二轨迹为无人机待切换至的轨迹；所述无人机从所述N+1个航点中确定距离所述无人机当前所处的位置最近的航点；所述无人机确定所述无人机当前所处的位置到经所述最近的航点做的所述第二轨迹的切线的垂足为目标点；所述无人机经由所述目标点，切换至沿所述第二轨迹运行。
第七方面，提供了一种轨迹切换的装置，包括：处理器和存储器，所述存储器用于存储计算机可执行指令，所述处理器用于执行所述计算机可执行指令，以实施以下操作：将第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点，所述第二轨迹为无人机待切换至的轨迹；从所述N+1个航点中确定距离无人机当前所处的位置最近的航点；确定所述无人机当前所处的位置到经所述最近的航点做的所述第二轨迹的切线的垂足为目标点；使所述 无人机经由所述目标点，切换至沿所述第二轨迹运行。
第八方面，提供了一种无人机，包括第七方面的轨迹切换的装置。
第九方面，提供了一种计算机可读存储介质，其上存储有指令，当指令在计算机上运行时，使得计算机执行第六方面的方法。
第十方面，提供了一种计算机程序，该计算机程序使得计算机执行第六方面的方法。
第六方面至第十方面，将待切换至的轨迹离散化出多个航点，确定距离该无人机当前所处的位置最近的航点，并经该最近的航点做的轨迹切线，将垂足作为目标点，经由该目标点，将无人机切换至沿该轨迹运行，可以使得无人机平滑地进行轨迹切换，可以提高用户体验。
附图说明
图1是本申请实施例的轨迹切换的方法的示意性流程图。
图2是本申请实施例提供的一种应用场景的示意图。
图3是本申请一个实施例的参考图像的示意图。
图4是本申请一个实施例的无人机的控制方法的示意性流程图。
图5是本申请实施例提供的无人机对参考对象进行环绕飞行的示意图。
图6是本申请另一实施例的无人机的控制方法的示意性流程图。
图7是本申请实施例提供的特征点跟踪的示意图。
图8是本申请实施例提供的三维坐标和像素坐标对应关系的示意图。
图9是本申请实施例提供的无人机对参考对象进行环绕飞行的示意图。
图10是本申请实施例提供的无人机对目标对象进行环绕飞行的示意图。
图11是本申请另一实施例的无人机的控制方法的示意性流程图。
图12是本申请实施例提供的特征点跟踪的示意图。
图13是本申请一个实施例的三阶贝塞尔曲线拟合生成圆形轨迹的原理的示意图。
图14是本申请一个实施例的第一圆心与四个控制点的示意图。
图15是本申请一个实施例的第一轨迹和第二轨迹的示意图。
图16是本申请一个实施例的确定目标点的原理的示意图。
图17是本申请另一实施例的轨迹切换的方法的示意性流程图。
图18是本申请一个实施例的轨迹切换的装置的示意性框图。
图19是本申请另一实施例的轨迹切换的装置的示意性框图。
具体实施方式
下面将结合附图，对本申请实施例中的技术方案进行描述。
需要说明的是，当组件被称为“固定于”另一个组件，它可以直接在另一个组件上或者也可以存在居中的组件。当一个组件被认为是“连接”另一个组件，它可以是直接连接到另一个组件或者可能同时存在居中组件。
除非另有定义，本文所使用的所有的技术和科学术语与属于本发明的技术领域的技术人员通常理解的含义相同。本文中在本发明的说明书中所使用的术语只是为了描述具体的实施例的目的，不是旨在于限制本发明。本文所使用的术语“及/或”包括一个或多个相关的所列项目的任意的和所有的组合。
下面结合附图，对本发明的一些实施方式作详细说明。在不冲突的情况下，下述的实施例及实施例中的特征可以相互组合。
除非另有定义，本文所使用的所有的技术和科学术语与属于本申请的技术领域的技术人员通常理解的含义相同。本文中在本申请的说明书中所使用的术语只是为了描述具体的实施例的目的，不是旨在于限制本申请。
本申请实施例的轨迹切换的方法和装置可以应用于带有单个或多个摄像头的无人机或其他载具，尤其是具有自主操作功能的载具，例如无人驾驶的汽车、自动飞行的无人机、无人地面机器人和无人船等。为了方便解释，本文以无人机来进行示意性说明。可以理解的是，本文中的无人机可以被同等地替代为上述载具。
在本申请的一些实施例中，无人机逐步精确地测算POI的位置(可以是三维位置也可以是省略高度的二维位置)，并可以自动环绕POI进行拍摄。由于初始阶段，无人机可能不知道POI的位置，一般按照预设半径(例如300m)来进行环绕。在随后的测算中，可以每次得到关于POI的位置一个测算结果，该测算结果会随测算次数的增加逐渐准确，直至在满足一定精度条件时，给出最终的测算结果。在这个过程中，因为POI的位置的测算可能不会一次就准确，所以环绕圆心和半径可以是变化的。这给航线规划以及轨迹的切换和控制带来了一定的挑战。
图1是本申请实施例的轨迹切换的方法100的示意性流程图。如图1所 示，方法100可以包括以下步骤。
S110，无人机沿第一轨迹运行，该第一轨迹为圆弧轨迹，该第一轨迹所在圆的圆心为第一圆心，半径为第一半径，该第一圆心位于第一位置(POI)。
S120，该无人机确定第二位置，该第二位置是相对该第一位置更新后的兴趣点的位置。
S130，该无人机根据该第二位置确定第二轨迹，该第二轨迹也为圆弧轨迹，该第一轨迹与该第二轨迹相切，该第二轨迹所在圆的圆心为第二圆心，半径为第二半径，该第二圆心位于第二位置。
S140，该无人机由该第一轨迹切换至沿第二轨迹运行。
本申请实施例的轨迹切换的方法，通过更新兴趣点的位置，在两段圆弧轨迹间进行切换，两段圆弧轨迹相切，使得无人机在自主性飞行时的各段轨迹平滑联接，能够提升无人机的自主性操作，提高用户体验。
具体而言，无人机先沿圆心为第一位置(初始的位置或者先测算的POI的位置)、半径为第一半径的圆弧、或圆形或多于一个完整圆形的轨迹运行。之后，无人机再次测算POI的位置，测算结果为第二位置，即第二位置是相对第一位置更新后的POI的位置。无人机根据新的第二位置确定新的轨迹即第二轨迹，该第二轨迹与第一轨迹相切(例如内切或外切)。无人机沿圆心为第二位置、半径为第二半径的圆弧、或圆形或多于一个完整圆形的轨迹运行。
应理解，第一轨迹与第二轨迹相切的切点可以是无人机当前所处的位置，即无人机立即切换轨迹。第一轨迹与第二轨迹相切的切点也可以是第一轨迹上无人机当前所处的位置之后的位置，即无人机继续沿第一轨迹运行一段之后再切换轨迹，本申请各实施例对此不做限定。
还应理解，本文中兴趣点还可以被称作目标对象。
可选地，在本申请一些实施例中，该第一位置和/或该第二位置可以为通过该无人机上设置的双目相机或单目相机根据视觉测量算法计算得到的。即，第一位置和/或第二位置可以是逐步精确地测算POI的位置的过程中测算得到的。
具体地，第一位置和/或第二位置可以是通过以下步骤S1和S2得到的。
步骤S1，无人机获取目标对象的指示信息，其中，所述指示信息包括目标对象在拍摄装置输出的参考图像中的位置信息。
图2是本申请实施例提供的一种应用场景的示意图。如图2所示，无人机20搭载有拍摄装置21，拍摄装置21具体可以是相机、摄像机等。具体的，拍摄装置21可通过云台22搭载在无人机20上，或者，拍摄装置21通过其他固定装置固定在无人机20上。该拍摄装置21可实时拍摄获得视频数据或图像数据，并将该视频数据或图像数据通过无人机20的无线通讯接口23发送给控制终端24，该控制终端24具体可以是无人机20对应的遥控器，也可以是用户终端例如智能手机、平板电脑等。另外，该无人机20还可以包括控制装置，该控制装置可以包括通用或者专用的处理器，此处只是示意性说明，并不限定该无人机的具体结构。
可选的，拍摄装置21拍摄的图像中包括如图2所示的目标对象31，此处，将拍摄装置21输出的某一帧图像记为参考图像，该无人机20的处理器可获取该目标对象的指示信息，该指示信息包括该目标对象在该参考图像中的位置信息。
作为一种可能的方式，所述获取目标对象的指示信息,包括：接收控制终端发送的指示信息，其中，所述指示信息是所述控制终端检测用户在显示所述参考图像的交互界面上的目标对象选择操作确定的。
具体的，拍摄装置21输出该参考图像后，通过无线通讯接口23将该参考图像发送给控制终端24，控制终端24将该参考图像显示在交互界面中，使得用户可以在该交互界面中对该参考图像中的目标对象进行选择。图3是本申请一个实施例的参考图像的示意图。如图3所示，30表示该交互界面中显示的参考图像，参考图像30中包括目标对象31，用户在该交互界面中对目标对象31进行选择的一种可能方式是：用户选择点32，并从点32开始滑动到点33，此处只是示意性说明，本实施例不限定具体的选择操作。控制终端24可根据用户在该交互界面中的选择操作，确定出用户在该交互界面中框选的区域34，并确定出区域34在该参考图像30中的位置信息，例如，控制终端24可确定出该区域34的左上角即点32在该参考图像30中的位置信息和该区域34的大小例如长、宽，或者，控制终端24可确定该区域34的左上角即点32在该参考图像30中的位置信息和该区域34的右下角即点33在该参考图像30中的位置信息。进一步，控制终端24可将该区域34在该参考图像30中的位置信息作为该目标对象31在该参考图像30中的位置信息即该目标对象31的指示信息发送给无人机20。
作为另一种可能的方式，所述获取目标对象的指示信息,包括：对参考图像中的目标对象进行识别以获取目标对象的指示信息。
具体地，无人机的处理器可以对拍摄装置21输出的参考图像中的目标对象进行识别，通过识别获取目标对象的指示信息。进一步地，无人机的处理器可以将参考图像输入到已经训练好的神经网络模型，并获取所述神经网络模型输出的目标对象的指示信息。
步骤S2，根据所述指示信息确定所述目标对象的位置信息。
具体地，在获取到目标对象的指示信息之后，无人机可以根据所述指示信息确定目标对象的位置信息，其中，所述目标对象的位置信息可以为三维位置信息或者二维位置信息；所述目标对象的位置信息可以为基于世界坐标系中的位置信息；另外，所述目标对象的位置信息也可以为基于全局坐标系中的位置信息，所述位置信息可以至少包括经度和纬度；此外，所述目标对象的位置信息还可以为基于无人机的机体坐标系中的位置信息。
作为一种可实现方式，所述根据所述指示信息确定所述目标对象的位置信息包括：根据所述指示信息确定目标对象相对于无人机的朝向，根据所述朝向和无人机与目标对象之间的水平距离或者无人机的对地高度值确定目标对象的位置信息。
具体地，本实施例可根据目标对象31在该参考图像30中的位置信息、承载拍摄装置21的云台的姿态确定目标对象31相对于无人机20的朝向；再根据所述朝向、目标对象31与无人机20之间的水平距离确定目标对象31的位置信息。其中，拍摄装置21具有的视场角(FOV)是已知的，根据目标对象31在参考图像中的位置信息可以确定目标对象31相对于拍摄装置21的光轴的角度，例如：若目标对象31在参考图像的正中心，则说明目标对象31相对于拍摄装置的光轴的角度为0，若拍摄装置21的FOV在水平方向为20度，若目标对象31在参考图像的最左边，则说明目标对象31相对于拍摄装置21的光轴的水平角度为10度，垂直方向上也类似，此处不再赘述；另外,拍摄装置21的云台22的姿态也决定了拍摄装置21的光轴的朝向，结合目标对象31相对于拍摄装置21的光轴的角度以及光轴的朝向，可以获得目标对象31相对于无人机20的朝向。进一步,根据目标对象31相对于无人机20的朝向、目标对象31与无人机20之间的水平距离确定目标对象31的位置信息。在某些实施例中，根据目标对象31相对于无人机20的朝向、目 标对象31与无人机20之间的水平距离或者无人机20的对地高度值确定目标对象31的位置信息。
继续参考图2，根据目标对象31相对于无人机20的朝向可确定目标对象31相对于无人机20在俯仰方向上的角度，例如图2所示的α角，然后，可以获取无人机20上配置的距离传感器测量的无人机的对地高度值，例如图2所示的h，根据所述α角和对地高度值即可以确定目标对象相对于无人机在垂直方向上的位置信息，另外，根据目标对象31相对于无人机20的朝向还可确定目标对象31相对于无人机20在偏航方向上的角度例如β角，根据β角和目标对象31与无人机20之间的水平距离L即可以确定目标对象相对于无人机在水平方向上的位置信息；根据目标对象相对于无人机在垂直方向上的位置信息和目标对象相对于无人机在水平方向上的位置信息即可确定出目标对象相对于无人机的位置信息，进一步，根据目标对象相对于无人机的位置信息和无人机的位置信息可确定出目标对象的位置信息，该目标对象的位置信息可以是该目标对象在世界坐标系中的位置，也可以是该目标对象在全局坐标系中的位置。
另外，在某些实施例，也可以根据目标对象31与无人机20之间的水平距离L和所述α角确定目标对象相对于无人机在垂直方向上的位置信息。其中，目标对象的指示信息可以指示目标对象在参考图像中对应的图像区域的大小，所述目标对象31与无人机20之间的水平距离可以根据所述图像区域的大小确定。
在步骤S2后，可以根据所述目标对象的位置信息控制所述无人机环绕所述目标对象移动。
具体的，当无人机20的处理器确定出目标对象31的位置信息后，以该目标对象31为中心，根据无人机20与目标对象31之间的位置关系生成环绕轨迹，并控制无人机20在该环绕轨迹上移动，即控制无人机20在该环绕轨迹上飞行，实现环绕目标对象31飞行。在无人机20环绕该目标对象31飞行的过程中，拍摄装置21可实时拍摄目标对象31，并将拍摄获得的图像数据或视频数据通过无线通讯接口23发送给控制终端24，以便用户浏览观看。
本实施例通过获取拍摄装置拍摄的目标对象在该拍摄装置输出的参考图像中的位置信息，确定该目标对象的位置信息，并根据该目标对象的位置 信息控制无人机环绕该目标对象移动，使得无人机不需要移动到环绕中心去记录环绕中心的位置，即可实现该无人机环绕该目标对象移动，简化了无人机实现对目标对象环绕移动的过程，同时提高了无人机的操作安全性。
本申请实施例提供一种无人机的控制方法。图4是本申请另一实施例提供的无人机的控制方法的流程图。如图4所示，在图1所示实施例的基础上，本实施例提供了根据所述指示信息确定所述目标对象的位置信息的另一种可实现方式，具体的，根据所述指示信息确定所述目标对象的位置信息，可以包括如下步骤：
步骤S401、控制所述无人机对参考对象进行环绕移动。
在本实施例中，可在无人机的正前方预设距离处取一个点为参考对象，该参考对象具体为虚拟目标点，控制无人机对该参考对象进行环绕飞行。图5为本申请实施例提供的无人机对参考对象进行环绕飞行的示意图。如图5所示，50表示无人机正前方预设距离处的参考对象，51表示无人机的机头，无人机内的处理器具体可控制该无人机对参考对象50进行环绕飞行。
作为一种可能的方式，所述控制所述无人机对参考对象进行环绕移动，包括：根据预设的环绕半径确定参考对象，控制所述无人机对参考对象进行环绕移动。
具体的，控制该无人机以参考对象50为环绕中心，以预设的环绕半径例如500米为半径，生成一条环形轨迹，如图5所示的环形轨迹53，并控制该无人机在该环形轨迹53上对参考对象50进行环绕飞行。可选的，该无人机可沿着逆时针方向在该环形轨迹53上飞行，也可以沿着顺时针方向在该环形轨迹53上飞行。可选的，当无人机内的处理器接收到控制终端发送的目标对象的指示信息后，即可根据该预设的环绕半径确定参考对象，并控制该无人机对参考对象进行环绕飞行。也就是说，当用户在参考图像中框选出目标对象后，无人机即可对参考对象进行环绕飞行。
作为另一种可能的方式，所述控制所述无人机对参考对象进行环绕移动，包括：在接收到所述控制终端发送的启动控制指令后，控制所述无人机对参考对象进行环绕移动。
具体的，如图3所示，当用户框选目标对象31后，该交互界面中可显示启动控制按键35，该启动控制按键35具体可以是该交互界面中的一个图标，也就是说，当用户框选目标对象31后，无人机并不立即对参考对象进 行环绕飞行，而是等到用户在该交互界面中点击该启动控制按键35后，该无人机才开始对参考对象进行环绕飞行。具体的，当用户在该交互界面中点击该启动控制按键35时，控制终端根据用户的点击操作生成启动控制指令，并将该启动控制指令发送给无人机，当无人机内的处理器接收到该启动控制指令后，控制无人机对参考对象进行环绕飞行，具体的环绕控制方式可以是如图5所示的方式，此处不再赘述。
步骤S402、在所述无人机环绕参考对象移动的过程中，获取所述拍摄装置输出的多帧第一目标图像，其中，所述第一目标图像中包括所述目标对象。
如图5所示，以该无人机沿着顺时针方向在该环形轨迹53上飞行为例，当该无人机在对参考对象50进行环绕飞行的过程中，该无人机的拍摄装置还可以对目标对象31进行拍摄，并输出包括该目标对象31的目标图像，本实施例将该无人机在对参考对象50进行环绕飞行的过程中拍摄的目标图像记为第一目标图像，且该无人机的拍摄装置输出的第一目标图像可以是多帧。具体的，该无人机在对参考对象50进行环绕飞行的过程中，该无人机的处理器可获取到拍摄装置输出的多帧第一目标图像，该第一目标图像中包括目标对象31。此处不限定目标对象31相对于拍摄装置的光轴偏移的角度，只要保证目标对象31在拍摄装置的拍摄画面中即可。
步骤S403、根据所述目标对象的指示信息和所述多帧第一目标图像确定所述目标对象的位置信息。
该无人机的处理器可根据上述实施例中获取到的目标对象31的指示信息，以及上述步骤中获取到的多帧第一目标图像确定目标对象31的位置信息。
可选的，所述根据所述目标对象的指示信息和所述多帧第一目标图像确定所述目标对象的位置信息，可以包括如图6所示的如下步骤：
步骤S601、获取所述参考图像的目标区域中的特征点，其中，所述目标区域为所述参考图像中所述指示信息指示的图像区域。
当无人机接收到控制终端发送的目标对象的指示信息后，可根据该目标对象的指示信息确定出该参考图像的目标区域，该目标区域具体为该指示信息指示的图像区域。例如图3所示，无人机接收到控制终端发送的区域34在该参考图像30中的位置信息或者通过识别获取区域34在该参考图像30 中的位置信息后，该无人机的处理器可在该参考图像30中确定出目标区域，该目标区域具体可以是区域34，也就是说，无人机可以将用户在交互界面中框选的区域作为该目标区域。进一步，该无人机的处理器可获取该目标区域中的特征点，可选的，该处理器可根据预设的特征点提取算法确定出该目标区域中的特征点，该特征点提取算法包括如下至少一种：Harris角点检测算法、尺度不变特征变换(Scale-invariant feature transform，SIFT)、加速稳健特征(Speeded Up Robust Features，SURT)算法、快速特征点提取和描述的算法(Oriented FAST and Rotated BRIEF，ORB)等。可选的，本实施例采用Harris角点检测算法提取该目标区域中的特征点。
步骤S602、基于所述参考图像的目标区域中的特征点利用跟踪算法获取每一帧第一目标图像的特征点。
在获取到该目标区域中的特征点后，利用跟踪算法对该目标区域中的特征点进行跟踪,即利用跟踪算法确定该目标区域中的特征点在每一帧第一目标图像中的位置。该跟踪算法具体可以是(Kanade-Lucas-Tomasi Feature Tracker，KLT)特征跟踪算法。
图7是本申请实施例提供的特征点跟踪的示意图。如图7所示，A、B、C、D、E、F、G分别表示参考图像30的目标区域即区域34中的特征点，特征点A、B、C、D、E、F、G也是目标对象31的特征点。71、72、73分别表示在该无人机环绕参考对象飞行的过程中，拍摄装置依次输出的第一目标图像。根据KLT特征跟踪算法可确定出参考图像30中目标对象31的特征点例如A、B、C、D、E、F、G分别在第一目标图像71、第一目标图像72、第一目标图像73中的位置。例如，拍照装置先输出参考图像30，后续依次输出第一目标图像71、第一目标图像72、第一目标图像73；参考图像30、第一目标图像71、第一目标图像72、第一目标图像73可以是相邻图像，也可以是非相邻图像。
如图5所示，在该无人机环绕参考对象50飞行的过程中，目标对象31相对于该无人机的位置不断变化，导致目标对象31在拍摄装置依次输出的第一目标图像中的位置不断变化，从而使得第一目标图像71、第一目标图像72、第一目标图像73中目标对象31对应的特征点在对应的第一目标图像中的位置不断变化。此处只是示意性说明，并不限定区域34中特征点的个数，也不限定第一目标图像的个数，以及区域34中的特征点在每一帧第一目标 图像中的位置。
步骤S603、根据每一帧第一目标图像的特征点在对应的第一目标图像中的位置信息确定目标对象的位置信息。
例如，根据第一目标图像71、第一目标图像72、第一目标图像73中目标对象31对应的特征点在对应的第一目标图像中的位置信息确定目标对象31的位置信息，确定出的该目标对象31的位置信息具体为目标对象31在三维空间中的三维坐标。此处，将根据第一目标图像71、第一目标图像72、第一目标图像73中目标对象31对应的特征点在对应的第一目标图像中的位置信息确定出的目标对象31的位置信息记为第一位置信息。
可以理解，在拍摄装置输出第一目标图像73之后还会输出新的第一目标图像，根据KLT特征跟踪算法可确定出目标对象31的特征点在该新的第一目标图像中的位置；进一步根据第一目标图像71、第一目标图像72、第一目标图像73、新的第一目标图像中目标对象31对应的特征点在对应的第一目标图像中的位置信息可确定出又一个目标对象31的位置信息，此处，将该目标对象31的位置信息记为第二位置信息。上述的第一位置信息和此处的第二位置信息可能相同，也可能不同，但可以理解的是，随着拍摄装置不断输出新的第一目标图像，根据第一目标图像71、第一目标图像72、第一目标图像73、以及拍摄装置后续不断输出的第一目标图像中目标对象31对应的特征点在对应的第一目标图像中的位置信息确定出的目标对象31的位置信息的精准度不断提高。一种可能的方式是，拍摄装置每输出一帧新的第一目标图像，无人机的处理器即可确定出该目标对象31的一个新的位置信息。
可选的，所述根据每一帧第一目标图像的特征点在对应的第一目标图像中的位置信息确定目标对象的位置信息,包括：基于每一帧第一目标图像的特征点在对应的第一目标图像中的位置信息利用拟合算法确定所述目标对象的位置信息。
图8是本申请实施例提供的三维坐标和像素坐标对应关系的示意图。如图8所示，80表示目标对象，81、82、83表示拍摄装置环绕目标对象80按照箭头所示的方向移动的过程中，拍摄装置先后输出的第一目标图像，可以理解，目标对象80上的三维点可映射到第一目标图像81、82、83中，该三维点在第一目标图像81、82、83中的映射点具体可以是第一目标图像81、 82、83中的特征点，从第一目标图像81到第一目标图像83的过程中，可跟踪到的特征点的个数在减少。
例如，点A、点B和点C分别为目标对象80上的三维点，点a1、点b1和点c1表示第一目标图像81中的特征点，点a1与点A对应，点b1与点B对应,点c1和点C对应；点a2、点b2和点c2表示第一目标图像82中的特征点，点a2与点A对应，点b2与点B对应，点c2和点C对应；点a3和点b3表示第一目标图像83中的特征点，点a3与点A对应，点b3与点B对应。此处只是示意性说明，并不限定目标对象80、目标对象80上的三维点、以及目标对象80上的三维点在第一目标图像中的映射点。可以理解的是，目标对象80在不同第一目标图像中的位置不同，目标对象80上的同一个三维点在不同第一目标图像中的映射点在对应的第一目标图像中的位置也不同。
根据世界坐标系和像素平面坐标系的转换关系，可得到目标对象80上的三维点在世界坐标系中的三维坐标(x w,y w,z w)与该三维点在第一目标图像中的映射点在该第一目标图像中的位置信息例如像素坐标(μ,υ)的关系，该关系具体如下公式(1)所示：
Figure PCTCN2018097944-appb-000001
其中，z c表示该三维点在相机坐标系Z轴上的坐标，K表示相机的内参，R表示相机的旋转矩阵，T表示相机的平移矩阵。在本实施例中，(μ,υ)、K、R、T为已知量，z c和(x w,y w,z w)为未知量。在拍摄装置拍摄不同的第一目标图像时，K是不变的，R、T可以是变化的。
具体的，根据点a1在第一目标图像81中的像素坐标、以及该拍摄装置拍摄第一目标图像81时对应的R、T，可建立一个如公式(1)所示的方程，根据点a2在第一目标图像82中的像素坐标、以及该拍摄装置拍摄第一目标图像82时对应的R、T，可建立另一个如公式(1)所示的方程，根据点a3在第一目标图像83中的像素坐标、以及该拍摄装置拍摄第一目标图像83时对应的R、T，可建立再一个如公式(1)所示的方程，随着该拍摄装置不断输出新的第一目标图像，建立的如公式(1)所示的方程逐渐增加，可以 理解，当方程组中方程的个数大于未知量的个数时，可求解出相应的未知量。也就是说，利用拟合算法对这些方程进行求解即可计算出三维点A在世界坐标系中的三维坐标。同理，可计算出三维点B和三维点C在世界坐标系中的三维坐标，此处不再赘述。可以理解，该拍摄装置输出的第一目标图像越多，基于该多帧第一目标图像中的特征点的像素坐标利用拟合算法得到的三维点在世界坐标系中的三维坐标越精确。当确定出目标对象80上的多个三维点例如三维点A、B、C在世界坐标系中的三维坐标后，根据三维点A、B、C在世界坐标系中的三维坐标可确定出目标对象80在世界坐标系中的三维坐标。无人机可以根据目标对象80在世界坐标系中的三维坐标获取目标对象的位置信息。例如，当目标对象31的位置信息为基于全局坐标系中的位置时，可以根据无人机的位置信息和目标对象80在世界坐标系中的三维坐标确定目标对象31的位置信息。当目标对象31的位置信息为基于无人机的机体坐标系中的位置时，可以将目标对象80在世界坐标系中的三维坐标转换到机体坐标系以获取基于所述机体坐标系的位置信息。
另外，所述方法还包括：在获取每一帧第一目标图像的特征点之后，从每一帧第一目标图像的特征点中确定满足预设要求的目标特征点；相应的，所述根据每一帧第一目标图像的特征点在对应的第一目标图像中的位置信息确定目标对象的位置信息,包括：根据每一帧第一目标图像的目标特征点在对应的第一目标图像中的位置信息确定所述目标对象的位置信息。
如图7所示，在获取到第一目标图像71、第一目标图像72、第一目标图像73中的特征点例如A、B、C、D、E、F、G之后，从第一目标图像71、第一目标图像72、第一目标图像73的特征点中确定满足预设要求的目标特征点，例如，每个特征点在第一目标图像71和参考图像30之间的偏移量可能是不同的，假设特征点A在第一目标图像71和参考图像30之间的偏移量记为h1、特征点B在第一目标图像71和参考图像30之间的偏移量记为h2、依次类推，特征点G在第一目标图像71和参考图像30之间的偏移量记为h7，计算h1、h2、…h7的平均值和方差，平均值记为u，方差记为δ 2，根据高斯分布选取偏移量在[u-3δ,u+3δ]内的特征点为目标特征点，假设h1在[u-3δ,u+3δ]外，则将第一目标图像71中的特征点A删除，保留第一目标图像71中的特征点B、C、D、E、F、G，将第一目标图像71中的特征点B、C、D、E、F、G作为第一目标图像71的目标特征点。同理，可计算出 第一目标图像72和第一目标图像73中的目标特征点，此处不再赘述。
在其他实施例中，根据每个特征点在第一目标图像71和参考图像30之间的偏移量例如h1、h2、…h7，计算出h1、h2、…h7的平均值和方差后，根据高斯分布选取偏移量在[u-3δ,u+3δ]内的特征点为有效点，例如，h1在[u-3δ,u+3δ]外，则将第一目标图像71中的特征点A删除，将第一目标图像71中的特征点B、C、D、E、F、G作为有效点，进一步从该有效点中确定出目标特征点，从该有效点中确定出目标特征点的一种可能的方式是，计算有效点对应的偏移量的平均值，即计算h2、…h7的平均值记为u1。此处将区域34在参考图像30中的位置信息记为ROI 0，根据ROI 0和u1可确定出区域34在第一目标图像71中的位置信息记为ROI 1，具体的，ROI 1＝ROI 0+u1，进一步根据区域34在第一目标图像71中的位置信息ROI 1，以及有效点B、C、D、E、F、G在第一目标图像71中的位置信息，确定出有效点B、C、D、E、F、G中哪些点在该区域34内，哪些点不在该区域34内，将有效点B、C、D、E、F、G中不在该区域34内的点进一步剔除掉，剩余的有效点作为该第一目标图像71的目标特征点，同理，可计算出第一目标图像72和第一目标图像73中的目标特征点，此处不再赘述。
通过上述方法确定出第一目标图像71、第一目标图像72和第一目标图像73中的目标特征点后，根据目标特征点在对应的第一目标图像中的位置信息确定目标对象31在世界坐标系中的三维坐标，具体原理与图8所示的原理一致，此处不再赘述。
本实施例通过控制无人机对参考对象进行环绕飞行，在无人机环绕参考对象飞行的过程中获取拍摄装置输出的多帧第一目标图像，根据目标对象的指示信息和多帧第一目标图像确定该目标对象的位置信息，在拍摄装置不断输出第一目标图像时，根据目标对象的指示信息和拍摄装置不断输出的第一目标图像可不断的确定出该目标对象的位置信息，且该目标对象的位置信息的准确度不断提高；另外，在获取到拍摄装置输出的每一帧第一目标图像的特征点之后，从每一帧第一目标图像的特征点中确定满足预设要求的目标特征点，在根据每一帧第一目标图像的目标特征点在对应的第一目标图像中的位置信息确定该目标对象的位置信息时，可提高该目标对象的位置信息的精确度，同时去除不满足该预设要求的特征点，还可降低相应的计算量。
本申请实施例提供一种无人机的控制方法。在上述实施例的基础上，所 述方法还包括：根据每一帧第一目标图像的特征点在对应的第一目标图像中的位置信息确定所述无人机在环绕参考对象移动过程中所述拍摄装置相对于目标对象的视差；相应的，所述根据所述目标对象的位置信息控制所述无人机环绕所述目标对象移动,包括：当所述视差大于第一预设视差阈值时，根据所述确定出的所述目标对象的位置信息确定所述无人机对所述目标对象进行环绕移动的环绕轨迹，并控制所述无人机在所述环绕轨迹上移动。
如图7所示，在获取到第一目标图像71、第一目标图像72、第一目标图像73中的特征点例如A、B、C、D、E、F、G之后，根据特征点A、B、C、D、E、F、G分别在第一目标图像71、第一目标图像72、第一目标图像73中的位置信息，可确定出如图5所示的无人机在环绕参考对象50飞行过程中，该无人机的拍摄装置相对于目标对象31的视差，例如，第一目标图像71是无人机在m1位置时拍摄装置拍摄的图像，第一目标图像72是无人机在m2位置时拍摄装置拍摄的图像，第一目标图像73是无人机在m3位置时拍摄装置拍摄的图像。根据特征点A、B、C、D、E、F、G分别在第一目标图像71、第一目标图像72中的位置信息，可确定出无人机从m1位置到m2位置的过程中，该无人机的拍摄装置相对于目标对象31的视差；具体的，将特征点A在第一目标图像71中的像素坐标记为(μ 1,υ 1)，将特征点A在第一目标图像72中的像素坐标记为(μ 2,υ 2)，根据如下公式(2)可计算出特征点A的视差，特征点A的视差记为parallaxA：
Figure PCTCN2018097944-appb-000002
其中，R 21表示相机在拍摄第一目标图像72时的姿态相对于相机在拍摄第一目标图像71时的姿态在旋转方向上的变化。c x和c y表示相机光心位置，可以理解，该相机光心在第一目标图像71和第一目标图像72中的位置相同。f表示该相机的焦距。同理，可计算出特征点B、C、D、E、F、G的视差，对特征点A、B、C、D、E、F、G的视差取平均值，该平均值为第一目标图像72的视差，第一目标图像72的视差为无人机从m1位置到m2位置的过程中，该无人机的拍摄装置相对于目标对象31的视差。
同理，根据特征点A、B、C、D、E、F、G分别在第一目标图像71、第一目标图像73中的位置信息，可确定出第一目标图像73的视差，该第一 目标图像73的视差为无人机从m1位置到m3位置的过程中，该无人机的拍摄装置相对于目标对象31的视差。可以理解，随着无人机沿着环形轨迹53飞行的过程中，该无人机的拍摄装置相对于目标对象31的视差不断变大，利用拟合算法不断的确定目标对象31的三维坐标，所述视差越大，确定出的目标对象的三维坐标的准确度越高，当该无人机的拍摄装置相对于目标对象31的视差大于第一预设视差阈值时，停止拟合算法，获取最新确定出的目标对象31的三维坐标，即目标对象31的精准三维坐标，并根据最新确定出的目标对象31的三维坐标确定无人机对该目标对象31进行环绕飞行的环绕轨迹，该环绕轨迹不同于无人机对参考对象50的环绕轨迹53。
图9是本申请实施例提供的无人机对参考对象进行环绕飞行的示意图。如图9所示，假设无人机沿环绕轨迹53飞行到m3位置时，该无人机的拍摄装置相对于目标对象31的视差大于第一预设视差阈值，则根据最新确定出的目标对象31的三维坐标和预设环绕参数例如环绕半径确定无人机对该目标对象31进行环绕飞行的目标轨迹91，并控制无人机沿着目标轨迹91飞行。
另外，所述方法还包括：确定所述视差的变化速度；根据所述视差的变化速度调节所述无人机环绕参考对象移动的速度。
可选的，所述确定所述视差的变化速度,包括：根据多帧第一目标图像中相邻的两帧第一目标图像的特征点在对应的第一目标图像中的位置信息确定所述视差的变化速度。
例如，第一目标图像71和第一目标图像72是拍摄装置拍摄的多帧第一目标图像中相邻的两帧第一目标图像，将第一目标图像71的视差记为PA i-1，第一目标图像72的视差记为PA i，视差的变化速度记为parallax_speed，parallax_speed＝(PA i-PA i-1)/t，t表示第一目标图像71和第一目标图像72之间的时间间隔，如果拍摄装置拍摄第一目标图像的频率是固定的例如30HZ，则parallax_speed还可以表示为parallax_speed＝(PA i-PA i-1)，即在图像频率固定的情况下，衡量(PA i-PA i-1)/t的大小与衡量PA i-PA i-1的大小其意义是一致的。
具体的，当无人机沿着环形轨迹53开始飞行时，该无人机可按照预设的较小的速度例如2m/s飞行，但是，如果目标对象31距离无人机较远，当无人机沿着环形轨迹53飞了较长时间后，目标对象31在拍摄装置拍摄的第一目标图像中的位置可能变化较小，或者几乎没变化，在这种情况下，可以 根据视差的变化速度，调整无人机沿着环形轨迹53飞行的飞行速度。例如，第一预设视差阈值记为T1，假设T1＝20,无人机从开始沿着环形轨迹53飞行需要在例如t＝2秒内确定出目标对象31的三维坐标，也就是说，需要该无人机的拍摄装置相对于目标对象31的视差在t＝2秒内达到第一预设视差阈值T1，则期望的视差的变化速度为T1/t＝10,假设根据parallax_speed＝(PA i-PA i-1)计算出当前的parallax_speed为2.5，则需要提高无人机的飞行速度，无人机需要达到的飞行速度＝无人机当前的飞行速度*(期望的视差的变化速度/当前的parallax_speed)即2m/s*(10/2.5)＝8m/s，也就是说，需要将无人机的飞行速度提高到8m/s。
此外，所述方法还包括：当所述视差大于第二预设视差阈值时，根据所述确定出的所述目标对象的位置信息调整所述无人机对参考对象进行环绕移动的半径，其中，所述第一预设视差阈值大于所述第二预设视差阈值。
如图9所示，如果无人机沿环绕轨迹53飞行到m3位置时，该无人机的拍摄装置相对于目标对象31的视差大于第一预设视差阈值，则可根据最新确定出的目标对象31的三维坐标确定无人机对该目标对象31进行环绕飞行的目标轨迹91，但是此时无人机距离该目标轨迹91可能较远，无人机需要从当前位置例如m3位置飞行到目标轨迹91上的一点再开始沿着目标轨迹91飞行。
作为一种可替换方式，如图10所示，假设无人机沿环绕轨迹53飞行到m2位置时，该无人机的拍摄装置相对于目标对象31的视差大于第二预设视差阈值，该第二预设视差阈值小于第一预设视差阈值。此时，利用拟合算法已经可以确定出目标对象31的三维坐标，即目标对象31的粗略三维坐标，并根据该目标对象31的三维坐标和预设环绕参数例如环绕半径确定出无人机对该目标对象31进行环绕飞行的目标轨迹91，即粗略的目标轨迹91，则从m2位置开始，可以不断调整无人机对参考对象50进行环绕飞行的半径，例如不断减小无人机对参考对象50进行环绕飞行的半径，在该无人机以不断减小的环绕半径对参考对象50进行环绕飞行的过程中，该无人机的拍摄装置相对于目标对象31的视差还在不断变化。当该无人机的拍摄装置相对于目标对象31的视差大于第一预设视差阈值时，无人机可能会到达目标轨迹91(准确的目标轨迹)上一点例如m4，或者无人机可能会到达距离目标轨迹91较近的一点，使得无人机可以从该点平滑过渡到目标轨迹91上。
本实施例通过每一帧第一目标图像的特征点在对应的第一目标图像中的位置信息确定无人机在环绕参考对象飞行过程中该拍摄装置相对于目标对象的视差，根据该视差的变化速度调节无人机环绕参考对象飞行的飞行速度，使得无人机可以在较短的时间内确定出目标对象的三维坐标，尤其是当目标对象距离无人机较远、无人机环绕参考对象飞行的飞行速度较小时，通过该视差的变化速度可提高该无人机的飞行速度，提高计算目标对象的三维坐标的效率；另外，通过设置至少两个视差阈值例如第一预设视差阈值和第二预设视差阈值，第一预设视差阈值大于第二预设视差阈值，当该视差大于第二预设视差阈值时，通过调整该无人机对参考对象进行环绕飞行的半径，可使得当该视差大于第一预设视差阈值时，无人机到达对目标对象进行环绕飞行的环绕轨迹上，或者到达距离该环绕轨迹较近的位置，从而使得无人机可以从对参考对象进行环绕飞行的环绕轨迹上平滑过渡到对目标对象进行环绕飞行的环绕轨迹上。
本申请实施例提供一种无人机的控制方法。图11为本申请另一实施例提供的无人机的控制方法的流程图。如图11所示，在上述实施例的基础上，所述方法还包括：在获取到所述指示信息之后，根据所述指示信息控制拍摄装置的拍摄姿态以使所述目标对象处于所述拍摄装置的拍摄画面中心。
如图3所示，当用户在参考图像30中框选出目标对象31时，目标对象31可能不在拍摄装置的拍摄画面中心，在本实施例中，当无人机获取到目标对象31的指示信息，例如接收到控制终端24发送的区域34在该参考图像30中的位置信息后，根据区域34在该参考图像30中的位置信息可确定出目标对象31相对于拍摄装置21的光轴的角度，根据该角度可调整无人机的姿态和/或云台的姿态来控制拍摄装置的拍摄姿态，以使目标对象31相对于拍摄装置的光轴的角度为0，即目标对象31处于所述拍摄装置的拍摄画面中心。
在一些实施例中，当用户框选目标对象31后，无人机即可对参考对象进行环绕飞行；因此，当无人机获取到目标对象31的指示信息后即可调整无人机的姿态和/或云台的姿态，以使目标对象31处于所述拍摄装置的拍摄画面中心，也就是说，在无人机对参考对象进行环绕飞行的过程中调整无人机的姿态和/或云台的姿态，以使目标对象31处于所述拍摄装置的拍摄画面中心，直到该无人机确定出目标对象31的三维坐标。
在另外一些实施例中，当用户框选目标对象31后，无人机并不立即对 参考对象进行环绕飞行，而是等到用户在该交互界面中点击该启动控制按键35后，该无人机才开始对参考对象进行环绕飞行。例如，无人机在t1时刻获取到该目标对象的指示信息，用户在t1时刻之后的t2时刻点击了启动控制按键35即无人机从t2时刻开始对参考对象进行环绕飞行，无人机在t2时刻之后的t3时刻确定出了目标对象31的三维坐标。
具体的，无人机可以在t1时刻到t2时刻之间调整无人机的姿态和/或云台的姿态，以使目标对象31处于所述拍摄装置的拍摄画面中心，因为在t1时刻到t2时刻之间无人机可能并没有动，但是目标对象31发生了移动，导致目标对象31在所述拍摄装置的拍摄画面中的位置发生了变化。或者，无人机还可以在t2时刻到t3时刻之间调整无人机的姿态和/或云台的姿态，以使目标对象31处于所述拍摄装置的拍摄画面中心。再或者，无人机还可以在t1时刻到t3时刻之间调整无人机的姿态和/或云台的姿态，以使目标对象31处于所述拍摄装置的拍摄画面中心。
另外，所述方法还包括：在获取到所述指示信息之后，获取所述拍摄装置输出的多帧第二目标图像，其中，所述第二目标图像中包括目标对象。
例如，当用户框选目标对象31后即无人机获取到目标对象31的指示信息后，无人机即可对参考对象进行环绕飞行，在无人机对象进行环绕飞行时获取拍摄装置输出的多帧第二目标图像，则此时的所述多帧第二目标图像包括所述多帧第一目标图像。
再例如，当用户框选目标对象31后，无人机并不立即对参考对象进行环绕飞行，而是等到用户在该交互界面中点击该启动控制按键35后，该无人机才开始对参考对象进行环绕飞行，则无人机获取到目标对象31的指示信息后，拍摄装置输出的多帧第二目标图像可能是拍摄装置在t1时刻到t2时刻之间拍摄的，也有可能是在t2时刻到t3时刻之间拍摄的，还有可能是在t1时刻到t3时刻之间拍摄的。也就是说，多帧第二目标图像至少包括多帧第一目标图像。
相应的，所述根据所述指示信息控制拍摄装置的拍摄姿态，包括如下步骤：
步骤S1101、基于所述参考图像的目标区域中的特征点利用跟踪算法获取每一帧第二目标图像的特征点。
具体的，利用跟踪算法计算该目标区域中的每个特征点在相邻目标图像 例如第二目标图像之间的偏移量，如果该特征点在前一帧目标图像相对于后一帧目标图像的偏移量和该特征点在后一帧目标图像相对于前一帧目标图像的偏移量大小相等、方向相反，即可确定该特征点是跟踪正确的特征点。
如图12所示，A、B、C、D、E、F、G分别表示参考图像30的目标区域即区域34中的特征点，特征点A、B、C、D、E、F、G也是目标对象31的特征点。121表示无人机在获取到所述指示信息之后，拍摄装置输出的多帧第二目标图像中的一个第二目标图像，此处只是示意性说明。根据KLT特征跟踪算法可确定出参考图像30中目标对象31的特征点例如A、B、C、D、E、F、G分别在第二目标图像121中的位置。
步骤S1102、根据每一帧第二目标图像的特征点确定目标对象在对应第二目标图像的位置信息。
根据特征点例如A、B、C、D、E、F、G分别在第二目标图像121中的位置，可确定出目标对象31在第二目标图像121中的位置信息，例如目标对象31的中心点N1在第二目标图像121中的位置信息。
步骤S1103、根据所述目标对象在对应第二目标图像的位置信息控制拍摄装置的拍摄姿态。
根据目标对象31的中心点N1在第二目标图像121中的位置信息以及第二目标图像121的中心点N的位置信息，可确定出目标对象31的中心点N1相对于第二目标图像121的中心点N在水平方向上的距离Δμ，以及目标对象31的中心点N1相对于第二目标图像121的中心点N在垂直方向上的距离Δυ，进一步，根据Δμ和拍摄装置在水平方向上的FOV，可确定出目标对象31相对于拍摄装置的光轴在水平方向上偏移的角度，根据Δυ和拍摄装置在垂直方向上的FOV，可确定出目标对象31相对于拍摄装置的光轴在垂直方向上偏移的角度。根据目标对象31相对于拍摄装置的光轴在水平方向和垂直方向上偏移的角度，通过调整无人机和/云台的姿态来调整该拍摄装置的拍摄姿态，以使拍摄装置的光轴对准目标对象31，目标对象31位于第二目标图像121的画面中心。
在其他实施例中，可以不限于将目标对象31调整在第一目标图像或第二目标图像的画面中心，还可以将目标对象31调整在第一目标图像或第二目标图像中的预设区域，也就是说，通过调整无人机和/云台的姿态，使得目标对象31相对于拍摄装置的光轴在水平方向和垂直方向上偏移的角度均为 非0的预设角度。
本实施例通过控制拍摄装置的拍摄姿态以使目标对象处于所述拍摄装置的拍摄画面中心，可避免无人机在对参考对象进行环绕飞行时该目标对象移动到该拍摄装置的拍摄画面外而导致无法正常确定目标对象的三维坐标；另外，也可以防止目标对象在移动过程中从拍摄装置的拍摄画面中消失。
在本申请一些实施例中，圆弧或圆形的轨迹，例如第一轨迹和/或第二轨迹可以由贝塞尔曲线拟合生成。可选地，贝塞尔曲线可以为三阶贝塞尔曲线。例如，第一轨迹所在圆的M分之一是由三阶贝塞尔曲线拟合生成的，第一轨迹所在圆的M分之一上包括四个控制点；和/或，第二轨迹所在圆的M分之一是由三阶贝塞尔曲线拟合生成的，第二轨迹所在圆的M分之一上包括四个控制点。
首先简单地说明三阶贝塞尔曲线拟合生成圆形轨迹的原理。
三阶贝塞尔曲线的参数方程为：
P(t)＝A(1-t) 3+B·3(1-t) 2·t+C·3(1-t)·t 2+D·t 3，
其中，t为曲线参数，t∈[0,1]；A、B、C、D为由M段三阶贝塞尔曲线拟合圆形时的四个控制点。
图13是本申请一个实施例的三阶贝塞尔曲线拟合生成圆形轨迹的原理的示意图。图13所示的示意图中M＝4，即由4段三阶贝塞尔曲线拟合完整的圆形，其圆心为O。如图13所示，P(0)和P(1)分别经过控制点A和控制点D，即曲线的两个端点。为了便于说明，假设控制点A在x轴上，坐标为(1,0)，控制点D在y轴上，坐标为(0,1)。控制点B的坐标为(h,1)和控制点C的坐标为(1,h)。选择合适位置的控制点，例如，选择h＝(4/3)tan(π/2M)，便能生成平滑的M分之一圆。其中，M＝4时，h＝(4/3)tan(π/8)≈0.552284749831。
应理解，M分之一圆的起始位置即控制点A可以为无人机当前所处的位置。M分之一圆的圆心可以为当前测算的兴趣点的第一位置。无论是圆弧轨迹、圆形轨迹还是，只需要指定圆心和曲线的起始位置两个点，便能使用四段1/4圆，拼接成一个完整的圆。
下面以第一轨迹为例，详细说明由三阶贝塞尔曲线拟合生成轨迹的过程。
取M＝4，将第一圆心标记为O 1，四个控制点包括第一控制点A 1、第二控制点B 1、第三控制点C 1和第四控制点D 1。第一控制点A 1为无人机当前 所处的位置，第一控制点A 1和第四控制点D 1为第一轨迹所在圆的四分之一的两个端点。第二控制点B 1在经第一控制点A 1的第一轨迹所在圆的切线上，并且第二控制点B 1和第四控制点D 1位于O 1A 1的同一侧，第二控制点B 1和第一控制点A 1之间的距离为h。第三控制点C 1在经第四控制点D 1的第一轨迹所在圆的切线上，并且第三控制点C 1和第一控制点A 1位于O 1D 1的同一侧，第三控制点C 1和第四控制点D 1之间的距离为h。其中
Figure PCTCN2018097944-appb-000003
实际中，第一控制点A 1可能不会刚好在x轴上，第一圆心O 1可能不会刚好在坐标(0,0)处。可以根据第一圆心O 1的位置(即兴趣点的二维位置或三维位置)以及第一控制点A 1(即无人机当前所处的位置)来推算点第二控制点B 1、第三控制点C 1和第四控制点D 1。图14是本申请一个实施例的第一圆心与四个控制点的示意图。本申请实施例只关心二维平面，所以忽略高度(即z轴)方向。
第一控制点A 1、第二控制点B 1、第三控制点C 1和第四控制点D 1满足以下关系：
Figure PCTCN2018097944-appb-000004
其中θ DA为向量O 1D 1与向量O 1A 1的夹角，假设顺时针取正，逆时针取负，则M＝4时
Figure PCTCN2018097944-appb-000005
B 1＝h(D 1-O 1)+A 1，C 1＝h(A 1-O 1)+D 1，其中
Figure PCTCN2018097944-appb-000006
假设O 1的坐标为(x 0,y 0)，A 1的坐标为(x A,y B)，则
Figure PCTCN2018097944-appb-000007
Figure PCTCN2018097944-appb-000008
Figure PCTCN2018097944-appb-000009
得到第一控制点A 1、第二控制点B 1、第三控制点C 1和第四控制点D 1的坐标，将其代入三阶贝塞尔曲线的参数方程
P(t)＝A(1-t) 3+B·3(1-t) 2·t+C·3(1-t)·t 2+D·t 3，t∈[0,1]
可以得到第一轨迹，展开并整理后记为
T 0(t)＝a 0+a 1·t+a 2·t 2+a 3·t 3。
应理解，本申请实施例还可以通过其它方法，例如最小二乘法拟合得到第一轨迹。已知第一轨迹的圆弧的圆心、半径以及其经过的点，可以通过各种方式得到第一轨迹的表达式，本申请实施例不再一一列举。
下面详细说明求得第二轨迹的过程。
在一个实施例中，由于能够得到第一轨迹与第二轨迹相切的切点，将该切点作为第五控制点A 2。另外，第二圆心标记为O 2，位于第二位置，即相对第一位置更新后的兴趣点的位置。得到这两个点的坐标以及第二半径后，就可以依照类似上述第一轨迹的计算过程，得到第二轨迹。
即M＝4，第二圆心标记为O 2，四个控制点包括第五控制点A 2、第六控制点B 2、第七控制点C 2和第八控制点D 2，第五控制点A 2为无人机当前所处的位置，第五控制点A 2和第八控制点D 2为第二轨迹所在圆的四分之一的两个端点，第六控制点B 2在经第五控制点A 2的第二轨迹所在圆的切线上，并且第六控制点B 2和第八控制点D 2位于O 2A 2的同一侧，第六控制点B 2和第五控制点A 2之间的距离为h，第七控制点C 2在经第八控制点D 2的第二轨迹所在圆的切线上，并且第七控制点C 2和第五控制点A 2位于O 2D 2的同一侧，第七控制点C 2和第八控制点D 2之间的距离为h，其中
Figure PCTCN2018097944-appb-000010
具体过程不再赘述。
在另一个实施例中，第二轨迹可以根据第一圆心、第一半径、第二圆心、第二半径和第一轨迹确定。假设第一圆心为O 1、第一半径为R 1、第二圆心为O 2、第二半径为R 2。图15是本申请一个实施例的第一轨迹和第二轨迹的示意图。
第二轨迹可以对第一轨迹进行圆心平移变换和半径尺度变换得到。具体而言，第一步对轨迹进行圆心平移变换，第二圆心为O 2至第一圆心为O 1的变化值为：Δt＝O 2-O 1，则可以得到轨迹的中间结果为S(t)＝T 0(t)+Δt。第二步对轨迹进行半径尺度变换，变换比例为
Figure PCTCN2018097944-appb-000011
则可以得到第二轨迹T 1(t)＝scale·(S(t)-S(0))+S(0)，对其进行整理后，可以为T 1(t)＝b 0+b 1·t+b 2·t 2+b 3·t 3。
应理解，已知第二轨迹的圆弧的圆心、半径以及其经过的点，便可以通 过各种方式得到第二轨迹的表达式，本申请实施例不再一一列举。由于第二轨迹和第一始轨迹是相切的，因此两轨迹在位置上是平滑的。
下面详细说明得到第二轨迹后，无人机进行轨迹切换的过程。
应理解，本申请实施例的第一轨迹、第二轨迹等仅是无人机规划的理想的运行轨迹，在实际运行中，无人机的实际位置通常与理想的运行轨迹有偏差。现有一种路径跟随技术，主要对寻找最近点进行一定的约束，使用上次的位置设定初值，迭代找到最近点，进行跟随控制。虽然现在采用路径跟随技术来纠正运行中的偏差，这种偏差仍然难以避免。另外，由于无人机通常具有较高的飞行速度，在真正的切换时刻，无人机会冲出运行轨迹一段距离。这就使得无人机在切换时通常不是平滑切换。
在本申请的一些实施例中，S140无人机由所述第一轨迹切换至沿第二轨迹运行，可以包括：所述无人机在所述第二轨迹上确定距离所述无人机当前所处的位置最近的点，作为目标点；所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。基于现有的方法，在第二轨迹上确定距离无人机当前所处的位置最近的点可以通过求第二轨迹的二阶导数来实现，此处不再赘述。
在本申请的另一些实施例中，S140无人机由所述第一轨迹切换至沿第二轨迹运行，可以包括：所述无人机将所述第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点；所述无人机从所述N+1个航点中确定距离所述无人机当前所处的位置最近的航点；所述无人机确定所述无人机当前所处的位置到经所述最近的航点做的所述第二轨迹的切线的垂足为目标点；所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
在一种具体的实现方式中，可以将第二轨迹上的至少部分轨迹离散化成多个航点。图16是本申请一个实施例的确定目标点的原理的示意图。例如图16所示的，N可以取值为20，那么该部分轨迹可以被20等分。在图16所示的例子中，该部分轨迹以参数t∈[0,1]来表示，但本申请实施例不限于此。除t＝0和t＝1始末两个点以外，在t＝1/20,2/20,3/20,…,19/20处分别取点，共21个航点。将t的取值分别代入第二轨迹的表达式T 1(t)＝b 0+b 1·t+b 2·t 2+b 3·t 3，则可以得到该21个航点的坐标点。应理解，这里将N取值为20仅是示例而非限定。
遍历该21个航点，找到与无人机当前所处的位置距离最接近的航点， 例如图16所示的为第六个航点(t＝5/20对应的航点)。经t＝5/20对应的航点做切线，并求取无人机当前所处的位置与切线之间的垂足，作为目标点。
具体求解过程可以根据以下步骤。
第一，求第二轨迹的一阶导数的表达式：
Figure PCTCN2018097944-appb-000012
第二，将与无人机当前所处的位置距离最接近的航点t＝5/20＝t6代入一阶导数的表达式，得到切线的斜率：
Figure PCTCN2018097944-appb-000013
将与无人机当前所处的位置距离最接近的航点t＝5/20＝t6代入第二轨迹的表达式T 1(t)，求得P 1(x 1,y 1)。
根据直线方程截距式y＝kx+c，求得c 1＝y 1-k 1x 1
所以切线方程为：y＝k 1x+c 1
将无人机当前所处的位置记为P 2(x 2,y 2)，无人机当前所处的位置与垂足的连线，必然垂直于切线，所以有：
k 2＝-1/k 1
c 2＝y 2-k 2x 2
所以垂直于切线的直线方程为：
y＝k 2x+c 2
联立方程求解
Figure PCTCN2018097944-appb-000014
即可以得到垂足P 3(x 3,y 3)，将其作为目标点。经由该目标点，将无人机由第一轨迹切换至沿第二轨迹运行。
应理解，在本申请的另一些实施例中也可以直接以与无人机当前所处的位置距离最接近的航点作为目标点。将垂足作为目标点相较于直接选用最近的航点作为目标点的情况，不会出现无人机后退导致的不平滑的问题。
在本申请的又一些实施例中，S140无人机由所述第一轨迹切换至沿第二轨迹运行，可以包括：所述无人机将所述第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点；所述无人机从所述N+1个航点中确定距离所述无人机当前所处的位置最近的，并且位于所述无人机当前所处的位置的运动 的前方的航点，作为目标点；所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。将位于无人机运动的前方且距离无人机当前所处的位置最近的的航点，作为目标点也可以避免在轨迹切换时出现无人机后退的问题。
在本申请一些实施例中，所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行，可以包括：通过比例-积分-微分(proportion-integral-derivative，PID)控制，所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。其中，PID控制是用来纠正系统响应，执行调节控制的一种技术。
应理解，本申请实施例的无人机由当前所处的位置切换至预定轨迹的过程，可以适用于更广泛的场景，而不局限于切换至的轨迹为圆弧轨迹。因此，本申请实施例还提供了一种轨迹切换的方法。图17是本申请另一实施例的轨迹切换的方法1700的示意性流程图。该方法1700可以包括以下步骤。
S1710，无人机将第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点，该第二轨迹为无人机待切换至的轨迹。
S1720，该无人机从该N+1个航点中确定距离该无人机当前所处的位置最近的航点。
S1730，该无人机确定该无人机当前所处的位置到经该最近的航点做的该第二轨迹的切线的垂足为目标点。
S1740，该无人机经由该目标点，切换至沿该第二轨迹运行。
本申请实施例的轨迹切换的方法，将待切换至的轨迹离散化出多个航点，确定距离该无人机当前所处的位置最近的航点，并经该最近的航点做的轨迹切线，将垂足作为目标点，经由该目标点，将无人机切换至沿该轨迹运行，可以使得无人机平滑地进行轨迹切换，可以提高用户体验。
可选地，在本申请的一些实施例中，该无人机经由该目标点，切换至沿该第二轨迹运行，可以包括：通过PID控制，该无人机经由该目标点，由该第一轨迹切换至沿该第二轨迹运行。
可选地，在本申请的一些实施例中，该第二轨迹可以为圆弧轨迹，该第二轨迹所在圆的圆心为第二圆心，半径为第二半径，该第二圆心位于第二位置。
可选地，在本申请的一些实施例中，在该无人机经由该目标点，切换至 沿该第二轨迹运行之前，该方法1700还可以包括：该无人机沿第一轨迹运行，该第一轨迹为圆弧轨迹，该第一轨迹所在圆的圆心为第一圆心，半径为第一半径，该第一圆心位于第一位置，该第一轨迹与该第二轨迹相切，该第二位置是相对该第一位置更新后的兴趣点的位置。
可选地，在本申请的一些实施例中，该第二位置可以为通过该无人机上设置的双目相机或单目相机根据视觉测量算法计算得到的。
可选地，在本申请的一些实施例中，该第二轨迹可以是由贝塞尔曲线拟合生成的。
可选地，在本申请的一些实施例中，该贝塞尔曲线可以为三阶贝塞尔曲线。
可选地，在本申请的一些实施例中，该第二轨迹所在圆的M分之一可以是由该三阶贝塞尔曲线拟合生成的，该第二轨迹所在圆的M分之一上包括四个控制点。
可选地，在本申请的一些实施例中，M＝4，该第二圆心标记为O 2，该四个控制点包括第一控制点A 2、第二控制点B 2、第三控制点C 2和第四控制点D 2，该第五控制点A 2为该无人机当前所处的位置，该第五控制点A 2和该第八控制点D 2为该第二轨迹所在圆的四分之一的两个端点，该第六控制点B 2在经该第五控制点A 2的该第二轨迹所在圆的切线上，并且该第六控制点B 2和该第八控制点D 2位于O 2A 2的同一侧，该第六控制点B 2和该第五控制点A 2之间的距离为h，该第七控制点C 2在经该第八控制点D 2的该第二轨迹所在圆的切线上，并且该第七控制点C 2和该第五控制点A 2位于O 2D 2的同一侧，该第七控制点C 2和该第八控制点D 2之间的距离为h，其中
Figure PCTCN2018097944-appb-000015
可选地，在本申请的一些实施例中，该第二轨迹可以是根据该第一圆心、该第一半径、该第二圆心、该第二半径和该第一轨迹确定的。
可选地，在本申请的一些实施例中，该第二轨迹是对该第一轨迹进行圆心平移变换和半径尺度变换得到的。
上文详细说明了本申请实施例的轨迹切换的方法，下面详细说明本申请实施例的轨迹切换的装置。
图18是本申请一个实施例的轨迹切换的装置1800的示意性框图。如图18所示，轨迹切换的装置1800包括：处理器1810和存储器1820，所述存储器1810用于存储计算机可执行指令，所述处理器1820用于执行所述计算机可执行指令，以实施以下操作：使无人机沿第一轨迹运行，所述第一轨迹为圆弧轨迹，所述第一轨迹所在圆的圆心为第一圆心，半径为第一半径，所述第一圆心位于第一位置；确定第二位置，所述第二位置是相对所述第一位置更新后的兴趣点的位置；根据所述第二位置确定第二轨迹，所述第二轨迹也为圆弧轨迹，所述第一轨迹与所述第二轨迹相切，所述第二轨迹所在圆的圆心为第二圆心，半径为第二半径，所述第二圆心位于第二位置；使所述无人机由所述第一轨迹切换至沿第二轨迹运行。
可选地，在一些实施例中，所述第一位置和/或所述第二位置为通过所述无人机上设置的双目相机或单目相机根据视觉测量算法计算得到的。
可选地，在一些实施例中，所述第一轨迹和/或所述第二轨迹是由贝塞尔曲线拟合生成的。
可选地，在一些实施例中，所述贝塞尔曲线为三阶贝塞尔曲线。
可选地，在一些实施例中，所述第一轨迹所在圆和/或所述第二轨迹所在圆的M分之一是由所述三阶贝塞尔曲线拟合生成的，所述第一轨迹所在圆和/或所述第二轨迹所在圆的M分之一上包括四个控制点。
可选地，在一些实施例中，M＝4，所述第一圆心标记为O 1，所述四个控制点包括第一控制点A 1、第二控制点B 1、第三控制点C 1和第四控制点D 1，所述第一控制点A 1为所述无人机当前所处的位置，所述第一控制点A 1和所述第四控制点D 1为所述第一轨迹所在圆的四分之一的两个端点，所述第二控制点B 1在经所述第一控制点A 1的所述第一轨迹所在圆的切线上，并且所述第二控制点B 1和所述第四控制点D 1位于O 1A 1的同一侧，所述第二控制点B 1和所述第一控制点A 1之间的距离为h，所述第三控制点C 1在经所述第四控制点D 1的所述第一轨迹所在圆的切线上，并且所述第三控制点C 1和所述第一控制点A 1位于O 1D 1的同一侧，所述第三控制点C 1和所述第四控制点D 1之间的距离为h，
其中
Figure PCTCN2018097944-appb-000016
可选地，在一些实施例中，M＝4，所述第二圆心标记为O 2，所述四个控 制点包括第五控制点A 2、第六控制点B 2、第七控制点C 2和第八控制点D 2，所述第五控制点A 2为所述无人机当前所处的位置，所述第五控制点A 2和所述第八控制点D 2为所述第二轨迹所在圆的四分之一的两个端点，所述第六控制点B 2在经所述第五控制点A 2的所述第二轨迹所在圆的切线上，并且所述第六控制点B 2和所述第八控制点D 2位于O 2A 2的同一侧，所述第六控制点B 2和所述第五控制点A 2之间的距离为h，所述第七控制点C 2在经所述第八控制点D 2的所述第二轨迹所在圆的切线上，并且所述第七控制点C 2和所述第五控制点A 2位于O 2D 2的同一侧，所述第七控制点C 2和所述第八控制点D 2之间的距离为h，
其中
Figure PCTCN2018097944-appb-000017
可选地，在一些实施例中，所述处理器1810具体用于：根据所述第一圆心、所述第一半径、所述第二圆心、所述第二半径和所述第一轨迹确定所述第二轨迹。
可选地，在一些实施例中，所述处理器1810具体用于：对所述第一轨迹进行圆心平移变换和半径尺度变换得到所述第二轨迹。
可选地，在一些实施例中，所述处理器具1810体用于：在所述第二轨迹上确定距离所述无人机当前所处的位置最近的点，作为目标点；使所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
可选地，在一些实施例中，所述处理器1810具体用于：将所述第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点；从所述N+1个航点中确定距离所述无人机当前所处的位置最近的航点；确定所述无人机当前所处的位置到经所述最近的航点做的所述第二轨迹的切线的垂足为目标点；使所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
可选地，在一些实施例中，所述处理器1810具体用于：将所述第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点；从所述N+1个航点中确定距离所述无人机当前所处的位置最近的，并且位于所述无人机当前所处的位置的运动的前方的航点，作为目标点；使所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
可选地，在一些实施例中，所述处理器1810具体用于：通过比例-积分-微分PID控制，使所述无人机经由所述目标点，由所述第一轨迹切换至沿 所述第二轨迹运行。
应理解，本申请各实施例的轨迹切换的装置可以基于模块实现。例如，轨迹切换的装置1800可以包括控制模块，用于使无人机沿第一轨迹运行，以及使所述无人机由所述第一轨迹切换至沿第二轨迹运行等。轨迹切换的装置1800可以包括计算模块，用于确定第二位置，以及根据所述第二位置确定第二轨迹，所述第二轨迹也为圆弧轨迹等。轨迹切换的装置中的各模块可以用于执行本申请相应实施例的方法，此处不再赘述。
本申请实施例还提供了一种无人机，包括：上述轨迹切换的装置1800。
图19是本申请一个实施例的轨迹切换的装置1900的示意性框图。如图19所示，轨迹切换的装置1900包括：处理器1910和存储器1920，所述存储器1920用于存储计算机可执行指令，所述处理器1910用于执行所述计算机可执行指令，以实施以下操作：将第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点，所述第二轨迹为无人机待切换至的轨迹；从所述N+1个航点中确定距离无人机当前所处的位置最近的航点；确定所述无人机当前所处的位置到经所述最近的航点做的所述第二轨迹的切线的垂足为目标点；使所述无人机经由所述目标点，切换至沿所述第二轨迹运行。
可选地，在一些实施例中，所述处理器1910具体用于：通过比例-积分-微分PID控制，使所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
可选地，在一些实施例中，所述第二轨迹为圆弧轨迹，所述第二轨迹所在圆的圆心为第二圆心，半径为第二半径，所述第二圆心位于第二位置。
可选地，在一些实施例中，在使所述无人机经由所述目标点，切换至沿所述第二轨迹运行之前，所述处理器1910还用于：使所述无人机沿第一轨迹运行，所述第一轨迹为圆弧轨迹，所述第一轨迹所在圆的圆心为第一圆心，半径为第一半径，所述第一圆心位于第一位置，所述第一轨迹与所述第二轨迹相切，所述第二位置是相对所述第一位置更新后的兴趣点的位置。
可选地，在一些实施例中，所述第二位置为通过所述无人机上设置的双目相机或单目相机根据视觉测量算法计算得到的。
可选地，在一些实施例中，所述第二轨迹是由贝塞尔曲线拟合生成的。
可选地，在一些实施例中，所述贝塞尔曲线为三阶贝塞尔曲线。
可选地，在一些实施例中，所述第二轨迹所在圆的M分之一是由所述 三阶贝塞尔曲线拟合生成的，所述第二轨迹所在圆的M分之一上包括四个控制点。
可选地，在一些实施例中，M＝4，所述第二圆心标记为O 2，所述四个控制点包括第一控制点A 2、第二控制点B 2、第三控制点C 2和第四控制点D 2，所述第五控制点A 2为所述无人机当前所处的位置，所述第五控制点A 2和所述第八控制点D 2为所述第二轨迹所在圆的四分之一的两个端点，所述第六控制点B 2在经所述第五控制点A 2的所述第二轨迹所在圆的切线上，并且所述第六控制点B 2和所述第八控制点D 2位于O 2A 2的同一侧，所述第六控制点B 2和所述第五控制点A 2之间的距离为h，所述第七控制点C 2在经所述第八控制点D 2的所述第二轨迹所在圆的切线上，并且所述第七控制点C 2和所述第五控制点A 2位于O 2D 2的同一侧，所述第七控制点C 2和所述第八控制点D 2之间的距离为h，
其中
Figure PCTCN2018097944-appb-000018
可选地，在一些实施例中，所述处理器1910具体用于：根据所述第一圆心、所述第一半径、所述第二圆心、所述第二半径和所述第一轨迹确定所述第二轨迹。
可选地，在一些实施例中，所述处理器1910具体用于：对所述第一轨迹进行圆心平移变换和半径尺度变换得到所述第二轨迹。
应理解，本申请各实施例的轨迹切换的装置可以基于模块实现。例如，轨迹切换的装置1900可以包括计算模块，用于将第二轨迹上的至少部分轨迹进行N等分得到N+1个航点，从所述N+1个航点中确定距离无人机当前所处的位置最近的航点，以及确定所述无人机当前所处的位置到经所述最近的航点做的所述第二轨迹的切线的垂足为目标点等。轨迹切换的装置1900可以包括控制模块，用于使所述无人机经由所述目标点，切换至沿所述第二轨迹运行等。轨迹切换的装置中的各模块可以用于执行本申请相应实施例的方法，此处不再赘述。
本申请实施例还提供了一种无人机，包括：上述轨迹切换的装置1900。
应理解，本申请实施例中提及的处理器可以是中央处理单元(central processing unit，CPU)，还可以是其他通用处理器、数字信号处理器(digital signal processor，DSP)、专用集成电路(application specific integrated circuit， ASIC)、现成可编程门阵列(field programmable gate array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。
还应理解，本申请实施例中提及的存储器可以是易失性存储器或非易失性存储器，或可包括易失性和非易失性存储器两者。其中，非易失性存储器可以是只读存储器(read-only memory，ROM)、可编程只读存储器(programmable ROM，PROM)、可擦除可编程只读存储器(erasable PROM，EPROM)、电可擦除可编程只读存储器(electrically EPROM，EEPROM)或闪存。易失性存储器可以是随机存取存储器(random access memory，RAM)，其用作外部高速缓存。通过示例性但不是限制性说明，许多形式的RAM可用，例如静态随机存取存储器(static RAM，SRAM)、动态随机存取存储器(dynamic RAM，DRAM)、同步动态随机存取存储器(synchronous DRAM，SDRAM)、双倍数据速率同步动态随机存取存储器(double data rate SDRAM，DDR SDRAM)、增强型同步动态随机存取存储器(enhanced SDRAM，ESDRAM)、同步连接动态随机存取存储器(synchlink DRAM，SLDRAM)和直接内存总线随机存取存储器(direct rambus RAM，DR RAM)。
需要说明的是，当处理器为通用处理器、DSP、ASIC、FPGA或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件时，存储器(存储模块)集成在处理器中。
应注意，本文描述的存储器旨在包括但不限于这些和任意其它适合类型的存储器。
本申请实施例还提供一种计算机可读存储介质，其上存储有指令，当指令在计算机上运行时，使得计算机执行上述各方法实施例的方法。
本申请实施例还提供一种计算机程序，该计算机程序使得计算机执行上述各方法实施例的方法。
本申请实施例还提供一种计算设备，该计算设备包括上述计算机可读存储介质。
应理解，本申请各实施例的电路、子电路、子单元的划分只是示意性的。本领域普通技术人员可以意识到，本文中所公开的实施例描述的各示例的电路、子电路和子单元，能够再行拆分或组合。
在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其任意 组合来实现。当使用软件实现时，可以全部或部分地以计算机程序产品的形式实现。计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行计算机指令时，全部或部分地产生按照本申请实施例的流程或功能。计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。计算机指令可以存储在计算机可读存储介质中，或者从一个计算机可读存储介质向另一个计算机可读存储介质传输，例如，计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(Digital Subscriber Line，DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集成的服务器、数据中心等数据存储设备。可用介质可以是磁性介质(例如，软盘、硬盘、磁带)、光介质(例如，高密度数字视频光盘(Digital Video Disc，DVD))、或者半导体介质(例如，固态硬盘(Solid State Disk，SSD))等。
应理解，说明书通篇中提到的“一个实施例”或“一实施例”意味着与实施例有关的特定特征、结构或特性包括在本申请的至少一个实施例中。因此，在整个说明书各处出现的“在一个实施例中”或“在一实施例中”未必一定指相同的实施例。此外，这些特定的特征、结构或特性可以任意适合的方式结合在一个或多个实施例中。
应理解，在本申请的各种实施例中，上述各过程的序号的大小并不意味着执行顺序的先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本申请实施例的实施过程构成任何限定。
应理解，在本申请实施例中，“与A相应的B”表示B与A相关联，根据A可以确定B。但还应理解，根据A确定B并不意味着仅仅根据A确定B，还可以根据A和/或其它信息确定B。
应理解，本文中术语“和/或”，仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如，A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B这三种情况。另外，本文中字符“/”，一般表示前后关联对象是一种“或”的关系。
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、或者计算机软件和电子硬件的结 合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本申请的范围。
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的系统、装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
在本申请所提供的几个实施例中，应该理解到，所揭露的系统、装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。
以上所述，仅为本申请的具体实施方式，但本申请的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本申请的保护范围之内。因此，本申请的保护范围应以所述权利要求的保护范围为准。

Claims (52)

  1. 一种轨迹切换的方法，其特征在于，包括：
    无人机沿第一轨迹运行，所述第一轨迹为圆弧轨迹，所述第一轨迹所在圆的圆心为第一圆心，半径为第一半径，所述第一圆心位于第一位置；
    所述无人机确定第二位置，所述第二位置是相对所述第一位置更新后的兴趣点的位置；
    所述无人机根据所述第二位置确定第二轨迹，所述第二轨迹也为圆弧轨迹，所述第一轨迹与所述第二轨迹相切，所述第二轨迹所在圆的圆心为第二圆心，半径为第二半径，所述第二圆心位于第二位置；
    所述无人机由所述第一轨迹切换至沿第二轨迹运行。
  2. 根据权利要求1所述的方法，其特征在于，所述第一位置和/或所述第二位置为通过所述无人机上设置的双目相机或单目相机根据视觉测量算法计算得到的。
  3. 根据权利要求1或2所述的方法，其特征在于，所述第一轨迹和/或所述第二轨迹是由贝塞尔曲线拟合生成的。
  4. 根据权利要求3所述的方法，其特征在于，所述贝塞尔曲线为三阶贝塞尔曲线。
  5. 根据权利要求3或4所述的方法，其特征在于，所述第一轨迹所在圆和/或所述第二轨迹所在圆的M分之一是由所述三阶贝塞尔曲线拟合生成的，所述第一轨迹所在圆和/或所述第二轨迹所在圆的M分之一上包括四个控制点。
  6. 根据权利要求5所述的方法，其特征在于，M＝4，所述第一圆心标记为O 1，所述四个控制点包括第一控制点A 1、第二控制点B 1、第三控制点C 1和第四控制点D 1，
    所述第一控制点A 1为所述无人机当前所处的位置，所述第一控制点A 1和所述第四控制点D 1为所述第一轨迹所在圆的四分之一的两个端点，
    所述第二控制点B 1在经所述第一控制点A 1的所述第一轨迹所在圆的切线上，并且所述第二控制点B 1和所述第四控制点D 1位于O 1A 1的同一侧，所述第二控制点B 1和所述第一控制点A 1之间的距离为h，
    所述第三控制点C 1在经所述第四控制点D 1的所述第一轨迹所在圆的切线上，并且所述第三控制点C 1和所述第一控制点A 1位于O 1D 1的同一侧， 所述第三控制点C 1和所述第四控制点D 1之间的距离为h，
    其中
    Figure PCTCN2018097944-appb-100001
  7. 根据权利要求5所述的方法，其特征在于，M＝4，所述第二圆心标记为O 2，所述四个控制点包括第五控制点A 2、第六控制点B 2、第七控制点C 2和第八控制点D 2，
    所述第五控制点A 2为所述无人机当前所处的位置，所述第五控制点A 2和所述第八控制点D 2为所述第二轨迹所在圆的四分之一的两个端点，
    所述第六控制点B 2在经所述第五控制点A 2的所述第二轨迹所在圆的切线上，并且所述第六控制点B 2和所述第八控制点D 2位于O 2A 2的同一侧，所述第六控制点B 2和所述第五控制点A 2之间的距离为h，
    所述第七控制点C 2在经所述第八控制点D 2的所述第二轨迹所在圆的切线上，并且所述第七控制点C 2和所述第五控制点A 2位于O 2D 2的同一侧，所述第七控制点C 2和所述第八控制点D 2之间的距离为h，
    其中
    Figure PCTCN2018097944-appb-100002
  8. 根据权利要求1或2所述的方法，其特征在于，所述第二轨迹是根据所述第一圆心、所述第一半径、所述第二圆心、所述第二半径和所述第一轨迹确定的。
  9. 根据权利要求8所述的方法，其特征在于，所述第二轨迹是对所述第一轨迹进行圆心平移变换和半径尺度变换得到的。
  10. 根据权利要求1至9中任一项所述的方法，其特征在于，所述无人机由所述第一轨迹切换至沿第二轨迹运行，包括：
    所述无人机在所述第二轨迹上确定距离所述无人机当前所处的位置最近的点，作为目标点；
    所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
  11. 根据权利要求1至9中任一项所述的方法，其特征在于，所述无人机由所述第一轨迹切换至沿第二轨迹运行，包括：
    所述无人机将所述第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点；
    所述无人机从所述N+1个航点中确定距离所述无人机当前所处的位置 最近的航点；
    所述无人机确定所述无人机当前所处的位置到经所述最近的航点做的所述第二轨迹的切线的垂足为目标点；
    所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
  12. 根据权利要求1至9中任一项所述的方法，其特征在于，所述无人机由所述第一轨迹切换至沿第二轨迹运行，包括：
    所述无人机将所述第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点；
    所述无人机从所述N+1个航点中确定距离所述无人机当前所处的位置最近的，并且位于所述无人机当前所处的位置的运动的前方的航点，作为目标点；
    所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
  13. 根据权利要求10至12中任一项所述的方法，其特征在于，所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行，包括：
    通过比例-积分-微分PID控制，所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
  14. 一种轨迹切换的装置，其特征在于，包括：处理器和存储器，所述存储器用于存储计算机可执行指令，所述处理器用于执行所述计算机可执行指令，以实施以下操作：
    使无人机沿第一轨迹运行，所述第一轨迹为圆弧轨迹，所述第一轨迹所在圆的圆心为第一圆心，半径为第一半径，所述第一圆心位于第一位置；
    确定第二位置，所述第二位置是相对所述第一位置更新后的兴趣点的位置；
    根据所述第二位置确定第二轨迹，所述第二轨迹也为圆弧轨迹，所述第一轨迹与所述第二轨迹相切，所述第二轨迹所在圆的圆心为第二圆心，半径为第二半径，所述第二圆心位于第二位置；
    使所述无人机由所述第一轨迹切换至沿第二轨迹运行。
  15. 根据权利要求14所述的装置，其特征在于，所述第一位置和/或所述第二位置为通过所述无人机上设置的双目相机或单目相机根据视觉测量 算法计算得到的。
  16. 根据权利要求14或15所述的装置，其特征在于，所述第一轨迹和/或所述第二轨迹是由贝塞尔曲线拟合生成的。
  17. 根据权利要求16所述的装置，其特征在于，所述贝塞尔曲线为三阶贝塞尔曲线。
  18. 根据权利要求16或17所述的装置，其特征在于，所述第一轨迹所在圆和/或所述第二轨迹所在圆的M分之一是由所述三阶贝塞尔曲线拟合生成的，所述第一轨迹所在圆和/或所述第二轨迹所在圆的M分之一上包括四个控制点。
  19. 根据权利要求18所述的装置，其特征在于，M＝4，所述第一圆心标记为O 1，所述四个控制点包括第一控制点A 1、第二控制点B 1、第三控制点C 1和第四控制点D 1，
    所述第一控制点A 1为所述无人机当前所处的位置，所述第一控制点A 1和所述第四控制点D 1为所述第一轨迹所在圆的四分之一的两个端点，
    所述第二控制点B 1在经所述第一控制点A 1的所述第一轨迹所在圆的切线上，并且所述第二控制点B 1和所述第四控制点D 1位于O 1A 1的同一侧，所述第二控制点B 1和所述第一控制点A 1之间的距离为h，
    所述第三控制点C 1在经所述第四控制点D 1的所述第一轨迹所在圆的切线上，并且所述第三控制点C 1和所述第一控制点A 1位于O 1D 1的同一侧，所述第三控制点C 1和所述第四控制点D 1之间的距离为h，
    其中
    Figure PCTCN2018097944-appb-100003
  20. 根据权利要求18所述的装置，其特征在于，M＝4，所述第二圆心标记为O 2，所述四个控制点包括第五控制点A 2、第六控制点B 2、第七控制点C 2和第八控制点D 2，
    所述第五控制点A 2为所述无人机当前所处的位置，所述第五控制点A 2和所述第八控制点D 2为所述第二轨迹所在圆的四分之一的两个端点，
    所述第六控制点B 2在经所述第五控制点A 2的所述第二轨迹所在圆的切线上，并且所述第六控制点B 2和所述第八控制点D 2位于O 2A 2的同一侧，所述第六控制点B 2和所述第五控制点A 2之间的距离为h，
    所述第七控制点C 2在经所述第八控制点D 2的所述第二轨迹所在圆的切 线上，并且所述第七控制点C 2和所述第五控制点A 2位于O 2D 2的同一侧，所述第七控制点C 2和所述第八控制点D 2之间的距离为h，
    其中
    Figure PCTCN2018097944-appb-100004
  21. 根据权利要求14或15所述的装置，其特征在于，所述处理器具体用于：
    根据所述第一圆心、所述第一半径、所述第二圆心、所述第二半径和所述第一轨迹确定所述第二轨迹。
  22. 根据权利要求21所述的装置，其特征在于，所述处理器具体用于：
    对所述第一轨迹进行圆心平移变换和半径尺度变换得到所述第二轨迹。
  23. 根据权利要求14至22中任一项所述的装置，其特征在于，所述处理器具体用于：
    在所述第二轨迹上确定距离所述无人机当前所处的位置最近的点，作为目标点；
    使所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
  24. 根据权利要求14至22中任一项所述的装置，其特征在于，所述处理器具体用于：
    将所述第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点；
    从所述N+1个航点中确定距离所述无人机当前所处的位置最近的航点；
    确定所述无人机当前所处的位置到经所述最近的航点做的所述第二轨迹的切线的垂足为目标点；
    使所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
  25. 根据权利要求14至22中任一项所述的装置，其特征在于，所述处理器具体用于：
    将所述第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点；
    从所述N+1个航点中确定距离所述无人机当前所处的位置最近的，并且位于所述无人机当前所处的位置的运动的前方的航点，作为目标点；
    使所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
  26. 根据权利要求23至25中任一项所述的装置，其特征在于，所述处理器具体用于：
    通过比例-积分-微分PID控制，使所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
  27. 一种无人机，其特征在于，包括：权利要求14至26中任一项所述的轨迹切换的装置。
  28. 一种轨迹切换的方法，其特征在于，包括：
    无人机将第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点，所述第二轨迹为无人机待切换至的轨迹；
    所述无人机从所述N+1个航点中确定距离所述无人机当前所处的位置最近的航点；
    所述无人机确定所述无人机当前所处的位置到经所述最近的航点做的所述第二轨迹的切线的垂足为目标点；
    所述无人机经由所述目标点，切换至沿所述第二轨迹运行。
  29. 根据权利要求28所述的方法，其特征在于，所述无人机经由所述目标点，切换至沿所述第二轨迹运行，包括：
    通过比例-积分-微分PID控制，所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
  30. 根据权利要求28或29所述的方法，其特征在于，所述第二轨迹为圆弧轨迹，所述第二轨迹所在圆的圆心为第二圆心，半径为第二半径，所述第二圆心位于第二位置。
  31. 根据权利要求28至30中任一项所述的方法，其特征在于，在所述无人机经由所述目标点，切换至沿所述第二轨迹运行之前，所述方法还包括：
    所述无人机沿第一轨迹运行，所述第一轨迹为圆弧轨迹，所述第一轨迹所在圆的圆心为第一圆心，半径为第一半径，所述第一圆心位于第一位置，所述第一轨迹与所述第二轨迹相切，所述第二位置是相对所述第一位置更新后的兴趣点的位置。
  32. 根据权利要求28至31中任一项所述的方法，其特征在于，所述第二位置为通过所述无人机上设置的双目相机或单目相机根据视觉测量算法计算得到的。
  33. 根据权利要求28至32中任一项所述的方法，其特征在于，所述第 二轨迹是由贝塞尔曲线拟合生成的。
  34. 根据权利要求33所述的方法，其特征在于，所述贝塞尔曲线为三阶贝塞尔曲线。
  35. 根据权利要求30所述的方法，其特征在于，所述第二轨迹所在圆的M分之一是由所述三阶贝塞尔曲线拟合生成的，所述第二轨迹所在圆的M分之一上包括四个控制点。
  36. 根据权利要求35所述的方法，其特征在于，M＝4，所述第二圆心标记为O 2，所述四个控制点包括第一控制点A 2、第二控制点B 2、第三控制点C 2和第四控制点D 2，
    所述第五控制点A 2为所述无人机当前所处的位置，所述第五控制点A 2和所述第八控制点D 2为所述第二轨迹所在圆的四分之一的两个端点，
    所述第六控制点B 2在经所述第五控制点A 2的所述第二轨迹所在圆的切线上，并且所述第六控制点B 2和所述第八控制点D 2位于O 2A 2的同一侧，所述第六控制点B 2和所述第五控制点A 2之间的距离为h，
    所述第七控制点C 2在经所述第八控制点D 2的所述第二轨迹所在圆的切线上，并且所述第七控制点C 2和所述第五控制点A 2位于O 2D 2的同一侧，所述第七控制点C 2和所述第八控制点D 2之间的距离为h，
    其中
    Figure PCTCN2018097944-appb-100005
  37. 根据权利要求31所述的方法，其特征在于，所述第二轨迹是根据所述第一圆心、所述第一半径、所述第二圆心、所述第二半径和所述第一轨迹确定的。
  38. 根据权利要求37所述的方法，其特征在于，所述第二轨迹是对所述第一轨迹进行圆心平移变换和半径尺度变换得到的。
  39. 一种轨迹切换的装置，其特征在于，包括：处理器和存储器，所述存储器用于存储计算机可执行指令，所述处理器用于执行所述计算机可执行指令，以实施以下操作：
    将第二轨迹上的至少部分轨迹进行N等分，得到N+1个航点，所述第二轨迹为无人机待切换至的轨迹；
    从所述N+1个航点中确定距离无人机当前所处的位置最近的航点；
    确定所述无人机当前所处的位置到经所述最近的航点做的所述第二轨 迹的切线的垂足为目标点；
    使所述无人机经由所述目标点，切换至沿所述第二轨迹运行。
  40. 根据权利要求39所述的装置，其特征在于，所述处理器具体用于：
    通过比例-积分-微分PID控制，使所述无人机经由所述目标点，由所述第一轨迹切换至沿所述第二轨迹运行。
  41. 根据权利要求39或40所述的装置，其特征在于，所述第二轨迹为圆弧轨迹，所述第二轨迹所在圆的圆心为第二圆心，半径为第二半径，所述第二圆心位于第二位置。
  42. 根据权利要求39至41中任一项所述的装置，其特征在于，在使所述无人机经由所述目标点，切换至沿所述第二轨迹运行之前，所述处理器还用于：
    使所述无人机沿第一轨迹运行，所述第一轨迹为圆弧轨迹，所述第一轨迹所在圆的圆心为第一圆心，半径为第一半径，所述第一圆心位于第一位置，所述第一轨迹与所述第二轨迹相切，所述第二位置是相对所述第一位置更新后的兴趣点的位置。
  43. 根据权利要求39至42中任一项所述的装置，其特征在于，所述第二位置为通过所述无人机上设置的双目相机或单目相机根据视觉测量算法计算得到的。
  44. 根据权利要求39至43中任一项所述的装置，其特征在于，所述第二轨迹是由贝塞尔曲线拟合生成的。
  45. 根据权利要求44所述的装置，其特征在于，所述贝塞尔曲线为三阶贝塞尔曲线。
  46. 根据权利要求41所述的装置，其特征在于，所述第二轨迹所在圆的M分之一是由所述三阶贝塞尔曲线拟合生成的，所述第二轨迹所在圆的M分之一上包括四个控制点。
  47. 根据权利要求46所述的装置，其特征在于，M＝4，所述第二圆心标记为O 2，所述四个控制点包括第一控制点A 2、第二控制点B 2、第三控制点C 2和第四控制点D 2，
    所述第五控制点A 2为所述无人机当前所处的位置，所述第五控制点A 2和所述第八控制点D 2为所述第二轨迹所在圆的四分之一的两个端点，
    所述第六控制点B 2在经所述第五控制点A 2的所述第二轨迹所在圆的切 线上，并且所述第六控制点B 2和所述第八控制点D 2位于O 2A 2的同一侧，所述第六控制点B 2和所述第五控制点A 2之间的距离为h，
    所述第七控制点C 2在经所述第八控制点D 2的所述第二轨迹所在圆的切线上，并且所述第七控制点C 2和所述第五控制点A 2位于O 2D 2的同一侧，所述第七控制点C 2和所述第八控制点D 2之间的距离为h，
    其中
    Figure PCTCN2018097944-appb-100006
  48. 根据权利要求42所述的装置，其特征在于，所述处理器具体用于：根据所述第一圆心、所述第一半径、所述第二圆心、所述第二半径和所述第一轨迹确定所述第二轨迹。
  49. 根据权利要求48所述的装置，其特征在于，所述处理器具体用于：对所述第一轨迹进行圆心平移变换和半径尺度变换得到所述第二轨迹。
  50. 一种无人机，其特征在于，包括：权利要求39至49中任一项所述的轨迹切换的装置。
  51. 一种计算机可读存储介质，其特征在于，其上存储有计算机程序，所述计算机程序被执行以实现如权利要求1至13中任一项所述的方法。
  52. 一种计算机可读存储介质，其特征在于，其上存储有计算机程序，所述计算机程序被执行以实现如权利要求28至38中任一项所述的方法。
PCT/CN2018/097944 2018-08-01 2018-08-01 轨迹切换的方法和装置 WO2020024134A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2018/097944 WO2020024134A1 (zh) 2018-08-01 2018-08-01 轨迹切换的方法和装置
CN201880037381.7A CN110730934A (zh) 2018-08-01 2018-08-01 轨迹切换的方法和装置

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2018/097944 WO2020024134A1 (zh) 2018-08-01 2018-08-01 轨迹切换的方法和装置

Publications (1)

Publication Number Publication Date
WO2020024134A1 true WO2020024134A1 (zh) 2020-02-06

Family

ID=69217754

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/097944 WO2020024134A1 (zh) 2018-08-01 2018-08-01 轨迹切换的方法和装置

Country Status (2)

Country Link
CN (1) CN110730934A (zh)
WO (1) WO2020024134A1 (zh)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021212445A1 (zh) * 2020-04-24 2021-10-28 深圳市大疆创新科技有限公司 拍摄方法、可移动平台、控制设备和存储介质
CN111539345B (zh) * 2020-04-27 2023-09-26 阿波罗智能技术(北京)有限公司 用于确定变道动作的方法、装置、设备及可读存储介质
CN112731959A (zh) * 2020-12-01 2021-04-30 一飞智控(天津)科技有限公司 一种无人机编队圆弧引导切换目标动态规划方法
CN113655813B (zh) * 2021-10-20 2022-01-25 北京微纳星空科技有限公司 一种飞行纠偏控制方法、系统、存储介质和电子设备

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060015247A1 (en) * 2004-07-07 2006-01-19 The Boeing Company Bezier curve flightpath guidance using moving waypoints
CN105573315A (zh) * 2015-12-01 2016-05-11 珞石（北京）科技有限公司 用于工业机器人的笛卡尔空间轨迹的几何平滑方法
CN106325294A (zh) * 2016-08-22 2017-01-11 上海交通大学 基于贝塞尔曲线转接的无人机轨迹平滑方法
CN106603970A (zh) * 2016-11-11 2017-04-26 重庆零度智控智能科技有限公司 视频拍摄方法、系统及无人机
CN106657779A (zh) * 2016-12-13 2017-05-10 重庆零度智控智能科技有限公司 环绕拍摄方法、装置及无人机
CN107087427A (zh) * 2016-11-30 2017-08-22 深圳市大疆创新科技有限公司 飞行器的控制方法、装置和设备以及飞行器
CN107168305A (zh) * 2017-04-01 2017-09-15 西安交通大学 路口场景下基于Bezier和VFH的无人车轨迹规划方法
CN107807659A (zh) * 2017-10-24 2018-03-16 北京臻迪科技股份有限公司 一种无人机飞行控制方法及装置
CN107980108A (zh) * 2017-01-04 2018-05-01 深圳配天智能技术研究院有限公司 机器人运动轨迹规划方法及相关装置

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2280241A3 (en) * 2009-07-30 2017-08-23 QinetiQ Limited Vehicle control
CN102358287A (zh) * 2011-09-05 2012-02-22 北京航空航天大学 一种用于车辆自动驾驶机器人的轨迹跟踪控制方法
CN104492066B (zh) * 2014-12-18 2017-02-22 中国科学院自动化研究所 任务导向式主动训练控制方法
CN115113645A (zh) * 2016-07-04 2022-09-27 深圳市大疆创新科技有限公司 用于支持航空作业的方法
CN107272734A (zh) * 2017-06-13 2017-10-20 深圳市易成自动驾驶技术有限公司 无人机飞行任务执行方法、无人机及计算机可读存储介质

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060015247A1 (en) * 2004-07-07 2006-01-19 The Boeing Company Bezier curve flightpath guidance using moving waypoints
CN105573315A (zh) * 2015-12-01 2016-05-11 珞石（北京）科技有限公司 用于工业机器人的笛卡尔空间轨迹的几何平滑方法
CN106325294A (zh) * 2016-08-22 2017-01-11 上海交通大学 基于贝塞尔曲线转接的无人机轨迹平滑方法
CN106603970A (zh) * 2016-11-11 2017-04-26 重庆零度智控智能科技有限公司 视频拍摄方法、系统及无人机
CN107087427A (zh) * 2016-11-30 2017-08-22 深圳市大疆创新科技有限公司 飞行器的控制方法、装置和设备以及飞行器
CN106657779A (zh) * 2016-12-13 2017-05-10 重庆零度智控智能科技有限公司 环绕拍摄方法、装置及无人机
CN107980108A (zh) * 2017-01-04 2018-05-01 深圳配天智能技术研究院有限公司 机器人运动轨迹规划方法及相关装置
CN107168305A (zh) * 2017-04-01 2017-09-15 西安交通大学 路口场景下基于Bezier和VFH的无人车轨迹规划方法
CN107807659A (zh) * 2017-10-24 2018-03-16 北京臻迪科技股份有限公司 一种无人机飞行控制方法及装置

Also Published As

Publication number Publication date
CN110730934A (zh) 2020-01-24

Similar Documents

Publication Publication Date Title
CN112567201B (zh) 距离测量方法以及设备
US11347217B2 (en) User interaction paradigms for a flying digital assistant
US20210116943A1 (en) Systems and methods for uav interactive instructions and control
US10630962B2 (en) Systems and methods for object location
WO2020024134A1 (zh) 轨迹切换的方法和装置
US9342927B2 (en) Augmented reality system for position identification
WO2020014909A1 (zh) 拍摄方法、装置和无人机
WO2019113966A1 (zh) 一种避障方法、装置和无人机
WO2020113423A1 (zh) 目标场景三维重建方法、系统及无人机
WO2020014987A1 (zh) 移动机器人的控制方法、装置、设备及存储介质
CN110533719B (zh) 基于环境视觉特征点识别技术的增强现实定位方法及装置
Eynard et al. Real time UAV altitude, attitude and motion estimation from hybrid stereovision
WO2018112848A1 (zh) 飞行控制方法和装置
CA3069813C (en) Capturing, connecting and using building interior data from mobile devices
TWI726536B (zh) 影像擷取方法及影像擷取設備
Qayyum et al. Inertial-kinect fusion for outdoor 3d navigation
KR102506411B1 (ko) 차량의 위치 및 자세 추정 방법, 장치 및 이를 위한 기록매체
US11514597B1 (en) Single-camera stereoaerophotogrammetry using UAV sensors
Kobzili et al. Multi-rate robust scale estimation of monocular SLAM
Jouir et al. Vision-only egomotion estimation in 6DOF using a sky compass
Haige et al. Improved Stereo Vision Robot Locating and Mapping Method
US20220343536A1 (en) Three-dimensional map estimation apparatus and obstacle detection apparatus
WO2021111613A1 (ja) ３次元地図作成装置、３次元地図作成方法、及び３次元地図作成プログラム
KR20220138824A (ko) 포인트-라인 피처를 이용한 슬램의 최적화 방법 및 장치
KR20220123901A (ko) 무인 비행체 또는 항공기에 의해 촬영된 항공 영상에 기반하여 hd 맵을 생성하는 방법 및 시스템

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18928737

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18928737

Country of ref document: EP

Kind code of ref document: A1