WO2022261901A1 - 无人机降落控制方法、装置、无人机、系统及存储介质 - Google Patents

无人机降落控制方法、装置、无人机、系统及存储介质 Download PDF

Info

Publication number
WO2022261901A1
WO2022261901A1 PCT/CN2021/100685 CN2021100685W WO2022261901A1 WO 2022261901 A1 WO2022261901 A1 WO 2022261901A1 CN 2021100685 W CN2021100685 W CN 2021100685W WO 2022261901 A1 WO2022261901 A1 WO 2022261901A1
Authority
WO
WIPO (PCT)
Prior art keywords
parking
uav
scene image
unmanned aerial
aerial vehicle
Prior art date
Application number
PCT/CN2021/100685
Other languages
English (en)
French (fr)
Inventor
周游
徐彬
陈伟航
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/CN2021/100685 priority Critical patent/WO2022261901A1/zh
Publication of WO2022261901A1 publication Critical patent/WO2022261901A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/12—Target-seeking control

Definitions

  • the present application relates to the technical field of UAV intelligent control, and in particular to a UAV landing control method, a control device, a UAV, equipment for parking a UAV, a UAV landing control system, and a computer-readable storage medium .
  • the application provides a method for controlling the landing of drones, a control device, drones, equipment for parking drones, and a landing control system for drones and computer readable storage media.
  • a UAV landing control method including: during the landing process of the UAV, acquiring the scene image of the parking area collected by the sensor of the UAV, the parking area Including a plurality of parking stands, the plurality of parking stands including at least two parking stands with different identifications; identifying the identifications of the parking stands in the scene image, and obtaining the identification distribution information of the plurality of parking stands; Based on the identified identification distribution information, determine the target parking stand of the UAV among the plurality of parking stands observed in the scene image; and control the UAV to land on the target parking stand.
  • another drone landing control method including: during the landing process of the drone, acquiring the scene image of the parking area collected by the sensor of the drone, the parking The airport includes a plurality of parking stands, and the multiple parking stands include at least two parking stands with different identifications; identifying the identifications of the parking stands in the scene image, and obtaining the identification distribution information of the plurality of parking stands ; Based on the identification distribution information obtained by identification, determine a target parking area in the parking area observed by the scene image; control the UAV to move to the target parking area.
  • a control device used for the landing control of the drone, including a memory, a processor, and a computer program stored in the memory and operable on the processor, When the processor executes the program, the steps of the methods described in the first aspect and the second aspect of the embodiments of the present application are implemented.
  • a UAV includes a sensor for collecting scene images of the parking lot and a control device; wherein, the control device is used for landing of the UAV Control, including a memory, a processor, and a computer program stored on the memory and operable on the processor, when the processor executes the program, the steps of the methods described in the first aspect and the second aspect of the embodiment of the present application are implemented .
  • a device for parking a drone includes a plurality of parking stands, and the plurality of parking stands include at least two parking stands with different identifications, so The identification distribution information of the plurality of parking stands can be used to locate each of the parking stands.
  • a drone landing control system includes a drone and equipment for parking the drone; the drone includes a A sensor and a control device for collecting scene images of the parking lot; wherein the control device is used for landing control of the unmanned aerial vehicle, including a memory and a processor and a computer program stored on the memory and operable on the processor, the When the processor executes the program, the steps of the method described in the first aspect and the second aspect of the embodiment of the present application are realized; the device includes a plurality of parking positions, and the plurality of parking positions include at least two parking positions with different identifications. For parking stands, the identification distribution information of the plurality of parking stands can be used to locate each of the parking stands.
  • a computer-readable storage medium stores computer instructions, and when the computer instructions are executed, the first aspect and the second aspect of the embodiments of the present application are implemented. The steps of the method described in the aspect.
  • the UAV landing control method provided in the embodiment of the present application obtains the scene image of the parking area containing multiple parking bays during the UAV landing process, and then obtains a plurality of parking spaces based on the recognition of the scene images
  • the identification distribution information of the parking stand determines the target parking stand corresponding to the UAV, and then controls the UAV to land at the target parking stand.
  • the UAV landing control method can determine the target parking position corresponding to the UAV according to the identification distribution information of multiple parking positions in the collected scene image, and can be applied to each on each UAV, so that each UAV can automatically and accurately locate the corresponding target parking position, thereby allowing multiple UAVs to automatically land on their corresponding target parking positions at the same time and ensure the safety of the UAV landing. , to overcome the time-consuming, low efficiency and other problems existing in related technologies.
  • Fig. 1 is a flowchart of a landing control method for a drone according to an exemplary embodiment of the present application.
  • Fig. 2 is a schematic diagram of an airport including multiple parking bays according to an exemplary embodiment of the present application.
  • Fig. 3 is a schematic diagram of multiple groups of different types of identifiers according to an exemplary embodiment of the present application.
  • Fig. 4 is a schematic diagram showing a principle of determining a target parking position according to an exemplary embodiment of the present application.
  • Fig. 5A is a representation of identification distribution information according to an exemplary embodiment of the present application.
  • Fig. 5B is a representation of another identification distribution information according to an exemplary embodiment of the present application.
  • Fig. 6 is a schematic diagram showing another principle of determining a target parking position according to an exemplary embodiment of the present application.
  • Fig. 7 is a flow chart of controlling a UAV to land at a target stand according to an exemplary embodiment of the present application.
  • Fig. 8A is a schematic diagram showing the principle of controlling the UAV to move to the target parking area according to an exemplary embodiment of the present application.
  • Fig. 8B is a schematic diagram showing another principle of controlling the UAV to move to the target parking area according to an exemplary embodiment of the present application.
  • FIG. 9A and FIG. 9B are schematic diagrams of target parking areas determined by a drone at different heights according to an exemplary embodiment of the present application.
  • Fig. 10 is a schematic diagram of a scene image collected by a sensor of an unmanned aerial vehicle according to an exemplary embodiment of the present application.
  • Fig. 11 is a flow chart of obtaining identifier distribution information according to an exemplary embodiment of the present application.
  • Fig. 12 is a schematic diagram of a principle of correcting a recognized marker according to an exemplary embodiment of the present application.
  • Fig. 13 is a flow chart showing another method for obtaining identifier distribution information according to an exemplary embodiment of the present application.
  • Fig. 14 is a flow chart of another drone landing control method according to an exemplary embodiment of the present application.
  • Fig. 15 is a flowchart showing a drone landing to an idle parking space according to an exemplary embodiment of the present application.
  • Fig. 16 is a schematic structural diagram of a control device for landing control of a drone according to an exemplary embodiment of the present application.
  • first, second, third, etc. may be used in the present disclosure to describe various information, the information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of the present disclosure, first information may also be called second information, and similarly, second information may also be called first information. Depending on the context, the word “if” as used herein may be interpreted as “at” or “when” or “in response to a determination.”
  • the UAV formation flight is composed of multiple UAVs forming a queue to execute a specific flight mode, which has a grand scene, can create a shocking atmosphere, and provide visual stimulation for the viewers. Therefore, the UAV formation flight has also become An important application scenario of drones, for example, has become a common celebration and entertainment performance program on holidays and so on.
  • the number of UAVs in the UAV formation flight queue is often hundreds or even thousands. How to carry out efficient and safe landing control management is still a difficult problem in the field of UAV intelligent control.
  • the first method is to make multiple drones in the drone queue land randomly and disorderly. This landing control method has a certain risk of collision, and after landing, multiple drones need to be checked one by one. UAVs are numbered and then recycled according to the number, which not only has a relatively large safety risk, but also has relatively low recovery efficiency.
  • the second way is to make multiple drones in the drone queue line up and land one by one. Although this can eliminate a certain risk of drone collision, it is very time-consuming.
  • the third way is to re-form multiple drones in the drone queue to form an orderly queue, and then control multiple drones to land simultaneously based on global positioning information.
  • the positioning method is not accurate enough, and the error is usually in the range of tens of meters. Therefore, in order to avoid collisions between multiple drones, there are high requirements for the size and flatness of the landing site, and a large parking area is required. The ground of the parking lot is flat.
  • UAVs perform specific logistics transportation tasks, and it is also necessary to provide multiple parking areas for multiple UAVs to land at different parking positions at high altitudes to complete the predetermined cargo transportation tasks. Improving the reliability of the UAV's full-scenario work, improving the control performance during take-off and landing, and improving the landing accuracy of the UAV in a multi-stop scenario have become key research directions for related technologies.
  • an embodiment of the present application provides a landing control method for a drone.
  • the method can be applied to the landing control of multiple unmanned aerial vehicles in the unmanned aerial vehicle queue, and the multiple unmanned aerial vehicles can be fixed-wing unmanned aerial vehicles, helicopter unmanned aerial vehicles, flapping-wing unmanned aerial vehicles, multiple Various types of UAVs such as rotor UAVs.
  • the application scenario of this method may be that UAVs fly in formation to perform specific tasks, such as UAV formation flight performances, etc., which is not limited in this embodiment of the present application.
  • the UAV landing control method provided in the embodiment of the present application may include the following steps:
  • Step 101 during the landing process of the UAV, obtain the scene image of the parking area collected by the sensor of the UAV;
  • Step 102 identifying the identification of the parking stand in the scene image, and obtaining the identification distribution information of a plurality of parking stands;
  • Step 103 based on the identified identification distribution information, determine the target parking stand of the UAV among the plurality of parking stands observed in the scene image;
  • Step 104 controlling the UAV to land to the target parking space.
  • the parking lot mentioned in step 101 may be a dedicated site for parking drones, or a device for parking drones, such as a drone landing pad or the like.
  • the parking lot includes a plurality of parking stands, and each parking stand is used to park an unmanned aerial vehicle.
  • the plurality of parking stands include at least two parking stands with different identifications, that is, the identifications of the plurality of parking stands. There are at least two types of identifications, which will be described in conjunction with FIG. 2 .
  • Fig. 2 is a schematic diagram of a parking lot including multiple parking stands, where 201 represents a parking stand, and each ellipse 202 in the parking stand 201 represents a parking stand. In the schematic diagram shown in Fig.
  • each parking stand has an identification, and there are two types of identification - identification 1 and identification 2, which means that a plurality of parking positions include at least two different identifications. stand.
  • the plurality of parking positions in the parking lot described in FIG. more than one species.
  • some parking positions may have no signs, some parking positions may have one or more signs, and no sign is a special sign type to realize multiple parking positions.
  • the parking stands include at least two parking stands with different identifiers, which is not limited in this embodiment of the present application.
  • the type of the logo is associated with the appearance attribute of the parking stand. That is, the multiple parking stands included in the parking place have different appearance attributes, so the different appearance attributes of the multiple parking stands can be used as different types of identifications.
  • the type of the logo is associated with one or more of the following appearance attributes of the parking stand: outline, color, size, material, identifier and so on. That is, the multiple parking stands contained in the parking place may have different appearance attributes such as outlines, colors, sizes, materials, and identifiers, and parking stands with the same appearance attribute are the same type of identification.
  • the plurality of parking stands contained in the parking lot are parking stands with a circular outline and a parking stand with a quadrangular outline, then the circular outline is one type of identification, and the quadrilateral outline is another type of identification;
  • the parking lot contains a plurality of parking stands, some of which have the identifier " ⁇ ", and some of the parking stands have the identifier " ⁇ ", then the identifier " ⁇ " is a type of identification, and the identification
  • the symbol " ⁇ " is another type of identification; for another example, if the multiple parking positions contained in the parking place are white parking positions and black parking positions, then white is a type of identification, and black is Another type of identification.
  • only two types of the mark are used as an example for illustration. Those skilled in the art should understand that when the mark has multiple types and is similar to two types of marks, this embodiment of the present application does not make any repeat.
  • the type of the identification may be associated with an identifier located on the parking stand.
  • the identifier can be set on the parking stand when the parking stand leaves the factory, such as being directly set on the parking stand by printing, or on an independent object such as a sticker.
  • the UAV landing control method described in the embodiment of the present application is applied, it is placed on multiple parking bays in the parking area, which is not limited in the embodiment of the present application.
  • the identifier associated with the type of the identifier may have different orientations, outlines, colors, sizes, materials, and the like. As shown in FIG. 3 , taking two different types of identifiers as an example, schematic diagrams of identifiers with different orientations, outlines, colors, sizes, materials, etc. are respectively given in FIGS. 3A-3E . In FIG. 3A, the identifier is an arrow indicating a direction, and the arrow points to an identifier in a specific direction, which represents a type of identification; in FIG.
  • the identifier has a different Outline, the identifier of a contour represents a type of identification; in Figure 3C, the identifier has different colors, and the identifier of the same color represents a type of identification; in Figure 3D, The identifiers have different sizes. Similar to the previous one, an identifier of one size represents a type of identification; in FIG. 3E , different grayscales represent that the identifiers have different materials, for example, for For some materials, the reflectivity is weak, while for other materials, the reflectivity is strong, you can use identifiers made of different materials with a preset difference in reflectivity, each representing a type of logo .
  • the type of the mark may also be associated with the appearance attribute of objects around the parking stand in some embodiments.
  • objects with different appearance attributes such as flags of different colors, can be placed around the parking lot, and the different appearance attributes of objects around the parking lot can be used to represent the different types of signs. This is not limited.
  • the parking stands included in the parking lot have different signs, which can be various types of signs, and the type of the signs can be associated with the appearance attributes of the parking stands, and can also be associated with the parking stands. It is easy to realize the setting of different types of signs, which can improve the flexibility and scope of application of the UAV landing control method provided by the embodiment of the present application, and reduce the need for The implementation difficulty of the above method.
  • the scene image of the parking area collected by the sensor of the UAV is acquired, and the acquired image may be the scene image of the parking area including all parking positions, or It may be a scene image of an airport field including some parking stands, which is not limited in this embodiment of the present application.
  • the identifier distribution information identified in step 102 may include the identifiers of all the parking stands in the plurality of parking stands in the airport, and may also be the identifiers of some of the parking stands. logo. Still referring to FIG. 2 , the identification distribution information obtained by identification may be the identification 201 including all parking stands, or the identification 203 including part of the parking stands.
  • the identification distribution information obtained by identification includes the identification of all parking stands, or in step 101
  • the scene image of the parking lot collected in the above includes some of the parking stands, and the identified distribution information of the identification includes the identification of some of the parking stands, all of which can realize the UAV landing control method described in the embodiment of the present application. , so as to improve the applicability and flexibility of the method described in the embodiment of the present application, so that the method described in the embodiment of the present application is easy to implement and has high reliability.
  • identifying the identity of the parking stand in the scene image to obtain the identity distribution information of a plurality of parking stands can be realized by referring to related technologies, such as based on image moments, image feature recognition, etc., this paper
  • the embodiment of the application does not limit the identification method adopted.
  • the identification distribution information is obtained based on identification, and the target parking position of the UAV is determined among the plurality of parking positions observed in the scene image, which can be based on multiple way to achieve.
  • the types of the marks of a plurality of said parking spaces including two types, one is an arrow facing left, and the other is an arrow facing upwards as an example several kinds of target stops that can realize the determination of the drone are introduced. bit example.
  • these embodiments are only exemplary illustrations, and other ways can also be used to determine the target parking position of the UAV among the multiple parking positions observed by the scene image. The embodiment does not limit this.
  • the UAV landing control method provided in the embodiment of the present application further includes: obtaining the preset identification distribution information of multiple parking stands in the area where the target parking stand is located;
  • Step 103 based on the identified identification information, determining the target parking position of the UAV among the plurality of parking positions observed in the scene image, including:
  • the target parking position of the UAV is determined among the plurality of parking positions observed in the scene image.
  • FIG. 4 shows a schematic diagram of the principle of determining the target parking stand of the UAV among the plurality of parking stands observed in the scene image based on the preset marker distribution information.
  • 401 is the preset identification distribution information of a plurality of parking stands in the area where the target parking stand is located
  • 402 is the target parking stand
  • 403 is the identification based on step 102
  • the acquired scene image contains The identity distribution information of multiple UAVs.
  • the preset identification distribution information 401 may be acquired first.
  • the timing of obtaining the preset identifier distribution information is not limited in the embodiment of the present application. It may be that the preset identifier distribution information is obtained before the drone takes off, or it may be during the flight of the drone. Acquiring the preset identifier distribution information sent by the control center, or obtaining the preset identifier distribution information from other devices or systems based on various communication methods before the UAV lands.
  • the preset logo distribution information 401 may include the logo distribution of all parking bays in the airport, or may include the logo distribution of some parking bays in the airport, which is not limited in this embodiment of the present application.
  • the preset marker distribution information includes the marker distribution of the target parking stand 402 and multiple parking stands around it, it is equivalent to an accurate "map" and can be used as reference information. Comparing the identification distribution information 403 of multiple parking stands obtained based on the identification in step 102 with the preset identification distribution information 401, the current position of the drone and the target parking stand can be located. , and then based on the guidance of the preset marker distribution information, step 104 is executed to control the UAV to land on the target stand.
  • the preset identifier distribution information and the identified identifier distribution information may be expressed in various forms.
  • the preset logo distribution information and the logo distribution information obtained by the identification may be the appearance attribute of the parking stand associated with the logo and/or the surrounding area of the parking stand
  • the appearance attribute of the object and the appearance attribute of the identifiers located on the plurality of parking stands are represented by the form.
  • the identification is associated with identifiers located on a plurality of the parking stands—a leftward arrow and an upward arrow, then, the preset identification distribution information and the identification distribution obtained by the identification
  • the information may be shown as 501 and 502 in FIG. 5A, respectively.
  • the preset marker distribution information may include marker distribution information of all parking stands in the kiosk airport, that is, a global marker distribution information representing a global parking stand situation.
  • a scene image under the drone may be collected through a sensor of the drone.
  • Identification distribution information of a plurality of parking stands is obtained from the scene image.
  • the identified identification distribution information can be expressed in the form of coded information.
  • compare the identified identification distribution information with the aforementioned global identification distribution information and calculate which area of the global identification distribution information the currently identified identification distribution information belongs to. In other words, based on the comparison result, it can be deduced that It is known which position in the global area of the parking lot the scene observed by the current UAV belongs to.
  • the relative positional relationship between the UAV and the target parking space or the target parking area in the parking area can be known, and then, based on the relative positional relationship, the UAV is controlled to adjust the posture in the air and gradually approach the target parking space or the target parking area.
  • the area is close.
  • the preset identifier distribution information and the identifier distribution information obtained by the identification can also be expressed in binary, octal, decimal, etc. according to the number of types of identifiers. Expressed in equal base numbers.
  • the two types of logos can be represented by binary numbers 0 and 1, as shown in Figure 5B, where "0" represents a leftward arrow, and "1" represents a Arrows above; for another example, there are four types of identifiers, and 00, 01, 10, and 00 may be used to represent the four types of identifiers respectively.
  • the representation form of the preset identifier distribution information and the identifier distribution information obtained by the identification can be determined based on the number of identifier types and actual application scenarios. There is no limit to this.
  • the preset identification distribution information and the identification distribution information obtained by the identification are represented in the form of binary numbers, octal numbers, decimal numbers, etc., which can save certain calculation and storage resources. . And comparing the identification distribution information obtained by the identification with the preset identification distribution information, when determining the target parking position of the UAV among the plurality of parking positions observed in the scene image, based on the Compared with the number, the error rate is low, which can improve the accuracy of the UAV landing control method described in the embodiment of the present application.
  • step 103 determining the target parking stand of the UAV among the plurality of parking stands observed in the scene image based on the identified identification distribution information includes:
  • a target parking stand is determined among the plurality of parking stands observed in the scene image, wherein the identification distribution of the parking stands in preset orientations around the target parking stand conforms to a preset Identifies the distribution criteria.
  • FIG. 6 shows identification distribution information 601 of multiple parking stands obtained by identifying the collected scene images, where 602 is the target parking stand.
  • the preset marker distribution information may not be acquired, but the target parking stand is determined based on preset marker distribution conditions.
  • the corresponding preset identification distribution condition is: the target parking lot corresponding to the UAV is an arrow facing left, and the adjacent upper left corner position is The logo and the logo at the lower left corner adjacent to it are the same as itself, and the logo at the upper right corner adjacent to it and the logo at the lower right corner adjacent to it are different from itself. Then, based on the preset identification distribution condition, it can be uniquely determined that the target parking stand corresponding to the UAV is at 602 . For multiple drones in the drone queue, as long as the preset identification distribution conditions corresponding to each drone are different, it is possible for the multiple drones to uniquely locate each drone. The corresponding target parking position.
  • the preset identification distribution conditions corresponding to each unmanned aerial vehicle can be more complicated, for example, with more
  • the target parking stand is defined by the identification of the position, that is, the target parking stand is defined by the identification of a plurality of positions on the first floor, the second floor, or even the third floor, etc. around the target parking stand.
  • the target parking stand with whether the target parking stand is located at the border of the airport field and at which azimuth, and/or use more types of signs to indicate the parking stands at different locations, and then
  • the identification of multiple parking stands around the target parking stand has a more complex combination to define the target parking stand, etc., which is not limited in this embodiment of the present application.
  • the target parking stand is determined among the plurality of parking stands observed in the scene image, and there is no need to obtain multiple parking stands in the area where the target parking stand is located.
  • the target parking position can be determined by the preset identification distribution information.
  • the relative position between the current position of the drone and the target parking stand can be obtained, and then based on the guidance of the scene image including the target parking stand, Step 104 is executed to control the UAV to land at the target parking space.
  • the target parking stand of the UAV is determined among the plurality of parking stands observed in the scene image by identifying the identification distribution information obtained and the preset identification distribution conditions, Not only can the relative position of the current position of the drone and the target parking stand be located, but also there is no need to obtain the preset identification distribution information of a plurality of parking stands in the area where the target parking stand is located.
  • the advantages of fast positioning and accurate positioning can improve the landing efficiency of multiple drones in the drone queue, reduce the requirements for the required landing site, and achieve efficient and safe landing control.
  • the above-mentioned determination of the target parking stand of the UAV among the multiple parking stands observed in the scene image may be based on multiple obtained information of the area where the target parking stand is located. It can also be realized based on the preset identification distribution information of the above-mentioned parking stands, or it can be realized only based on the preset identification distribution conditions. Of course, it can also be based on other methods, based on the identified identification distribution information, in the above-mentioned
  • the target parking position of the UAV is determined from the plurality of parking positions observed in the scene image.
  • the determination of the target parking stand may be based on only one method, or the above-mentioned solutions may be adopted at the same time, so as to avoid the loss or error of the preset marker distribution information or the preset marker distribution conditions, etc., resulting in
  • the drone landing method provided in the embodiment of the present application cannot be executed, and the reliability and robustness of the method provided in the embodiment of the present application are improved.
  • step 104 controlling the UAV to land on the target parking space, as shown in FIG. 7 , includes:
  • Step 701 determining the relative positional relationship between the UAV and the target parking stand according to the scene image
  • Step 702 based on the relative positional relationship, control the UAV to land on the target stand.
  • FIG. 4 and Figure 6 are taken as examples for illustration.
  • 401 is the preset identification distribution information, that is, the "map" of the airport that contains multiple parking positions; 403 corresponds to the dotted box part in 401, that is, the UAV The current position; and 402 is the target stand corresponding to the drone. Then, from the current position 403 of the drone and the relative positional relationship between the target parking space 402 in the "map" 401, it can be determined that the drone should be controlled along the tail of the "upward arrow". Direction movement, so as to move to the sky above the target parking stand and land on the target parking stand.
  • 601 is the identification distribution information corresponding to the scene image
  • 602 is the target parking stand.
  • the relative positional relationship between the current position of the drone and the target parking stand can be known.
  • the target parking stand should be in the center of the scene image. Therefore, based on this condition, it can be known that the UAV should be controlled to move in the direction of the "upward arrow" tail, so as to move to the sky above the target parking stand and land on the target parking stand.
  • the UAV landing control method provided in the embodiment of the present application further includes: based on the identification distribution information obtained by identification, determining the target parking area in the parking area observed by the scene image; Controlling the unmanned aerial vehicle to move to the target parking area.
  • the target parking area can be determined based on the identification distribution area obtained first, and the target parking area is an area containing a plurality of parking stand identifications, Controlling the unmanned aerial vehicle to move to the target parking area.
  • the size of the target parking area can be preset based on actual application needs, for example, it can be an area containing N*M marks in the center of the target parking stand, M and N are positive integers, and M and N can be Same or different.
  • the size of the target parking area can be maintained until the target is accurately located in the target parking area.
  • the parking position is used to control the drone to accurately land on the target parking position. That is, as shown in FIG. 8A , during the landing process of the drone to the target parking position (the position indicated by the black arrow), it moves to the target parking areas 802, 803 and 804 of the same size in sequence until it is at the target parking space. Accurately locate the target parking stand in the center of the target parking area 803 in the parking area 803 , and then land towards the target parking stand.
  • the size of the target parking area may be gradually reduced until the target parking area only includes the target parking space, Then, it is realized to control the UAV to accurately land on the target parking position. That is, as shown in FIG. 8B , during the landing process of the drone to the target parking position (the position indicated by the black arrow), it moves to the target parking areas 802, 803 and 804 whose sizes gradually decrease until Landing towards the target parking stand 804 .
  • control of the UAV to move to the target parking area may be that the UAV maintains a constant flight height and only changes its position on a plane parallel to the horizontal plane to achieve the target The movement of the parking area; it may also be that the UAV not only has a change in flying height, but also changes in position on a plane parallel to the horizontal plane, which is not limited in this embodiment of the present application.
  • step 701 determines the target parking area in the parking field observed by the scene image, which may be based on the collected scene image, and performs identification again.
  • Recognition so as to determine the target parking area, may also be using the recognition result of the scene image in step 102 to determine the target parking area, which is not limited in this embodiment of the present application.
  • the target parking area where the target parking stand is located is determined based on the result of identifying the logo contained in the scene image and the target parking space, first control the UAV to move to the target parking area, and then control the UAV to land at the target parking space, that is, first control the UAV to move to a larger size
  • the target area that is, the target parking area
  • the UAV landing control method described in the embodiment of the present application further includes:
  • Step 103 based on the identification distribution information obtained by identification, determine the target parking position of the UAV among the plurality of parking positions observed in the scene image;
  • Step 104 the control of the UAV to The landing at the target stand includes:
  • the target parking area is determined in the parking area observed by the scene image, and the UAV is controlled to fly to the target. Parking area movement;
  • the drone landing control method also includes:
  • the second flying height is smaller than the first flying height.
  • the flying height of the UAV is obtained, based on the identification of multiple parking positions contained in the scene image identification, determine the target parking area or target parking position corresponding to the UAV at different flight altitudes, and guide the landing control of the UAV.
  • 901 is the scene image of the parking area collected by the sensor of the UAV, and the identification distribution information of a plurality of parking stands is obtained
  • 902 is the target parking stand of the UAV.
  • the identification distribution information of the plurality of parking positions can be obtained based on the recognition of the scene image, and then the current position of the UAV can be determined.
  • the position of the target parking stand can determine the relative positional relationship between the current position of the UAV and the target parking stand, so as to guide the UAV to move to the target parking area where the target parking stand is located.
  • the unmanned aerial vehicle As the unmanned aerial vehicle lands, when the unmanned aerial vehicle is at a lower flight altitude, it can be more accurately positioned based on the identification distribution information of a plurality of the parking stands obtained from the recognition of the scene image. The location of the target parking stand, so when the UAV is at the second flying height, the UAV can be controlled to land at the determined target parking stand.
  • the first flying height may include multiple flying sub-heights, and at each flying sub-height, a corresponding sub-target parking area is determined in the parking area observed by the scene image, Controlling the UAV to move to the sub-target parking area.
  • the first flying altitude including a first sub-flying altitude and a second sub-flying altitude as an example, and the first sub-flying altitude is greater than the second sub-flying altitude, it will be described with reference to FIG. 9A and FIG. 9B .
  • the UAV is at the first sub-flying height, based on the identified identification distribution information 901, as shown in FIG.
  • the sub-target parking area determines the first sub-target parking area in the parking area observed by the scene image 903. Control the UAV to move to the first sub-target parking area 903; when the UAV is at the second sub-flying height, obtain the identification distribution information 901 based on the identification, as shown in FIG. 9B , determining a second sub-target parking area 904 in the parking area observed by the scene image, and controlling the UAV to move to the second sub-target parking area 904 .
  • the sub-target parking area can be reduced as its corresponding sub-flying height decreases, and then more areas where non-target parking stands are located can be excluded, and the position of the target parking stand is gradually and accurately determined. .
  • the scene image used to obtain the marker distribution information may be the scene image obtained at a preset flying height after the UAV receives the landing signal.
  • Scene image of multiple parking bays may include: acquiring the scene images collected by the UAV at different flight heights.
  • the UAV When the UAV is at different flight heights, obtain scene images corresponding to the different flight heights that contain a plurality of parking bays, and based on the recognition of the signs contained in the scene images, it is possible to obtain The identification distribution information corresponding to the different flying heights of the UAV. For example, when the UAV is located at a high altitude above the first threshold, in the hollow between the first threshold and the second threshold, and near the ground below the second threshold, the parking spaces containing a plurality of parking spaces are respectively collected. Scene image of the airport. By identifying the scene image, the identification distribution information corresponding to the flight height can be obtained, and the relative positional relationship between the UAV and the target parking stand at different flight heights can be determined.
  • the UAV landing control method provided in the embodiment of the application is based on identifying the signs contained in the collected scene images, and determining the target parking area and the target parking position among multiple parking positions, it is based on this application
  • the method provided by the embodiment, the positioning error of the target parking area and the target parking position is far smaller than the positioning error caused by positioning devices such as GPS in the related art, and can be compared with the unmanned aerial vehicle landing control method adopted in the related art. Achieve more precise landing control.
  • the target stoppage determined based on the recognition of the scene image corresponding to the higher flight altitude The error of the area will be slightly larger than the error of the target parking area determined by the lower flight altitude, and as the flight altitude continues to decrease, the error of the target parking area determined based on the recognition of the scene image will become smaller and smaller until Accurately locate the target parking stand.
  • controlling the UAV to move to the target parking area and controlling the UAV to land on the target parking stand can be continuously accurate as the flying height decreases, and finally the UAV is controlled accurately. Landing on the target parking stand.
  • the obtained identification distribution information is gradually more accurate, and the determined target parking area and target parking position can be gradually more accurate.
  • the precise landing of the unmanned aerial vehicle is realized.
  • This method can be applied to the simultaneous landing of multiple UAVs in the UAV queue, and then can overcome the defects in related technologies that the UAVs take a long time to land or require a large landing site, etc. Efficient and safe landing control of multiple UAVs in a platoon.
  • each unmanned aerial vehicle all implements the scheme of this embodiment, so in the process of landing, each unmanned aerial vehicle all can be more accurately adjusted to its corresponding target parking position or the sky over the target parking area, along with the unmanned aerial vehicle Gradually landing, more precise landing control can be realized based on this scheme.
  • the UAV may not be located above the parking area containing multiple parking positions, so it is impossible to directly obtain the scene image of the parking area containing multiple parking positions, so in some implementations
  • the scene image of the parking lot collected by the sensor of the unmanned aerial vehicle may include:
  • the scene image is acquired.
  • the acquisition of the location of the UAV can be implemented with reference to related technologies. For example, it can be based on the Global Positioning System (Global Positioning System, GPS) of the drone, or it can be based on the visual-inertial odometer (Visual-Inertial Odometry, VIO), of course, it can also be based on other methods to obtain the drone. Human-machine positioning. When it is determined based on the acquired location that the UAV is not in the area corresponding to the parking area, the UAV can be controlled to move to the area under the guidance of the location information of the UAV. The area corresponding to the parking lot, and then acquire the scene image. Of course, when the UAV is not in the area corresponding to the parking area, other methods can also be used to first make the UAV in the area corresponding to the parking area, and then acquire the scene image. The embodiment does not limit this either.
  • GPS Global Positioning System
  • VIO Visual-Inertial Odometry
  • the scene image of the parking lot collected by the sensor of the unmanned aerial vehicle may be inconsistent with the real scene of the parking lot, and the inconsistent situation includes the scene
  • the image has a rotation transformation relative to the scene of the real parking field.
  • the markers include a left-facing arrow and an upward-facing arrow, and the distribution of the markers is shown in FIG. 6 .
  • the acquired scene image of the parking area collected by the sensor of the drone may be as shown in Figure 10, that is, in the identification distribution information, the identification is an arrow facing the upper left corner and an arrow facing the lower left corner .
  • step 101 the acquisition of the scene image of the parking area collected by the sensor of the drone, as shown in FIG. 11 , includes:
  • Step 1101 acquiring orientation information of the drone when shooting the scene image
  • Step 1102 correcting the scene image based on the orientation information
  • Step 1103 identify the sign of the parking stand in the rectified scene image, and obtain the sign distribution information of a plurality of the parking stands.
  • the orientation information of the UAV may be obtained based on an azimuth positioning device of the UAV, for example, an electronic compass, etc., which is not limited in this embodiment of the present application.
  • the scene image can be adjusted so that the orientation of the scene image is consistent with the orientation of the signs in the real scene of the parking lot, and an accurate distribution of the signs can be obtained information, which can ensure the accuracy of the landing control of the drone.
  • the scene image may also be corrected only based on the markers contained in the scene image, so as to obtain accurate marker distribution information.
  • FIG. 12 when the marks are of two types, one is an arrow pointing left, and the other is an arrow pointing upwards.
  • the scene image collected by the sensor is shown in the left figure of Figure 12, since the logo only has the above two situations, it can be uniquely determined that the scene image needs to be rotated by a certain angle, so that as shown in Figure 12
  • the mark in the first row and first column shown is an upward arrow, and the orientation of the scene image can be corrected.
  • the corrected scene image is shown in the right figure of FIG. 12 .
  • other markers may also be set so that the scene image can be corrected based only on the marker of the scene image, so as to obtain accurate marker distribution information, which is not limited in this embodiment of the present application.
  • step 101 the acquisition of the scene image of the parking area collected by the sensor of the drone, as shown in FIG. 13 , includes:
  • Step 1301 adjusting the orientation of the sensor of the drone based on preset orientation information
  • Step 1302 collect the scene image based on the adjusted sensor.
  • the preset orientation information may be an orientation completely consistent with the real scene of the airport, or an orientation with a known angular difference from the actual scene of the airport.
  • the adjustment of the orientation of the sensor of the UAV can be determined based on the azimuth positioning device of the UAV, for example, an electronic compass, etc. Of course, it can also be based on the adjusted The orientation of the marker in the scene image collected by the sensor is determined, which is not limited in this embodiment of the present application. Of course, those skilled in the art should understand that it is also possible to adjust the orientation of the drone as a whole.
  • the orientation of the collected scene image is consistent with the parking position.
  • the orientation of the signs in the real scene of the airport is consistent, and the accurate distribution information of the signs is obtained, which can ensure the accuracy of the landing control of the drone.
  • the scene image of the parking area containing multiple parking bays is acquired, and then based on the analysis of the scene
  • the identification distribution information of multiple parking stands obtained by image recognition determines the target parking stand corresponding to the UAV, and then controls the UAV to land on the target parking stand.
  • the identification distribution information of multiple parking positions in the system can determine the target parking position corresponding to the UAV, which can be applied to each UAV, so that each UAV can automatically and accurately locate the corresponding target parking position, so that it can It allows multiple drones to automatically land on their corresponding target parking positions at the same time and ensures the safety of the drone landing, overcoming the time-consuming and inefficient problems of related technologies.
  • the embodiment of the present application also provides another drone landing control method, as shown in Figure 14, the method includes:
  • Step 1401 during the landing process of the UAV, acquire the scene image of the parking area collected by the sensor of the UAV, the parking area includes a plurality of parking stands, and the plurality of parking stands include at least two parking bays;
  • Step 1402 identifying the logo of the parking stand in the scene image, and obtaining the distribution information of the logos of a plurality of parking stands;
  • Step 1403 based on the identification distribution information obtained, determine the target parking area in the parking field observed in the scene image;
  • Step 1404 controlling the UAV to move to the target parking area.
  • the target parking area is a target parking space of the drone.
  • the target parking area is an area containing vacant parking bays.
  • the method further includes:
  • Step 1501 determining idle parking spaces in the target parking area
  • Step 1502 controlling the UAV to land at any of the free parking spaces.
  • determining the parking space in the target parking area can be realized by referring to related technologies.
  • the idle parking spaces in the target parking area may be determined based on infrared sensors, or based on methods such as image feature recognition.
  • the identification distribution information is obtained, thereby determining the target parking area, and controlling the UAV to move to the target parking area , based on the acquired scene image, it is possible to accurately control the movement of the UAV to the target parking area, thereby realizing accurate landing of the UAV.
  • the identifier distribution information obtained by the identification includes identifiers of all or part of the plurality of parking stands in the airport.
  • the method further includes: acquiring preset identification distribution information of a plurality of parking stands in the target parking area; based on the identified identification distribution information, observing in the scene image Determining the target parking area of the UAV in a plurality of parking positions, including: according to the identification distribution information obtained by identification and the preset identification distribution information, determining in the plurality of parking positions observed in the scene image The target parking area for the drone.
  • determining the target parking area of the UAV in the plurality of parking positions observed in the scene image based on the identification distribution information obtained by identification includes: based on the identification obtained by identification The distribution information is to determine a target parking area among the plurality of parking bays observed in the scene image, wherein the logo distribution information of parking bays at preset orientations around the target parking area conforms to a preset logo distribution condition.
  • the method further includes: acquiring the height of the UAV; and determining the unmanned aerial vehicle in a plurality of parking positions observed in the scene image based on the identification distribution information obtained by identification.
  • the target parking position of the drone; the controlling the drone to move to the target parking area includes: when the drone is at the first height, based on the identified identification distribution information, in the scene Determining a target parking area in the parking area observed by the image, and controlling the UAV to move to the target parking area; the method also includes: when the UAV is at a second height, based on the identification obtained The identification distribution information, determine the target parking position of the UAV among the plurality of parking positions observed in the scene image; control the UAV to land to the target parking position; wherein, the second height is less than the first height.
  • the acquiring the scene images of the parking area collected by the sensors of the UAV includes: acquiring the scene images collected by the UAV at different heights.
  • the controlling the UAV to land at the target parking position includes: determining the relative positional relationship between the UAV and the target parking area according to the scene image; based on the Relative to the positional relationship, the UAV is controlled to move to the target parking area.
  • the obtaining the scene image of the parking area collected by the sensor of the UAV includes: obtaining the location of the UAV; when the location of the UAV is in the area corresponding to the parking area, obtaining The scene image.
  • the type of the logo is associated with the appearance attribute of the parking stand; and/or, the type of the logo is associated with the appearance attribute of objects around the parking stand.
  • the type of the logo is associated with one or more of the following appearance attributes of the parking stand: outline, color, size, material, and identifier.
  • the acquisition of the scene image of the parking area collected by the sensor of the drone includes: acquiring the orientation information of the drone when the scene image is taken; correcting the orientation information based on the orientation information.
  • the scene image identifying the sign of the parking stand in the rectified scene image, and obtaining the sign distribution information of a plurality of the parking stands.
  • the acquiring the scene image of the parking area collected by the sensor of the drone includes: adjusting the orientation of the sensor of the drone based on preset orientation information; The sensor captures the scene image.
  • the embodiment of the present application also provides a control device, the control device is used for landing control of the drone, as shown in Figure 16, the control device includes a memory 1601, a processor 1602, and a computer program stored in the memory 1601 and operable on the processor 1602.
  • the processor executes the program, any method embodiment described above is implemented.
  • One processor 1602 is taken as an example in FIG. 16 , and the processor may include a GPU and a CPU.
  • the processor 1602 and the memory 1601 in the electronic device may be connected through a communication bus or in other ways. In FIG. 16 , connection through a communication bus is taken as an example.
  • the processor 1602 of the electronic device integrates the application programs provided in the foregoing embodiments.
  • the memory 1601 in the electronic device can be used to store one or more programs, and the programs can be software programs, computer-executable programs, and modules.
  • the processor 1602 executes various functional applications and data processing of the device by running software programs, instructions and modules stored in the memory 1601, that is, implements the above-mentioned various method embodiments.
  • the memory 1601 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application program required by at least one function; the data storage area may store data created according to the use of the device, and the like.
  • the memory 1601 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid-state storage devices.
  • the memory 1601 may further include memory located remotely relative to the processor 1602, and these remote memories may be connected to the device through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
  • the processor 1602 executes various functional applications and data processing by running the programs stored in the memory 1601 to implement the method provided in the embodiment of the present application.
  • the control device provided by the embodiment of the present application can determine the target parking stand corresponding to the UAV according to the identification distribution information of multiple parking stands in the collected scene image, and can be applied to each unmanned parking lot.
  • an embodiment of the present application also provides an unmanned aerial vehicle, the unmanned aerial vehicle includes a sensor for collecting scene images of the parking lot and the control device described in any of the foregoing embodiments.
  • the scene image of the parking area containing multiple parking bays can be obtained, and then based on the The identification distribution information of a plurality of parking stands obtained through the recognition of the above-mentioned scene image determines the target parking stand corresponding to the UAV, and then controls the UAV to land on the target parking stand.
  • the UAV provided by the embodiment of the present application can determine the target parking stand corresponding to the UAV according to the identification distribution information of multiple parking stands in the collected scene image, and can allow multiple The drones automatically land on their corresponding target parking positions at the same time and ensure the safety of the drone landing, overcoming the time-consuming and inefficient problems of related technologies.
  • the embodiment of the present application also provides a device for parking the unmanned aerial vehicle.
  • parking stands and the identification distribution information of the plurality of parking stands can be used to locate each of the parking stands.
  • the type of the logo is associated with the appearance attribute of the parking stand; and/or, the type of the logo is associated with the appearance attribute of objects around the parking stand.
  • the type of the logo is associated with one or more of the following appearance attributes of the parking stand: outline, color, size, material, and identifier.
  • the device includes a sensor and a processor, and the processor is configured to execute a preset program based on the information collected by the sensor.
  • the device is foldable, and the sensor is used to notify the processor when it detects that the drone has landed on a preset parking position; the processor is used to control the device to fold , to pack the drone.
  • the parking position can be provided with a parking platform and a storage box, and when the drone lands on the parking platform, the parking platform can perform specific mechanical actions to store the drone into The storage box.
  • the preset program includes one or more of the following operating procedures: charging the drone; loading cargo to the drone; unloading the drone; cleaning The drone; replacing some parts of the drone; repairing the parts of the drone, etc.
  • the equipment provided based on the embodiments of the present application can provide conditions for the rapid landing control and fast packaging of the UAV, thereby helping to realize safe and efficient landing of the UAV Control and fast storage, saving users' time and labor costs.
  • the embodiment of the present application also provides a UAV landing control system, the system includes the UAV described in any of the above embodiments and the device for parking the UAV, the relevant content has been mentioned above Detailed description, here, this application does not repeat it.
  • the UAV landing control system provided by the embodiment of the present application, it can be realized that during the UAV landing process, the scene image of the parking area containing multiple parking bays can be obtained, and then based on The identification distribution information of multiple parking stands obtained by identifying the scene image determines the target parking stand corresponding to the UAV, and then controls the UAV to land on the target parking stand.
  • the UAV provided by the embodiment of the present application can determine the target parking stand corresponding to the UAV according to the identification distribution information of multiple parking stands in the collected scene image, and can allow multiple The drones automatically land on their corresponding target parking positions at the same time and ensure the safety of the drone landing, overcoming the time-consuming and inefficient problems of related technologies.
  • the present application also provides a computer-readable storage medium, where the storage medium stores a computer program, and when the computer program is executed by a processor, any method described above is implemented.
  • the computer readable medium may be a computer readable signal medium or a computer readable storage medium.
  • a computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination thereof. More specific examples (non-exhaustive list) of computer readable storage media include: electrical connections with one or more leads, portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), Erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CDROM), optical storage device, magnetic storage device, or any suitable combination of the above.
  • a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.
  • the signal medium of the computer-readable storage medium may comprise a data signal propagating in baseband or as part of a carrier wave carrying computer-readable program code thereon. Such propagated data signals may take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing.
  • a computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. .
  • Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
  • Computer program code for performing the operations of the present application may be written in one or more programming languages or combinations thereof, including object-oriented programming languages—such as Java, Smalltalk, C++, and conventional Procedural programming language—such as "C" or a similar programming language.
  • the program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server.
  • the remote computer can be connected to the user computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (such as through the Internet using an Internet service provider). connect).
  • LAN local area network
  • WAN wide area network
  • connect an external computer (such as through the Internet using an Internet service provider).

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)
  • Traffic Control Systems (AREA)

Abstract

一种无人机降落控制方法、控制装置、无人机、用于停放无人机的设备、无人机降落控制系统以及计算机可读存储介质，所述方法包括：在无人机降落过程中，获取所述无人机的传感器采集的停机场的场景图像，所述停机场包括多个停机位，多个所述停机位包括具有不同的标识的至少两个停机位；在所述场景图像中识别所述停机位的标识，得到多个所述停机位的标识分布信息；基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位；控制所述无人机向所述目标停机位降落。基于本申请实施例所提供的方法，能够允许多个无人机同时自动降落至各自对应的目标停机位上，克服相关技术所存在的耗时、效率低下等问题。

Description

无人机降落控制方法、装置、无人机、系统及存储介质 技术领域
本申请涉及无人机智能控制技术领域，尤其涉及一种无人机降落控制方法、控制装置、无人机、用于停放无人机的设备、无人机降落控制系统以及计算机可读存储介质。
背景技术
随着无人机技术的不断发展和进步，无人机的应用越来越广泛。在一些场景中，往往涉及多台无人机的高频起降。相关技术中，提供了具有多个停机位的停机场，以供无人机降落。例如，无人机编队飞行表演成了一个节假日常见的庆祝以及娱乐表演节目，表演结束后，多台无人机需要分别降落至不同的停机位。如何有效地控制无人机高效且安全地降落，是目前无人机智能控制技术领域的一个难题。
发明内容
为克服相关技术所存在的耗时、效率低下等问题，本申请提供了一种无人机降落控制方法、控制装置、无人机、用于停放无人机的设备、无人机降落控制系统以及计算机可读存储介质。
根据本申请实施例的第一方面，提供一种无人机降落控制方法，包括：在无人机降落过程中，获取所述无人机的传感器采集的停机场的场景图像，所述停机场包括多个停机位，多个所述停机位包括具有不同的标识的至少两个停机位；在所述场景图像中识别所述停机位的标识，得到多个所述停机位的标识分布信息；基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位；控制所述无人机向所述目标停机位降落。
根据本申请实施例的第二方面，提供另一种无人机降落控制方法，包括：在无人机降落过程中，获取所述无人机的传感器采集的停机场的场景图像，所述停机场包括多个停机位，多个所述停机位包括具有不同的标识的至少两个停机位；在所述场景图像中识别所述停机位的标识，得到多个所述停机位的标识分布信息；基于识别得到的所述标识分布信息，在所述场景图像观测的所述停机场中确定目标停机区域；控制 所述无人机向所述目标停机区域移动。
根据本申请实施例的第三方面，提供一种控制装置，所述控制装置用于无人机的降落控制，包括存储器和处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现本申请实施例第一方面和第二方面所述的方法的步骤。
根据本申请实施例的第四方面，提供一种无人机，所述无人机包括用于采集停机场的场景图像的传感器以及控制装置；其中，所述控制装置用于无人机的降落控制，包括存储器和处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现本申请实施例第一方面和第二方面所述的方法的步骤。
根据本申请实施例的第五方面，提供一种用于停放无人机的设备，所述设备包括多个停机位，多个所述停机位包括具有不同的标识的至少两个停机位，所述多个停机位的标识分布信息可用于定位每个所述停机位。
根据本申请实施例的第六方面，提供一种无人机降落控制系统，所述无人机降落控制系统包括无人机以及用于停放无人机的设备；所述无人机包括用于采集停机场的场景图像的传感器以及控制装置；其中，所述控制装置用于无人机的降落控制，包括存储器和处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现本申请实施例第一方面和第二方面所述的方法的步骤；所述设备包括多个停机位，多个所述停机位包括具有不同的标识的至少两个停机位，所述多个停机位的标识分布信息可用于定位每个所述停机位。
根据本申请实施例的第七方面，提供一种计算机可读存储介质，所述计算机可读存储介质上存储有计算机指令，所述计算机指令被执行时实现本申请实施例第一方面和第二方面所述的方法的步骤。
本申请的实施例提供的技术方案可以包括以下有益效果：
本申请实施例所提供的无人机降落控制方法，通过在无人机降落过程中，获取包含多个停机位的停机场的场景图像，然后基于对所述场景图像的识别而得到的多个停机位的标识分布信息，确定所述无人机所对应的目标停机位，进而控制所述无人机向所述目标停机位降落。可以看到，本申请实施例所提供的无人机降落控制方法能够根据所采集的场景图像中的多个停机位的标识分布信息，确定无人机所对应的目标停机位，可应用于每个无人机上，使得每个无人机能够自动准确定位所对应的目标停机 位，从而能够允许多个无人机同时自动降落至各自对应的目标停机位上且保证无人机降落的安全性，克服相关技术所存在的耗时、效率低下等问题。
应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，并不能限制本说明书。
附图说明
为了更清楚地说明本申请实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1是本申请根据一示例性实施例示出的一种无人机降落控制方法的流程图。
图2是本申请根据一示例性实施例示出的一种包括多个停机位的停机场的示意图。
图3是本申请根据一示例性实施例示出的多组不同类型的标识的示意图。
图4是本申请根据一示例性实施例示出的一种确定目标停机位的原理示意图。
图5A是本申请根据一示例性实施例示出的一种标识分布信息的表示形式。
图5B是本申请根据一示例性实施例示出的另一种标识分布信息的表示形式。
图6是本申请根据一示例性实施例示出的另一种确定目标停机位的原理示意图。
图7是本申请根据一示例性实施例示出的一种控制无人机向目标停机位降落的流程图。
图8A是本申请根据一示例性实施例示出的一种控制无人机向目标停机区域移动的原理示意图。
图8B是本申请根据一示例性实施例示出的另一种控制无人机向目标停机区域移动的原理示意图。
图9A和图9B是本申请根据一示例性实施例示出的一种无人机在不同的高度所确定的目标停机区域的示意图。
图10是本申请根据一示例性实施例示出的一种无人机的传感器所采集的场景 图像的示意图。
图11是本申请根据一示例性实施例示出的一种获取标识分布信息的流程图。
图12是本申请根据一示例性实施例示出的一种对所识别的标识进行矫正的原理示意图。
图13是本申请根据一示例性实施例示出的另一种获取标识分布信息的流程图。
图14是本申请根据一示例性实施例示出的另一种无人机降落控制方法的流程图。
图15是本申请根据一示例性实施例示出的一种无人机向空闲停机位降落的流程图。
图16是本申请根据一示例性实施例示出的一种用于无人机的降落控制的控制装置的结构示意图。
具体实施方式
这里将详细地对示例性实施例进行说明，其示例表示在附图中。下面的描述涉及附图时，除非另有表示，不同附图中的相同数字表示相同或相似的要素。以下示例性实施例中所描述的实施方式并不代表与本公开相一致的所有实施方式。相反，它们仅是与如所附权利要求书中所详述的、本公开的一些方面相一致的装置和方法的例子。
在本公开使用的术语是仅仅出于描述特定实施例的目的，而非旨在限制本公开。在本公开说明书和所附权利要求书中所使用的单数形式的“一种”、“所述”和“该”也旨在包括多数形式，除非上下文清楚地表示其他含义。还应当理解，本文中使用的术语“和/或”是指并包含一个或多个相关联的列出项目的任何或所有可能组合。
应当理解，尽管在本公开可能采用术语第一、第二、第三等来描述各种信息，但这些信息不应限于这些术语。这些术语仅用来将同一类型的信息彼此区分开。例如，在不脱离本公开范围的情况下，第一信息也可以被称为第二信息，类似地，第二信息也可以被称为第一信息。取决于语境，如在此所使用的词语“如果”可以被解释成为“在……时”或“当……时”或“响应于确定”。
随着无人机硬件、软件等各方面技术的不断发展，无人机的应用也越来越广泛。其中，无人机编队飞行由于由多个无人机构成队列执行特定的飞行模式，具有场面宏大、可营造具有震慑感的氛围、为观看者提供视觉刺激，因此，无人机编队飞行也成 为了无人机的一个重要应用场景，例如，在节假日成为了一个常见的庆祝和娱乐表演节目等等。然而，无人机编队飞行的队列中的无人机数量往往成百甚至上千，如何进行高效且安全的降落控制管理仍是无人机智能控制领域的一个难题。
相关技术中，无人机飞行完毕后如若需要降落，对无人机的降落控制通常通过以下几种方式来实现。第一种方式是令无人机队列中的多个无人机随机无序降落，这种降落控制方式具有一定的碰撞风险，且降落后还需要对多个无人机一架一架地核对无人机编号，然后对号回收，不仅具有比较大的安全风险且回收效率也比较低下。第二种方式是令无人机队列中的多个无人机逐一排队降落，这样虽然能够消除一定的无人机碰撞的风险，但是却十分耗时。第三种方式是对无人机队列中的多个无人机再次编队，形成有序的队列，然后控制多个无人机基于全球定位信息同时降落，在这个过程中，由于基于全球定位系统的定位方式不够准确，误差通常在几十米的范围，故为了避免多个无人机之间的相互碰撞，对降落场地的尺寸以及平整度具有较高的要求，需要较大的停机场地且停机场地地面平整。
在另一些应用场景中，无人机执行特定的物流运输任务，也需要提供多个停机坪，以供多台无人机高降落至不同的停机位，完成既定的货物运输任务。提升无人机全场景工作的可靠性，提升起飞、降落过程中的控制性能，提升无人机在多停机位场景下的降落准确度，成了相关技术重点研究方向。
为了克服相关技术所存在的上述缺陷，本申请实施例提供了一种无人机降落控制方法。该方法可以应用于无人机队列中的多个无人机的降落控制，所述多个无人机可以是固定翼式无人机、直升机式无人机、扑翼式无人机、多旋翼式无人机等多种类型的无人机。该方法的应用场景，可以是无人机编队飞行执行特定的任务，例如无人机编队飞行表演等等，本申请实施例对此不作限制。
如图1所示，本申请实施例所提供的无人机降落控制方法，可以包括如下步骤：
步骤101，在无人机降落过程中，获取所述无人机的传感器采集的停机场的场景图像；
步骤102，在所述场景图像中识别所述停机位的标识，得到多个所述停机位的标识分布信息；
步骤103，基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位；
步骤104，控制所述无人机向所述目标停机位降落。
其中，步骤101中所述的停机场可以是用于停放无人机的专用场地，也可以是用于停放无人机的设备，例如无人机停机坪等等。所述停机场包括多个停机位，每个停机位用于停放无人机。多个所述停机位包括具有不同的标识的至少两个停机位，也就是多个停机位所具有的标识，这些标识至少有两种类型，结合图2给出说明。图2是包括多个停机位的停机场的示意图，其中，201表示停机场，停机场201中的每个椭圆形202代表一个停机位。在图2所给出的示意图中，每个停机位都具有标识，且标识有两种类型——标识1和标识2，此即为多个所述停机位包括具有不同的标识的至少两个停机位。当然，图2所述的停机场中的多个停机位以及下述的各个实施例仅以两种类型的标识为例进行说明，本领域技术人员应当理解，所述标识的类型还可以是两种以上。此外，除了如图2所示的全部停机位都具有标识，还可以是部分停机位没有标识，部分停机位具有一种或多种标识，没有标识作为一种特殊的标识类型，实现多个所述停机位包括具有不同的标识的至少两个停机位，本申请实施例对此不作限制。
在一些实施例中，所述标识的类型与所述停机位的外观属性关联。也就是，停机场所包含的多个停机位，自身就具有不同的外观属性，那么，可以将多个停机位的不同的外观属性，作为不同类型的标识。
在一些实施例中，所述标识的类型与所述停机位的以下一种或者多种外观属性关联：轮廓、颜色、尺寸、材质、标识符等等。即，所述停机场所包含的多个停机位，可以具有不同的轮廓、颜色、尺寸、材质以及标识符等外观属性，具有同一外观属性的停机位，为同种类型的标识。
例如，所述停机场所包含的多个停机位，为圆形轮廓的停机位和四边形轮廓的停机位，那么，圆形轮廓即为一种类型的标识，四边形轮廓为另外一种类型的标识；又例如，所述停机场所包含的多个停机位，一部分停机位上具有标识符“↑”，一部分停机位上具有标识符“←”，那么标识符“↑”为一种类型的标识，标识符“←”为另一种类型的标识；再例如，所述停机场所包含的多个停机位，为白色的停机位和黑色的停机位，那么，白色即为一种类型的标识，黑色为另外一种类型的标识。这里，仅以所述标识具有两种类型为例进行示例性说明，本领域技术人员应当理解，当所述标识具有多种类型，与两种类型的标识相似，本申请实施例在此不做赘述。
可选的，所述标识的类型可以与位于所述停机位上的标识符关联。所述标识符 可以是所述停机位出厂时就被设置在所述停机位上，如以打印的方式直接设置在所述停机位上，也可以是位于贴纸等独立的对象之上，当需要应用本申请实施例所述的无人机降落控制方法时，被安置在所述停机场的多个停机位上，本申请实施例对此不作限制。
所述标识的类型与之关联的标识符，可以具有不同的朝向、轮廓、颜色、尺寸、材质等等。如图3所示，以所述标识具有两种不同的类型为例，在图3A-图3E中分别给出了具有不同朝向、轮廓、颜色、尺寸、材质等等的标识符的示意图。在图3A中，所述标识符为具有方向指示性的箭头，所述箭头指向一种特定方向的标识符，即代表着一种类型的标识；在图3B中，所述标识符具有不同的轮廓，一种轮廓的标识符代表着一种类型的标识；在图3C中，所述标识符具有不同的颜色，同一种颜色的标识符即代表着一种类型的标识；在图3D中，所述标识符具有不同的尺寸，同前面类似，一种尺寸的标识符代表着一种类型的标识；在图3E中，以不同的灰度代表所述标识符具有不同的材质，例如，对于某些材质来说，反射率较弱，而对于另外一些材质来说，反射率较强，则可以采用反射率具有预设差值的不同材料所制作的标识符，各自代表一种类型的标识。
当然，本领域技术人员应当理解，上述的举例仅为示例性说明，并非枚举，多个所述停机位的不同类型的标识，还可以根据实际应用场景的特点，进行数量以及类型的确定，本申请实施例对此不作限制。
所述标识的类型，除了可以与所述停机位的外观属性关联之外，在一些实施例中，所述标识的类型还可以与所述停机位的周边的物体的外观属性关联。例如，可以在所述停机位的周边放置具有不同外观属性的物体，如不同颜色旗子，以所述停机位的周边的物体的不同的外观属性表征所述不同类型的标识，本申请实施例对此不作限制。
通过上述各个实施例可以看到，所述停机场所包括的停机位具有的不同标识，可以是各种类型的标识，所述标识的类型可以与所述停机位的外观属性关联，还可以与所述停机位的周边的物体的外观属性关联，容易实现不同类型的标识的设置，能够提高本申请实施例所提供的无人机的降落控制方法的灵活性以及适用范围，降低本申请实施例所述方法的实现难度。
在一些实施例中，步骤101，在无人机降落过程中，获取所述无人机的传感器采集的停机场的场景图像，所获取的可以是包含全部停机位的停机场的场景图像，也 可以是包含部分停机位的停机场的场景图像，本申请实施例对此不作限制。
故在一些实施例中，步骤102中所述的识别得到的所述标识分布信息可以包括所述停机场的多个所述停机位中全部停机位的标识，也可以是部分所述停机位的标识。仍结合图2进行说明，识别得到的所述标识分布信息可以是包含全部停机位的标识201，也可以是包含部分停机位的标识203。
通过上述实施例可以看到，不论在步骤101中所采集的停机场的场景图像，包含的是全部停机位，进而所识别得到的所述标识分布信息包括全部停机位的标识，还是在步骤101中所采集的停机场的场景图像，包含的是部分停机位，进而所识别得到的所述标识分布信息包含部分停机位的标识，都能够实现本申请实施例所述的无人机降落控制方法，从而提高本申请实施例所述方法的适用性以及灵活性，使得本申请实施例所述的方法易于实现，且具有较高的可靠性。
对于步骤102中，在所述场景图像中识别所述停机位的标识，得到多个所述停机位的标识分布信息，可以参考相关技术来实现，例如基于图像矩、图像特征识别等等，本申请实施例对所采用的识别方法不作限制。
在得到了多个所述停机位的标识分布信息之后，基于识别得到所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位，可以基于多种方式来实现。以下，以多个所述停机位的标识的类型包括两种，一种是朝左的箭头，另一种是朝上的箭头为例，介绍几种能够实现确定所述无人机的目标停机位的实施例。当然，本领域技术人员应当理解，这些实施例仅为示例性说明，还可以通过其他方式来实现在所述场景图像观测的多个停机位中确定所述无人机的目标停机位，本申请实施例对此不作限制。
在一些实施例中，本申请实施例所提供的无人机降落控制方法，还包括：获取所述目标停机位的所在区域的多个停机位的预设标识分布信息；
步骤103，所述基于识别得到的所述标识信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位，包括：
根据识别得到的所述标识分布信息和所述预设标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位。
结合图4对上述实施例进行说明。图4给出了基于所述预设标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位的原理示意图。在图 4中，401为所述目标停机位的所在区域的多个停机位的预设标识分布信息，402为目标停机位，403为基于步骤102的识别，所获取的所述场景图像所包含的多个无人机的标识分布信息。
在上述实施例中，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位之前，可以先获取所述预设标识分布信息401。所述预设标识分布信息的获取时机，本申请实施例不作限制，可以是在所述无人机起飞前就获取所述预设标识分布信息，也可以是在所述无人机在飞行过程中获取由控制中心所发送的所述预设标识分布信息，还可以是在所述无人机降落之前，从其他设备或者系统，基于各种通信方式实现所述预设标识分布信息的获取。所述预设标识分布信息401，可以包含着所述停机场的全部停机位的标识分布，也可以包含着所述停机场的部分停机位的标识分布，本申请实施例对此不作限制。
由于该预设标识分布信息包含着所述目标停机位402以及其周围的多个停机位的标识分布，故相当于一个准确的“地图”，能够作为基准信息。将基于步骤102的识别，所获得的多个所述停机位的标识分布信息403与所述预设标识分布信息401进行比较，能够定位出所述无人机当前的位置与所述目标停机位的相对位置，进而能够基于所述预设标识分布信息的指引，执行步骤104，控制所述无人机向所述目标停机位降落。
从上述实施例可以看到，通过获取所述目标停机位的所在区域的多个所述停机位的预设标识分布信息，根据识别得到的所述标识分布信息和所述预设标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位，不仅能够定位出所述无人机当前的位置与所述目标停机位的相对位置，而且具有能够快速定位、准确定位的优点，进而能够提高无人机队列的多个无人机的降落效率、降低对所需的降落场地的要求，实现高效安全的降落控制。
在上述实施例中，所述预设标识分布信息以及所述识别得到的所述标识分布信息，可以以多种形式来表示。
在一些实施例中，所述预设标识分布信息以及所述识别得到的所述标识分布信息，可以是以所述标识关联的所述停机位的外观属性和\或所述停机位的周边的物体的外观属性、位于多个所述停机位上的标识符的外观属性的形态来表示。例如，所述标识关联的是位于多个所述停机位上的标识符——朝左的箭头以及朝上的箭头，那么，所述预设标识分布信息和所述识别得到的所述标识分布信息可以分别如图5A的501 和502所示。
在一些实施例中，所述预设标识分布信息可以包括所述亭机场中所有停机位的标识分布信息，即，一个表征全局停机位情况的全局标识分布信息。具体的，可以通过无人机的观感器采集所述无人机下方的场景图像。从所述场景图像中识别得到多个停机位的标识分布信息。其中，这个识别得到的标识分布信息可以以编码信息的形式表达。进一步的，将识别得到的标识分布信息和前述全局标识分布信息进行比较，计算出当前识别得到的标识分布信息属于全局标识分布信息的哪一块区域，换句话说，基于比对的结果，可以推导知道当前无人机观测的场景属于停机场的全局区域中的哪个位置。
基于这个位置，可以知道无人机与停机场中目标停机位或者目标停机区域的相对位置关系，进而，基于相对位置关系，控制无人机调整空中的位姿，逐步向目标停机位或者目标停机区域靠近。
当然，为了节省存储以及计算资源等等，在一些实施例中，所述预设标识分布信息和所述识别得到的所述标识分布信息也可以根据标识的类型数量，以二进制、八进制、十进制等等进制数来表示。例如，所述标识的类型为两种，则可以以二进制数0和1分别表示这两种类型的标识，如图5B所示，其中，“0”代表朝左的箭头，“1”代表朝上的箭头；又例如，所述标识的类型为四种，可以以00、01、10、00分别表示这四种类型的标识。当然，本领域技术人员应当理解，所述预设标识分布信息以及所述识别得到的所述标识分布信息的表现形式，可以基于所述标识类型的数量以及实际应用场景来确定，本申请实施例对此不作限制。
通过上述实施例可以看到，以二进制数、八进制、十进制等进制数的形式来表示所述预设标识分布信息和所述识别得到的所述标识分布信息，能够节省一定的计算以及存储资源。且将所述识别得到的所述标识分布信息与所述预设标识分布信息进行比对，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位时，基于进制数的比较，错误率低，能够提高本申请实施例所述的无人机降落控制方法的准确率。
在一些实施例中，步骤103，所述基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位，包括：
基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定目标停机位，其中，所述目标停机位周围预设方位的停机位的所述标识分布符合预设 的标识分布条件。
结合图6对上述实施例所述的在所述场景图像观测的多个停机位中确定所述无人机的目标停机位进行说明。图6为通过识别所采集的场景图像，得到的多个所述停机位的标识分布信息601，其中，602为目标停机位。在该实施例中，可以不获取所述预设标识分布信息，而是基于预设的标识分布条件，确定所述目标停机位。以图6为例，假设对于所述无人机，其对应的预设的标识分布条件为：该无人机所对应的目标停机位，为朝左的箭头，与其相邻的左上角位置处的标识和与其相邻的左下角位置处的标识与自己相同，与其相邻的右上角位置处的标识和与其相邻的右下角位置处的标识与自己不同。那么，基于该预设的标识分布条件，则可以唯一确定出该无人机所对应的目标停机位处于602处。对于无人机队列中的多个无人机而言，只要每个无人机所对应的预设的标识分布条件不同，就能够为所述多个无人机唯一定位每个无人机所对应的目标停机位。
当然，本领域技术人员应当理解，上述举例仅为示例性说明。当需要控制降落的无人机队列中的无人机的数量较多时，每个无人机所对应的预设的标识分布条件可以更复杂，例如以所述目标停机位周围的更多个的位置的标识来限定所述目标停机位，即以所述目标停机位为中心，以其周围第一层、第二层甚至第三层等等的多个位置的标识来限定所述目标停机位，和\或，以该目标停机位是否位于所述停机场的边界以及位于哪个方位的边界来限定所述目标停机位，和\或采用更多类型的标识来指示不同位置的停机位，进而使得所述目标停机位的周围的多个停机位的标识具有更复杂的组合情况来限定所述目标停机位等等，本申请实施例对此不作限制。
在上述实施例中，基于预设的标识分布条件，从而在所述场景图像观测的多个停机位中确定目标停机位，不需要获取所述目标停机位的所在区域的多个所述停机位的预设标识分布信息，就能确定出所述目标停机位。基于所确定的目标停机位在所述场景图像中的位置，能够获得所述无人机当前位置与所述目标停机位的相对位置，进而能够基于包含所述目标停机位的场景图像的指引，执行步骤104，控制所述无人机向所述目标停机位降落。
从上述实施例可以看到，通过识别得到的所述标识分布信息和所述预设的标识分布条件，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位，不仅能够定位出所述无人机当前的位置与所述目标停机位的相对位置，而且不需要获取所述目标停机位的所在区域的多个所述停机位的预设标识分布信息，具有能够快速定位、 准确定位的优点，进而能够提高无人机队列的多个无人机的降落效率、降低对所需的降落场地的要求，实现高效安全的降落控制。
此外，本领域技术人员应当理解，上述在所述场景图像观测的多个停机位中确定所述无人机的目标停机位，可以基于所获取的所述目标停机位的所在区域的多个所述停机位的预设标识分布信息来实现，也可以仅基于所述的预设的标识分布条件来实现，当然，也可以基于其他方法，基于所识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位。此外，确定所述目标停机位，可以仅基于一种方法，也可以同时采用上述各个方案，以避免所述预设标识分布信息或者所述预设的标识分布条件等等的丢失或者错误，导致本申请实施例所提供的无人机降落方法无法执行，提高本申请实施例所提供的方法的可靠性以及稳健性。
在一些实施例中，步骤104，控制所述无人机向所述目标停机位降落，如图7所示，包括：
步骤701，根据所述场景图像中确定所述无人机与所述目标停机位的相对位置关系；
步骤702，基于所述相对位置关系，控制所述无人机向所述目标停机位降落。
分别以图4和图6为例进行说明。在图4中，前文已经介绍，401为预设标识分布信息，即包含着多个停机位的所述停机场的“地图”；403对应着401中的虚线框部分，即所述无人机当前的位置；而402则为所述无人机所对应的目标停机位。那么，由所述无人机当前的位置403和所述目标停机位402在“地图”401中的相对位置关系，能够确定应该控制所述无人机沿与所述“朝上箭头”箭尾方向运动，从而运动至所述目标停机位的上空降落至所述目标停机位上。
在图6中，601为所述场景图像所对应的标识分布信息，602为目标停机位。基于该场景图像所对应的标识分布信息，可知所述无人机当前的位置与所述目标停机位的相对位置关系。此外，如果所述无人机当前位于所述目标停机位的正上方，那么，所述目标停机位应当处于所述场景图像的正中央。故基于该条件，可知，应该控制所述无人机沿与所述“朝上箭头”箭尾方向运动，从而运动至所述目标停机位的上空降落至所述目标停机位上。
通过上述实施例可以看到，根据所述相对位置关系，能够实现快速、简单且准确地控制所述无人机向所述目标停机位降落。当然，本领域技术人员应当理解，除了 基于所述相对位置关系，控制所述无人机向所述目标停机位降落，也可以基于所述停机位的真实尺寸、所述停机位周围的物体的真实尺寸以及位于所述停机位上的标识符的真实尺寸等等，结合所识别的标识或者所述场景图像，确定所识别的标识或者所述场景图像与真实空间尺寸的比例关系，进而确定所述无人机当前的位置与所述目标停机位的真实距离，从而更加准确地控制所述无人机向所述目标停机位降落，本申请实施例对此不作限制。
在一些实施例中，本申请实施例所提供的无人机降落控制方法，还包括：基于识别得到的所述标识分布信息，在所述场景图像观测的所述停机场中确定目标停机区域；控制所述无人机向所述目标停机区域移动。
仍结合图4和图6进行说明。由于所述无人机在降落初始状态，可能并非位于所述目标停机位的正上方。因此，在控制所述无人机向所述目标停机位降落之前，可以先基于识别得到的所述标识分布区域，确定目标停机区域，所述目标停机区域为包含多个停机位标识的区域，控制所述无人机向所述目标停机区域移动。所述目标停机区域的尺寸可以基于实际应用需要预先设定，例如，可以是以所述目标停机位中心的包含N*M个所述标识的区域，M和N为正整数，M和N可以相同，也可以不同。
此外，在一些实施例中，在所述无人机向所述目标停机区域不断移动的过程中，可以保持所述目标停机区域的尺寸，直至在所述目标停机区域中准确定位到所述目标停机位，控制所述无人机精准降落至所述目标停机位上。即如图8A所示，所述无人机在向所述目标停机位(黑色箭头所指的位置)降落的过程中，依次向相同尺寸的目标停机区域802、803和804移动，直至在目标停机区域803中准确定位到位于所述目标停机区域803中心的所述目标停机位，然后向所述目标停机位降落。
当然，在一些实施例，在所述无人机向所述目标停机区域不断移动的过程中，可以逐渐缩小所述目标停机区域的尺寸，直至所述目标停机区域仅包括所述目标停机位，进而实现控制所述无人机精准降落至所述目标停机位上。即如图8B所示，所述无人机在向所述目标停机位(黑色箭头所指的位置)降落的过程中，依次向尺寸逐渐减小的目标停机区域802、803以及804移动，直至向所述目标停机位804降落。
在上述实施例中，控制所述无人机向所述目标停机区域移动，可以是所述无人机保持飞行高度不变，仅在平行于水平面的平面上发生位置变化，实现至所述目标停机区域的移动；也可以是所述无人机既有飞行高度的变化，也在平行于水平面的平面上发生位置变化，本申请实施例对此不作限制。
此外，在上述实施例中，步骤701，基于识别得到的所述标识分布，在所述场景图像观测的所述停机场中确定目标停机区域，可以是基于所采集的场景图像，进行标识的再次识别，从而确定所述目标停机区域，也可以是利用步骤102中对所述场景图像的识别结果，确定所述目标停机区域，本申请实施例对此也不做限制。
通过上述实施例可以看到，在本申请实施例所提供的无人机降落控制方法中，基于对所述场景图像所包含的标识进行识别的结果，确定所述目标停机位所在的目标停机区域和所述目标停机位，先控制所述无人机向所述目标停机区域移动，再控制所述无人机向所述目标停机位降落，即先控制所述无人机向一个较大尺寸的目标区域(即目标停机区域)移动，然后再控制所述无人机向目标停机位降落，能够合理控制降落区域的精度，符合实际降落的精度需求，易于实现，且节省计算资源。
在一些实施例中，本申请实施例所述的无人机降落控制方法，还包括：
获取所述无人机的飞行高度；
步骤103，所述基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位；步骤104，所述控制所述无人机向所述目标停机位降落，包括：
当所述无人机处于第一飞行高度时，基于识别得到的所述标识分布信息，在所述场景图像观测的所述停机场中确定目标停机区域，控制所述无人机向所述目标停机区域移动；
所述无人机降落控制方法还包括：
当所述无人机处于第二飞行高度时，基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位；
控制所述无人机向所述目标停机位降落；
其中，所述第二飞行高度小于所述第一飞行高度。
在上述实施例中，不仅获取所述无人机的传感器所采集的停机场的场景图像，还获取所述无人机的飞行高度，基于对所述场景图像所包含的多个停机位的标识的识别，确定所述无人机在不同飞行高度所对应的目标停机区域或目标停机位，指引所述无人机的降落控制。
结合图9A对上述实施例进行说明。图9A中，901为对所述无人机的传感器采 集的停机场的场景图像进行识别，所获得的多个所述停机位的标识分布信息，902为所述无人机的目标停机位。当确定所述无人机处于较高的飞行高度时，可以基于对所述场景图像的识别，得到所述多个停机位的标识分布信息，进而确定所述无人机当前的位置.结合所述目标停机位的位置，能够确定所述无人机当前的位置与所述目标停机位的相对位置关系，从而可以指引所述无人机向所述目标停机位所在的目标停机区域移动。随着无人机的降落，当所述无人机处于较低的飞行高度时，能够基于对所述场景图像的识别而获得的多个所述停机位的标识分布信息，更加准确地定位出所述目标停机位的位置，故当所述无人机处于第二飞行高度时，可以控制所述无人机向所确定的目标停机位降落。
当然，本领域技术人员应当理解，所述第一飞行高度可以包括多个飞行子高度，在每个飞行子高度，在所述场景图像观测的所述停机场中确定对应的子目标停机区域，控制所述无人机向所述子目标停机区域移动。以所述第一飞行高度包括第一子飞行高度和第二子飞行高度且第一子飞行高度大于第二子飞行高度为例，结合图9A和图9B进行说明。当所述无人机处于第一子飞行高度时，基于识别得到的所述标识分布信息901，如图9A所示，在所述场景图像观测的所述停机场中确定第一子目标停机区域903，控制所述无人机向所述第一子目标停机区域903移动；当所述无人机处于第二子飞行高度时，基于识别得到的所述标识分布信息901，如图9B所示，在所述场景图像观测的所述停机场中确定第二子目标停机区域904，控制所述无人机向所述第二子目标停机区域904移动。其中，所述子目标停机区域可以随着其对应的子飞行高度的降低而减小，进而能够排除更多的非目标停机位所在的区域，逐步准确地确定出所述目标停机位所在的位置。
在上述实施例中，在不同的飞行高度，用于得到所述标识分布信息的所述场景图像，可以是在所述无人机在接收到降落信号后，在预设飞行高度所获取的包含多个停机位的场景图像。当然，在一些实施例中，所述获取所述无人机的传感器采集的停机场的场景图像，可以包括：获取所述无人机在不同飞行高度采集的所述场景图像。
当所述无人机处于不同的飞行高度时，获取与所述不同的飞行高度所对应的包含多个停机位的场景图像，基于对所述场景图像所包含的标识的识别，可以获得与所述无人机的不同飞行高度所对应的标识分布信息。例如，当所述无人机分别位于高于第一阈值的高空、在第一阈值与第二阈值之间的中空以及低于第二阈值的近地面时，分别采集包含多个停机位的停机场的场景图像。通过对所述场景图像进行识别，能够 得到与所述飞行高度所对应的标识分布信息，确定出所述无人机在不同飞行高度与所述目标停机位的相对位置关系。由于本申请实施例所提供的无人机降落控制方法，是基于对所采集的场景图像所包含的标识进行识别，在多个停机位中确定目标停机区域以及目标停机位的，故基于本申请实施例所提供的方法，对目标停机区域以及目标停机位的定位误差，远远小于相关技术中采用GPS等定位装置所带来的定位误差，能够比相关技术所采用的无人机降落控制方法实现更为精准的降落控制。
此外，由于当所述无人机在较高飞行高度时，距离所述包含多个停机位的停机场较远，故基于对较高的飞行高度所对应的场景图像的识别而确定的目标停机区域的误差会比较低的飞行高度所确定的目标停机区域的误差稍大，随着飞行高度的不断降低，基于所述场景图像的识别而确定的目标停机区域的误差会越来越小，直至准确定位出所述目标停机位。相应地，控制所述无人机向所述目标停机区域移动以及控制所述无人机向所述目标停机位降落，能够随着飞行高度的降低不断精确，最终控制所述无人机精准地降落至所述目标停机位上。以图9A和图9B为例，当901代表所述停机场，903、904以及902代表着在不同高度所采集的场景图像，且所对应的无人机的飞行高度逐渐减低，那么，最终能够准确地使所述无人机降落至所述目标停机位902上。
通过上述实施例可以看到，通过在不同的飞行高度获取不同的场景图像，进而使得所得到的标识分布信息逐步更为精准，能够使得所确定的目标停机区域以及目标停机位逐步更为精准，最终实现所述无人机的精准降落。该方法能够应用于无人机队列中的多个无人机的同时降落，进而可以克服相关技术中，无人机的降落耗时长，或者需要较大的降落场地等的缺陷，实现对无人机队列中的多个无人机的高效、安全的降落控制。
值得说明的是，应对有多台无人机预期同时分别降落在停机场的不同停机位，本实施例的方案能起到较优的控制效果。若每台无人机均实施本实施例的方案，那么在降落的过程中，每台无人机均可以较为准确的调整至自身对应的目标停机位或者目标停机区域上空，随着无人机逐渐降落，可以基于本方案实现更为精确地降落控制。减少多台无人机分别降落不同停机位时可能出现的位姿干涉情况，减少无人机之间互相碰撞的可能性，提升无人机的降落过程的安全性。
在某些情况下，所述无人机可能并没有位于所述包含多个停机位的停机场的上方，故无法直接获取所述包含多个停机位的停机场的场景图像，故在一些实施例中，所述获取无人机的传感器采集的停机场的场景图像，可以包括：
获取无人机的定位；
当所述无人机的定位处于所述停机场对应的区域，获取所述场景图像。
所述获取无人机的定位，可以参考相关技术来实现。例如，可以基于所述无人机的全球定位系统(Global Positioning System，GPS)，也可以基于视觉惯性里程计(Visual-Inertial Odometry，VIO)，当然，还可以基于其他方式，来获取所述无人机的定位。当基于所获取的定位，确定所述无人机并未处于所述停机场对应的区域时，可以在所述无人机的定位信息的指引下，先控制所述无人机移动至所述停机场对应的区域，然后再获取所述场景图像。当然，当所述无人机未处于所述停机场对应的区域，也可以基于其他方式，先使得所述无人机处于所述停机场对应的区域，然后再获取所述场景图像，本申请实施例对此也不做限制。
通过上述实施例可以看到，通过获取所述无人机的定位，当所述无人机的定位处于所述停机场对应的区域，获取所述场景图像，能够避免无效图像的采集，节省无人机的存储资源以及计算资源。
在某些情况下，步骤101中，所获取的无人机的传感器所采集的停机场的场景图像，可能会与真实的停机场的场景出现不一致的情况，所述不一致的情况包括所述场景图像相对于所述真实的停机场的场景存在旋转变换。结合图6和图10进行说明，当真实的停机场的场景中，所述标识包括朝左的箭头和朝上的箭头，标识的分布情况如图6所示。步骤101中，所获取的无人机的传感器所采集的停机场的场景图像可能如图10所示，即所述标识分布信息中，所述标识为朝左上角的箭头和朝左下角的箭头。出现这种不一致的原因在于，所述传感器采集场景图像的朝向与所述停机场的停机位的放置朝向不一致。为了能够克服这种不一致情况，实现前文各个实施例所述的无人机降落控制方法，可以在所述无人机的传感器采集所述停机场的场景图像时，调整所述无人机整体或者所述传感器的朝向，或者，调整所采集的场景图像的朝向，当然，也可以调整所识别的标识的朝向等等。
故，在一些实施例中，步骤101，所述获取所述无人机的传感器采集的停机场的场景图像，如图11所示，包括：
步骤1101，获取拍摄所述场景图像时所述无人机的朝向信息；
步骤1102，基于所述朝向信息矫正所述场景图像；
步骤1103，在矫正后的所述场景图像中识别所述停机位的标识，得到多个所述 停机位的标识分布信息。
其中，所述无人机的朝向信息，可以基于所述无人机的方位定位装置来获取，例如，电子罗盘等等，本申请实施例对此不作限制。在确定了所述无人机的朝向信息之后，可以对所述场景图像进行调整，使得所述场景图像的朝向与所述停机场的真实场景中的标识朝向一致，得到准确的所述标识分布信息，能够保证对所述无人机降落控制的准确性。
当然，在某些情况下，也可以仅基于所述场景图像所包含的标识，来对所述场景图像进行校正，进而得到准确的所述标识分布信息。如图12所示，当所述标识为两种类型，一种是朝左的箭头，另一种是朝上的箭头。当所述传感器采集的场景图像如图12的左图所示，由于所述标识仅有上述两种情况，那么，可以唯一确定，需要对所述场景图像旋转一定的角度，使得如图12所示的第一行第一列的所述标识为朝上的箭头，即可对所述场景图像进行朝向矫正，矫正后的所述场景图像如图12的右图所示。当然，还可以通过设置其他的标识，使得能够仅基于所述场景图像的标识，即可对所述场景图像进行矫正，从而得到准确的所述标识分布信息，本申请实施例对此不作限制。
在一些实施例中，步骤101，所述获取所述无人机的传感器采集的停机场的场景图像，如图13所示，包括：
步骤1301，基于预设的朝向信息，调整所述无人机的所述传感器的朝向；
步骤1302，基于调整后的所述传感器采集所述场景图像。
其中，所述的预设的朝向信息，可以是与所述停机场的真实场景完全一致的朝向，也可以是与所述停机场的真实场景具有已知角度差的朝向，本申请实施例对此不作限制。所述调整所述无人机的所述传感器的朝向，可以是基于所述无人机的所述无人机的方位定位装置来确定，例如，电子罗盘等等，当然，也可以基于调整后的所述传感器所采集的场景图像中的标识的朝向来确定，本申请实施例对此也不做限制。当然，本领域技术人员应当理解，也可以是对所述无人机整体进行朝向调整。
通过上述实施例可以看到，通过基于预设的朝向信息，调整所述无人机的所述传感器或者所述无人机整体的朝向，使得所采集的所述场景图像的朝向与所述停机场的真实场景中的标识朝向一致，得到准确的所述标识分布信息，能够保证对所述无人机降落控制的准确性。
通过上述各个实施例可以看到，基于本申请所提供的无人机降落控制方法，通过在无人机降落过程中，获取包含多个停机位的停机场的场景图像，然后基于对所述场景图像的识别而得到的多个停机位的标识分布信息，确定所述无人机所对应的目标停机位，进而控制所述无人机向所述目标停机位降落，能够根据所采集的场景图像中的多个停机位的标识分布信息，确定无人机所对应的目标停机位，可应用于每个无人机上，使得每个无人机能够自动准确定位所对应的目标停机位，从而能够允许多个无人机同时自动降落至各自对应的目标停机位上且保证无人机降落的安全性，克服相关技术所存在的耗时、效率低下等问题。
此外，本申请实施例还提供了另一种无人机降落控制方法，如图14所示，所述方法包括：
步骤1401，在无人机降落过程中，获取所述无人机的传感器采集的停机场的场景图像，所述停机场包括多个停机位，多个所述停机位包括具有不同的标识的至少两个停机位；
步骤1402，在所述场景图像中识别所述停机位的标识，得到多个所述停机位的标识分布信息；
步骤1403，基于识别得到的所述标识分布信息，在所述场景图像观测的所述停机场中确定目标停机区域；
步骤1404，控制所述无人机向所述目标停机区域移动。
在一些实施例中，所述目标停机区域为所述无人机的目标停机位。
在一些实施例中，所述目标停机区域为包含空闲停机位的区域。
在一些实施例中，如图15所示，所述方法还包括：
步骤1501，确定所述目标停机区域中的空闲停机位；
步骤1502，控制所述无人机向任意所述空闲停机位降落。
其中，确定所述目标停机区域中的空间停机位，可以参考相关技术来实现。例如，可以基于红外传感器，或者基于图像特征识别等方法，来确定所述目标停机区域中的空闲停机位。
在上述实施例中，通过获取所述场景图像，并对所述场景图像进行识别，得到所述标识分布信息，从而确定所述目标停机区域，控制所述无人机向所述目标停机区 域移动，能够实现基于所获取的场景图像，准确控制所述无人机向所述目标停机区域移动，进而实现所述无人机的准确降落。
在一些实施例中，所述识别得到的所述标识分布信息包括所述停机场的多个所述停机位中全部或者部分所述停机位的标识。
在一些实施例中，所述方法还包括：获取所述目标停机区域的多个所述停机位的预设标识分布信息；所述基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机区域，包括：根据识别得到的所述标识分布信息和所述预设标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机区域。
在一些实施例中，所述基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机区域，包括：基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定目标停机区域，其中，所述目标停机区域周围预设方位的停机位的所述标识分布信息符合预设的标识分布条件。
在一些实施例中，所述方法还包括：获取所述无人机的高度；所述基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位；所述控制所述无人机向所述目标停机区域移动，包括：当所述无人机处于第一高度时，基于识别得到的所述标识分布信息，在所述场景图像观测的所述停机场中确定目标停机区域，控制所述无人机向所述目标停机区域移动；所述方法还包括：当所述无人机处于第二高度时，基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位；控制所述无人机向所述目标停机位降落；其中，所述第二高度小于所述第一高度。
在一些实施例中，所述获取所述无人机的传感器采集的停机场的场景图像，包括：获取所述无人机在不同高度采集的所述场景图像。
在一些实施例中，所述控制所述无人机向所述目标停机位降落，包括：根据所述场景图像中确定所述无人机与所述目标停机区域的相对位置关系；基于所述相对位置关系，控制所述无人机向所述目标停机区域移动。
在一些实施例中，所述获取所述无人机的传感器采集的停机场的场景图像包括：获取无人机的定位；当所述无人机的定位处于所述停机场对应的区域，获取所述场景图像。
在一些实施例中，所述标识的类型与所述停机位的外观属性关联；和\或，所述标识的类型与所述停机位的周边的物体的外观属性关联。
在一些实施例中，所述标识的类型与所述停机位的以下一种或者多种外观属性关联：轮廓、颜色、尺寸、材质、标识符。
在一些实施例中，所述获取所述无人机的传感器采集的停机场的场景图像，包括：获取拍摄所述场景图像时所述无人机的朝向信息；基于所述朝向信息矫正所述场景图像；在矫正后的所述场景图像中识别所述停机位的标识，得到多个所述停机位的标识分布信息。
在一些实施例中，所述获取所述无人机的传感器采集的停机场的场景图像，包括：基于预设的朝向信息，调整所述无人机的所述传感器的朝向；基于调整朝向后的所述传感器采集所述场景图像。
上述各个实施例的相关内容，同本申请所提供的第一种无人机降落控制方法的相关内容类似，在前文中已经详细介绍，可以参考前文所述各个实施例，本申请在此不做赘述。
与前述所提供的方法的各个实施例相对应，本申请实施例还提供了一种控制装置，所述控制装置用于无人机的降落控制，如图16所示，所述控制装置包括存储器1601和处理器1602及存储在存储器1601上并可在处理器1602上运行的计算机程序，所述处理器执行所述程序时实现前文所述的任一方法实施例。该电子设备中的处理器1602可以是一个或多个，图16中以一个处理器1602为例，所述处理器可以包括GPU和CPU。所述电子设备中的处理器1602和存储器1601可以通过通信总线或其他方式连接，图16中以通过通信总线连接为例。
本实施例中电子设备的处理器1602中集成了上述实施例提供的所述应用程序。此外，该电子设备中的存储器1601作为一种计算机可读存储介质，可用于存储一个或多个程序，所述程序可以是软件程序、计算机可执行程序以及模块。处理器1602通过运行存储在存储器1601中的软件程序、指令以及模块，从而执行设备的各种功能应用以及数据处理，即实现上述各个方法实施例。
存储器1601可包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需的应用程序；存储数据区可存储根据设备的使用所创建的数据等。此外，存储器1601可以包括高速随机存取存储器，还可以包括非易失性存储器， 例如至少一个磁盘存储器件、闪存器件、或其他非易失性固态存储器件。在一些实例中，存储器1601可进一步包括相对于处理器1602远程设置的存储器，这些远程存储器可以通过网络连接至设备。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
处理器1602通过运行存储在存储器1601中的程序，从而执行各种功能应用以及数据处理，实现本申请实施例提供的方法。
通过上述各个实施例可以看到，基于本申请实施例所提供的一种控制装置，能够实现在无人机降落过程中，获取包含多个停机位的停机场的场景图像，然后基于对所述场景图像的识别而得到的多个停机位的标识分布信息，确定所述无人机所对应的目标停机位，进而控制所述无人机向所述目标停机位降落。可以看到，本申请实施例所提供的控制装置，能够根据所采集的场景图像中的多个停机位的标识分布信息，确定无人机所对应的目标停机位，可应用于每个无人机上，使得每个无人机能够自动准确定位所对应的目标停机位，从而能够允许多个无人机同时自动降落至各自对应的目标停机位上且保证无人机降落的安全性，克服相关技术所存在的耗时、效率低下等问题。
此外，本申请实施例还提供了一种无人机，所述无人机包括用于采集停机场的场景图像的传感器以及前文任一实施例所述的控制装置。
通过上述各个实施例可以看到，基于本申请实施例所提供的一种无人机，能够实现在无人机降落过程中，获取包含多个停机位的停机场的场景图像，然后基于对所述场景图像的识别而得到的多个停机位的标识分布信息，确定所述无人机所对应的目标停机位，进而控制所述无人机向所述目标停机位降落。可以看到，本申请实施例所提供的无人机，能够根据所采集的场景图像中的多个停机位的标识分布信息，确定无人机所对应的目标停机位，能够允许多个所述无人机同时自动降落至各自对应的目标停机位上且保证无人机降落的安全性，克服相关技术所存在的耗时、效率低下等问题。
与所述无人机相对应，本申请实施例还提供了一种用于停放无人机的设备，所述设备包括多个停机位，多个所述停机位包括具有不同的标识的至少两个停机位，所述多个停机位的标识分布信息可用于定位每个所述停机位。
在一些实施例中，所述标识的类型与所述停机位的外观属性关联；和\或，所述标识的类型与所述停机位的周边的物体的外观属性关联。
在一些实施例中，所述标识的类型与所述停机位的以下一种或者多种外观属性关联：轮廓、颜色、尺寸、材质、标识符。
上述所述用于停放无人机的设备的各个实施例的相关内容，在前文所提供的各个方法实施例中已经详细介绍，本申请实施例在此不再赘述。
在一些实施例中，所述设备包括传感器以及处理器，所述处理器用于基于所述传感器所采集的信息，执行预设的程序。
在一些实施例中，所述设备具备可折叠性，所述传感器用于当检测到无人机降落至预设的停机位上时通知所述处理器；所述处理器用于控制所述设备折叠，以对所述无人机打包。
其中，所述停机位可以设置有停机平台和收纳箱体，待所述无人机降落于所述停机平台时，所述停机平台可以执行特定的机械动作，以将所述无人机收纳进入所述收纳箱体。
在一些实施例中，所述预设的程序包括以下一种或者多种操作程序：给所述无人机充电；装载货物至所述无人机；卸除所述无人机的货物；清洁所述无人机；更换所述无人机的部分零部件；维修所述无人机的零部件等等。
通过上述实施例可以看到，基于本申请实施例所以供的设备，能够为所述无人机的快速降落控制以及快速打包提供条件，从而有助于对所述无人机实现安全高效的降落控制以及快速收纳，节省用户的时间成本以及人力成本。
此外，本申请实施例还提供一种无人机降落控制系统，所述系统包括前文任一实施例所述的无人机以及所述用于停放无人机的设备，相关内容在前文中已经详述，在此，本申请不做赘述。
通过上述实施例可以看到，基于本申请实施例所提供的一种无人机降落控制系统，能够实现在无人机降落过程中，获取包含多个停机位的停机场的场景图像，然后基于对所述场景图像的识别而得到的多个停机位的标识分布信息，确定所述无人机所对应的目标停机位，进而控制所述无人机向所述目标停机位降落。可以看到，本申请实施例所提供的无人机，能够根据所采集的场景图像中的多个停机位的标识分布信息，确定无人机所对应的目标停机位，能够允许多个所述无人机同时自动降落至各自对应的目标停机位上且保证无人机降落的安全性，克服相关技术所存在的耗时、效率低下等问题。
此外，本申请还提供了一种计算机可读存储介质，所述存储介质存储有计算机程序，所述计算机程序被处理器执行时实现前文所述的任一方法。
所述计算机可读介质可以是计算机可读信号介质或者计算机可读存储介质。计算机可读存储介质例如可以是但不限于电、磁、光、电磁、红外线、或半导体的系统、装置或器件，或者任意以上的组合。计算机可读存储介质的更具体的例子(非穷举的列表)包括：具有一个或多个导线的电连接、便携式计算机磁盘、硬盘、随机存取存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、光纤、便携式紧凑磁盘只读存储器(CDROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。在本文件中，计算机可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行系统、装置或者器件使用或者与其结合使用。
所述计算机可读存储介质的信号介质可以包括在基带中或者作为载波一部分传播的数据信号，其中承载了计算机可读的程序代码。这种传播的数据信号可以采用多种形式，包括但不限于电磁信号、光信号或上述的任意合适的组合。计算机可读的信号介质还可以是计算机可读存储介质以外的任何计算机可读介质，该计算机可读介质可以发送、传播或者传输用于由指令执行系统、装置或者器件使用或者与其结合使用的程序。
计算机可读介质上包含的程序代码可以用任何适当的介质传输，包括但不限于无线、电线、光缆、RF等等，或者上述的任意合适的组合。可以以一种或多种程序设计语言或其组合来编写用于执行本申请操作的计算机程序代码，所述程序设计语言包括面向对象的程序设计语言—诸如Java、Smalltalk、C++，还包括常规的过程式程序设计语言—诸如”C”语言或类似的程序设计语言。程序代码可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中，远程计算机可以通过任意种类的网络，包括局域网(LAN)或广域网(WAN)，连接到用户计算机，或者，可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。
本领域技术人员在考虑申请及实践这里申请的发明后，将容易想到本申请的其它实施方案。本申请旨在涵盖本申请的任何变型、用途或者适应性变化，这些变型、用途或者适应性变化遵循本申请的一般性原理并包括本申请未申请的本技术领域中的公知常识或惯用技术手段。说明书和实施例仅被视为示例性的，本申请的真正范围和 精神由下面的权利要求指出。
应当理解的是，本申请并不局限于上面已经描述并在附图中示出的精确结构，并且可以在不脱离其范围进行各种修改和改变。本申请的范围仅由所附的权利要求来限制。
以上所述仅为本申请的较佳实施例而已，并不用以限制本申请，凡在本申请的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本申请保护的范围之内。

Claims (24)

  1. 一种无人机降落控制方法，其特征在于，包括：
    在无人机降落过程中，获取所述无人机的传感器采集的停机场的场景图像，所述停机场包括多个停机位，多个所述停机位包括具有不同的标识的至少两个停机位；
    在所述场景图像中识别所述停机位的标识，得到多个所述停机位的标识分布信息；
    基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位；
    控制所述无人机向所述目标停机位降落。
  2. 根据权利要求1所述的方法，其特征在于，所述识别得到的所述标识分布信息包括所述停机场的多个所述停机位中全部或者部分所述停机位的标识。
  3. 根据权利要求1所述的方法，其特征在于，所述方法还包括：
    获取所述目标停机位的所在区域的多个所述停机位的预设标识分布信息；
    所述基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位，包括：
    根据识别得到的所述标识分布信息和所述预设标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位。
  4. 根据权利要求1所述的方法，其特征在于，所述基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位，包括：
    基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定目标停机位，其中，所述目标停机位周围预设方位的停机位的所述标识分布信息符合预设的标识分布条件。
  5. 根据权利要求1所述的方法，其特征在于，所述方法还包括：
    基于识别得到的所述标识分布信息，在所述场景图像观测的所述停机场中确定目标停机区域；
    控制所述无人机向所述目标停机区域移动。
  6. 根据权利要求1所述的方法，其特征在于，所述方法还包括：
    获取所述无人机的飞行高度；
    所述基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位；所述控制所述无人机向所述目标停机位降落，包括：
    当所述无人机处于第一飞行高度时，基于识别得到的所述标识分布信息，在所述场景图像观测的所述停机场中确定目标停机区域，控制所述无人机向所述目标停机区 域移动；
    所述方法还包括：
    当所述无人机处于第二飞行高度时，基于识别得到的所述标识分布信息，在所述场景图像观测的多个停机位中确定所述无人机的目标停机位；
    控制所述无人机向所述目标停机位降落；
    其中，所述第二飞行高度小于所述第一飞行高度。
  7. 根据权利要求1所述的方法，其特征在于，所述获取所述无人机的传感器采集的停机场的场景图像，包括：
    获取所述无人机在不同飞行高度采集的所述场景图像。
  8. 根据权利要求1所述的方法，其特征在于，所述控制所述无人机向所述目标停机位降落，包括：
    根据所述场景图像中确定所述无人机与所述目标停机位的相对位置关系；
    基于所述相对位置关系，控制所述无人机向所述目标停机位降落。
  9. 根据权利要求1所述的方法，其特征在于，所述获取所述无人机的传感器采集的停机场的场景图像包括：
    获取无人机的定位；
    当所述无人机的定位处于所述停机场对应的区域，获取所述场景图像。
  10. 根据权利要求1所述的方法，其特征在于，所述标识的类型与所述停机位的外观属性关联；和\或，所述标识的类型与所述停机位的周边的物体的外观属性关联。
  11. 根据权利要求1所述的方法，其特征在于，所述标识的类型与所述停机位的以下一种或者多种外观属性关联：轮廓、颜色、尺寸、材质、标识符。
  12. 根据权利要求1所述的方法，其特征在于，所述获取所述无人机的传感器采集的停机场的场景图像，包括：
    获取拍摄所述场景图像时所述无人机的朝向信息；
    基于所述朝向信息矫正所述场景图像；
    在矫正后的所述场景图像中识别所述停机位的标识，得到多个所述停机位的标识分布信息。
  13. 根据权利要求1所述的方法，其特征在于，所述获取所述无人机的传感器采集的停机场的场景图像，包括：
    基于预设的朝向信息，调整所述无人机的所述传感器的朝向；
    基于调整朝向后的所述传感器采集所述场景图像。
  14. 一种无人机降落控制方法，其特征在于，包括：
    在无人机降落过程中，获取所述无人机的传感器采集的停机场的场景图像，所述停机场包括多个停机位，多个所述停机位包括具有不同的标识的至少两个停机位；
    在所述场景图像中识别所述停机位的标识，得到多个所述停机位的标识分布信息；
    基于识别得到的所述标识分布信息，在所述场景图像观测的所述停机场中确定目标停机区域；
    控制所述无人机向所述目标停机区域移动。
  15. 根据权利要求14所述的方法，其特征在于，所述方法还包括：
    确定所述目标停机区域中的空闲停机位；
    控制所述无人机向任意所述空闲停机位降落。
  16. 一种控制装置，其特征在于，所述控制装置用于无人机的降落控制，包括存储器和处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现权利要求1至15任一所述的方法。
  17. 一种无人机，其特征在于，所述无人机包括用于采集停机场的场景图像的传感器以及控制装置；
    其中，所述控制装置用于无人机的降落控制，包括存储器和处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现权利要求1至15任一所述的方法。
  18. 一种用于停放无人机的设备，其特征在于，所述设备包括多个停机位，多个所述停机位包括具有不同的标识的至少两个停机位，所述多个停机位的标识分布信息可用于定位每个所述停机位。
  19. 根据权利要求18所述的设备，其特征在于，所述设备包括传感器以及处理器，所述处理器用于基于所述传感器所采集的信息，执行预设的程序。
  20. 根据权利要求18所述的设备，其特征在于，所述设备具备可折叠性，所述传感器用于当检测到无人机降落至预设的停机位上时通知所述处理器；所述处理器用于控制所述设备折叠，以对所述无人机打包。
  21. 根据权利要求18所述的设备，其特征在于，所述标识的类型与所述停机位的外观属性关联；和\或，所述标识的类型与所述停机位的周边的物体的外观属性关联。
  22. 根据权利要求18所述的设备，其特征在于，所述标识的类型与所述停机位的以下一种或者多种外观属性关联：轮廓、颜色、尺寸、材质、标识符。
  23. 一种无人机降落控制系统，其特征在于，所述无人机降落控制系统包括无人 机以及用于停放无人机的设备；
    所述无人机包括用于采集停机场的场景图像的传感器以及控制装置；其中，所述控制装置用于无人机的降落控制，包括存储器和处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现权利要求1至15任一所述的方法；
    所述设备包括多个停机位，多个所述停机位包括具有不同的标识的至少两个停机位，所述多个停机位的标识分布信息可用于定位每个所述停机位。
  24. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质上存储有计算机指令，所述计算机指令被执行时实现权利要求1至15任一项所述方法的步骤。
PCT/CN2021/100685 2021-06-17 2021-06-17 无人机降落控制方法、装置、无人机、系统及存储介质 WO2022261901A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/100685 WO2022261901A1 (zh) 2021-06-17 2021-06-17 无人机降落控制方法、装置、无人机、系统及存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/100685 WO2022261901A1 (zh) 2021-06-17 2021-06-17 无人机降落控制方法、装置、无人机、系统及存储介质

Publications (1)

Publication Number Publication Date
WO2022261901A1 true WO2022261901A1 (zh) 2022-12-22

Family

ID=84525929

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/100685 WO2022261901A1 (zh) 2021-06-17 2021-06-17 无人机降落控制方法、装置、无人机、系统及存储介质

Country Status (1)

Country Link
WO (1) WO2022261901A1 (zh)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105353765A (zh) * 2015-11-10 2016-02-24 浙江大华技术股份有限公司 一种控制无人机降落的方法及装置
CN106127201A (zh) * 2016-06-21 2016-11-16 西安因诺航空科技有限公司 一种基于视觉定位降落末端的无人机降落方法
WO2018112404A1 (en) * 2016-12-16 2018-06-21 The Government Of The United States Of America As Represented By The Secretary Of The Navy Autonomous airborne vehicle controlled by millmeter wave radar
CN108216624A (zh) * 2017-12-25 2018-06-29 上海歌尔泰克机器人有限公司 一种控制无人机降落的方法、装置及无人机
CN109521799A (zh) * 2018-09-28 2019-03-26 易瓦特科技股份公司 基于地面标识进行识别的飞行控制方法及装置
US10401874B1 (en) * 2018-11-10 2019-09-03 Skyways Air Transportation Inc Autonomous aircraft navigation
CN110687928A (zh) * 2019-09-09 2020-01-14 苏州臻迪智能科技有限公司 降落控制方法，系统，无人机及存储介质
US20200301445A1 (en) * 2019-03-21 2020-09-24 Wing Aviation Llc Geo-fiducials for uav navigation

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105353765A (zh) * 2015-11-10 2016-02-24 浙江大华技术股份有限公司 一种控制无人机降落的方法及装置
CN106127201A (zh) * 2016-06-21 2016-11-16 西安因诺航空科技有限公司 一种基于视觉定位降落末端的无人机降落方法
WO2018112404A1 (en) * 2016-12-16 2018-06-21 The Government Of The United States Of America As Represented By The Secretary Of The Navy Autonomous airborne vehicle controlled by millmeter wave radar
CN108216624A (zh) * 2017-12-25 2018-06-29 上海歌尔泰克机器人有限公司 一种控制无人机降落的方法、装置及无人机
CN109521799A (zh) * 2018-09-28 2019-03-26 易瓦特科技股份公司 基于地面标识进行识别的飞行控制方法及装置
US10401874B1 (en) * 2018-11-10 2019-09-03 Skyways Air Transportation Inc Autonomous aircraft navigation
US20200301445A1 (en) * 2019-03-21 2020-09-24 Wing Aviation Llc Geo-fiducials for uav navigation
CN110687928A (zh) * 2019-09-09 2020-01-14 苏州臻迪智能科技有限公司 降落控制方法，系统，无人机及存储介质

Similar Documents

Publication Publication Date Title
US11693428B2 (en) Methods and system for autonomous landing
US11656638B1 (en) Backup navigation system for unmanned aerial vehicles
CN113597591B (zh) 用于无人飞行器导航的地理基准
CN107240063B (zh) 一种面向移动平台的旋翼无人机自主起降方法
CA2920388C (en) Unmanned vehicle searches
Luo et al. A survey of intelligent transmission line inspection based on unmanned aerial vehicle
US20220074744A1 (en) Unmanned Aerial Vehicle Control Point Selection System
CN104298248A (zh) 旋翼无人机精确视觉定位定向方法
US20240159557A1 (en) Techniques for collaborative map construction between an unmanned aerial vehicle and a ground vehicle
US20240176367A1 (en) Uav dispatching method, server, dock apparatus, system, and storage medium
US20190187239A1 (en) System and method for automatic determination of location of an autonomous vehicle when a primary location system is offline
CN115981355A (zh) 一种可快速精准降落的无人机自动巡航方法及系统
WO2022261901A1 (zh) 无人机降落控制方法、装置、无人机、系统及存储介质
US11459117B1 (en) Drone-based cameras to detect wind direction for landing
Supriyono et al. Design, building and performance testing of GPS and computer vision combination for increasing landing precision of quad-copter drone
CN114384921B (zh) 一种基于无人机母车的车载无人机的升降方法
WO2021072766A1 (zh) 飞行控制方法、系统、无人飞行器及存储介质
CN114527779A (zh) 货物配送无人机的控制方法、系统和存储介质
US20240094743A1 (en) Pixel-by-pixel segmentation of aerial imagery for autonomous vehicle control
CN115019563B (zh) 一种无人机着陆航线规划方法及规划系统
US20240212205A1 (en) Using ambiguous semantic labeled images for uav positioning
EP4390316A1 (en) Using ambiguous semantic labeled images for uav positioning
JP2024027754A (ja) 情報処理装置、情報処理方法、及びプログラム
CN114995472A (zh) 一种基于目标识别无人机精准降落方法
Xu et al. A Fast Positioning Method for Docking Vehicles in Mixed Traffic Scenarios.

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: 21945487

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: 21945487

Country of ref document: EP

Kind code of ref document: A1