CN106647814A - System and method of unmanned aerial vehicle visual sense assistant position and flight control based on two-dimensional landmark identification - Google Patents

Abstract

The invention discloses a UAV visual aided positioning and flight control system and method based on two-dimensional code landmark recognition. The system includes a UAV body, a sensor module, a tracking trajectory generation module, a visual processing module, a sensor update module, The flight control module, the visual aid control switching module, the command output module and the camera, through the visual extraction of the two-dimensional code markers arranged on the specific position of the route, integrate the inertial navigation system to calculate the precise position and attitude information, and then assist and improve The accuracy of the traditional GPS integrated navigation system, while providing diversified information guidance for UAVs through two-dimensional code code information, expanding the diversity of flight missions, in addition, a cascade flight control system based on deviation-based adaptive compensation control is proposed , to achieve a smooth transition between the marker recognition state and the unrecognized state, improve the stability of flight control, and then improve the accuracy and speed of recognition.

Description

A vision-assisted positioning and flight control system for UAV based on two-dimensional code landmark recognition and methods

technical field

The invention belongs to the technical field of unmanned aerial vehicles, and more specifically relates to a visual aided positioning and flight control system and method for an unmanned aerial vehicle based on two-dimensional code identification.

Background technique

In recent years, with the development of intelligence science and control science, UAV has become a hot research topic. At present, drones are widely used in aerial photography, land surveying and mapping, geological rescue, fire rescue, traffic monitoring and other fields. UAVs not only have practical social application value, but also have important research significance in engineering and academia, such as agricultural plant protection, power inspection, forest fire prevention, disaster detection and other fields, and have broad development prospects.

In the automatic flight of UAVs, the traditional integrated navigation technology is limited by the GPS accuracy problem, and the position accuracy is about ±2m. In the occasions where the flight route and hovering accuracy are relatively high, such as express logistics delivery , Disaster relief support, combat on board, automatic return to charging and other applications often require the use of other equipment to assist in improving the accuracy of reaching the target point during flight, which has certain limitations.

Contents of the invention

In view of the above defects or improvement needs of the prior art, the present invention provides a UAV visual aided positioning and flight control system based on the precise recognition of two-dimensional code landmarks. As the key point, through the visual extraction of the two-dimensional code markers, the calculation of the precise position and attitude information is fused with the inertial navigation system, thereby assisting and improving the accuracy of the traditional GPS integrated navigation system, and at the same time through the coding of the two-dimensional code Information provides diversified information guidance for drones and expands the diversity of flight missions. In addition, a cascaded flight control system based on deviation-based adaptive compensation control is proposed to realize the smooth transition between the marker recognition state and the unrecognized state, improve the stability of flight control, and then improve the accuracy and rapidity of recognition. This solves the technical problem that the traditional integrated navigation technology in the prior art is limited by the GPS accuracy, the position accuracy is low, and other equipment needs to be assisted to improve the accuracy of reaching the target point during flight.

In order to achieve the above object, according to one aspect of the present invention, a vision-assisted positioning and flight control system for unmanned aerial vehicles based on two-dimensional code landmark recognition is provided, which is characterized in that it includes: an unmanned aerial vehicle body, a sensor module, a tracking track Generation module, vision processing module, sensor update module, flight control module, command output module, visual aid control switching module and camera:

The sensor module is used to obtain the position information of the UAV body and the first motion velocity vector of the UAV body;

The tracking track generation module is used to generate route tracking tracks according to preset mission waypoint information, and perform discrete processing on the route tracking tracks to obtain N expected waypoints, where N is a positive integer;

The vision processing module is used to acquire the position information, attitude information and encoding information of the two-dimensional code marker according to the image of the two-dimensional code marker acquired by the camera, and the position information, attitude information and encoding information Obtaining a deviation distance vector of the camera relative to the two-dimensional code marker and a second motion velocity vector of the camera relative to the two-dimensional code marker;

The sensor update module is used to use the position information of the UAV body, the first motion velocity vector, the deviation distance vector and the second motion velocity vector to perform multi-sensor information fusion through a Kalman filter algorithm, Obtain the target position information of the drone body filtered by the Kalman filter algorithm, the first target motion velocity vector, the target deviation distance vector and the target second motion velocity vector;

The flight control module is used to utilize the expected position of the target expected waypoint, the expected velocity vector of the target expected waypoint, the target position information, the target first motion velocity vector, the target deviation distance vector and the target The second motion velocity vector generates a guidance instruction through deviation adaptive compensation, and sends the guidance instruction to the instruction output module, wherein the target desired waypoint is the waypoint that the drone is currently heading to, and the guidance instruction including roll and pitch angles;

The visual aid control switching module is used to control the flight control module to calculate guidance instructions according to the information obtained by the sensor module and the visual processing module when the two-dimensional code marker is in the recognition state. When the two-dimensional code marker is in an unrecognized state, control the flight control module to calculate guidance instructions only according to the information obtained by the sensor module;

The command output module is used to output the guidance command.

Preferably, the camera is located at the bottom of the drone body, and the field of view of the camera is vertically downward.

Preferably, the visual processing module includes an image grayscale module, an image binarization module, a binary image processing module, a two-dimensional code information extraction module, and a position and attitude acquisition module,

The image grayscale module is used to convert the image of the two-dimensional code marker into a single-channel grayscale image;

The image binarization module is used to set a fixed threshold according to the single-channel grayscale image, and convert the grayscale image into a binary image;

The binary image processing module is used for performing contour detection on the binary image, traversing all polygons with four sides in the binary image, and removing polygons whose area is smaller than a preset threshold, and then converting the remaining edges Orthogonal projection of 4 polygons to obtain a standard square image;

The two-dimensional code information extraction module is used to extract binary coded information and corner information in the square image according to preset coded information rules;

The position and posture acquisition module is used to obtain the deviation distance vector of the camera relative to the two-dimensional code marker and the first position of the camera relative to the two-dimensional code marker according to the extracted binary code information and corner point information. Two motion velocity vectors.

According to another aspect of the present invention, a visual aided positioning and flight control method for UAV based on two-dimensional code landmark recognition is provided, which is characterized in that it includes:

S1: Obtain the position information of the UAV and the first motion velocity vector of the UAV;

S2: Generate route tracking trajectories according to preset mission waypoint information, and perform discrete processing on the route tracking trajectories to obtain N expected waypoints, where N is a positive integer;

S3: Acquire the position information, posture information and coding information of the two-dimensional code marker according to the image of the two-dimensional code marker obtained by the camera, and obtain the relative position of the camera relative to the coding information from the position information, posture information and coding information The deviation distance vector of the two-dimensional code marker and the second motion velocity vector of the camera relative to the two-dimensional code marker;

S4: Using the position information of the UAV, the first motion speed vector, the deviation distance vector and the second motion speed vector to perform multi-sensor information fusion through the Kalman filter algorithm, and obtain the Kalman filter algorithm The target position information of the filtered UAV, the first target motion velocity vector, the target deviation distance vector and the target second motion velocity vector;

S5: Utilize the expected position of the target's expected waypoint, the expected speed vector of the target's expected waypoint, the target position information, the target's first motion speed vector, the target's deviation distance vector, and the target's second motion speed vector A guidance instruction is generated by adaptive offset compensation, wherein the target desired waypoint is the waypoint the UAV is currently heading to, and the guidance instruction includes a roll angle and a pitch angle;

S6: Outputting the guidance instruction.

Preferably, the camera is located at the bottom of the drone, and the field of view of the camera is vertically downward.

Preferably, step S3 specifically includes the following sub-steps:

S301: Convert the image of the two-dimensional code marker into a single-channel grayscale image;

S302: Set a fixed threshold according to the single-channel grayscale image, and convert the grayscale image into a binary image;

S303: Perform contour detection on the binary image, traverse all polygons with 4 sides in the binary image, and eliminate polygons whose area is smaller than a preset threshold, and then perform the remaining polygons with 4 sides Orthogonal projection to get a standard square image;

S304: Extract binary coded information and corner point information in the square image according to preset coded information rules;

S305: Obtain a deviation distance vector of the camera relative to the two-dimensional code marker and a second motion velocity vector of the camera relative to the two-dimensional code marker according to the extracted binary code information and corner point information.

Generally speaking, compared with the prior art, the above technical solution conceived by the present invention mainly has the following technical advantages:

(1) By identifying the two-dimensional code markers arranged on a specific position on the ground, using two-dimensional code coding technology to obtain landmark information, and fusing various sensor information to improve the positioning accuracy of the drone, thereby assisting and improving The accuracy of the traditional GPS integrated navigation system, and because the two-dimensional code can provide rich landmark information and has encryption capabilities, it can provide diversified information guidance for UAVs, thereby expanding the diversity of flight missions;

(2) The same cascaded flight control system is used in the marker recognition state and the unrecognized state, and a deviation-based adaptive compensation control method is proposed, and the additional position information obtained in the marker target recognition state Compensation can achieve a smooth transition between the marker recognition state and the unrecognized state, improve the stability of flight control, and ensure that the rotor UAV can achieve fast and accurate recognition in various interference environments.

Description of drawings

Fig. 1 is a hardware structure diagram of a kind of unmanned aerial vehicle high-precision autonomous flight disclosed by the embodiment of the present invention;

Fig. 2 is a schematic structural diagram of a UAV visual aided positioning and flight control system based on two-dimensional code landmark recognition disclosed in an embodiment of the present invention;

Fig. 3 is an information interaction diagram of the visual aided positioning of a UAV based on two-dimensional code landmark recognition and the modules of the flight control system disclosed in the embodiment of the present invention;

4 is a schematic flow diagram of a method for visually aided positioning and flight control of a drone based on two-dimensional code landmark recognition disclosed in an embodiment of the present invention;

Fig. 5 is a schematic flow chart of a high-precision autonomous flight of a drone disclosed in an embodiment of the present invention.

detailed description

In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not constitute a conflict with each other.

Fig. 1 shows the hardware structure diagram of a kind of unmanned aerial vehicle high-precision autonomous flight disclosed by the embodiment of the present invention. In the hardware structure diagram shown in Fig. 1, the upper left part of Fig. 1 is a set of navigation sensors, which can include accelerometer, Gyroscopes, ultrasonic sensors, barometers, magnetometers, and GPS modules, etc., each of which can communicate with the flight control board in the lower left part of Figure 1 through IIC and SPI interfaces, and the lower right part of Figure 1 is the camera on the drone, which can The USB2.0 interface communicates with the vision processing board in the upper right part of Figure 1, and the vision processing board can communicate with the flight control board through the TTL serial port.

Among them, the flight control main board can adopt STM32F407 embedded processor, and the operating frequency is 168Mhz. Navigation sensors can specifically include: MPU6050 gyroscope and accelerometer, MS5611 high-precision barometer, M8NGPS receiver, US100 ultrasonic range finder. The visual processing main board can adopt S5P4418 high-performance processor, the operating frequency is 1.4Ghz, and it has 1GB DDR3 operating memory. The camera can be KS2A17, and the communication method with the visual processing main board can be USB2.0, and the maximum frame rate is 120fps under the resolution of 640×480. The vision processing main board and the flight control main board can use TTL serial port connection for data communication.

Figure 2 is a schematic structural diagram of a UAV visual aided positioning and flight control system based on two-dimensional code landmark recognition disclosed in an embodiment of the present invention, and Figure 3 is a structural diagram of a two-dimensional code-based landmark recognition disclosed in an embodiment of the present invention Information interaction diagram of each module of UAV visual aided positioning and flight control system. As shown in Figure 2 and Figure 3, the system of the present invention includes a UAV body, a sensor module, a tracking trajectory generation module, a visual processing module, a sensor update module, a flight control module, an instruction output module, a visual aid control switching module and Camera.

Wherein, the above-mentioned sensor module is used to obtain the position information of the UAV body and the first motion velocity vector of the UAV body;

The tracking track generating module is used to generate route tracking tracks according to preset mission waypoint information, and perform discrete processing on the above route tracking tracks to obtain N expected waypoints, where N is a positive integer;

The above-mentioned visual processing module is used to obtain the position information, attitude information and coding information of the two-dimensional code marker according to the image of the two-dimensional code marker acquired by the camera, and obtain the relative position of the camera relative to the two-dimensional code from the above position information, attitude information and coding information. The deviation distance vector of the two-dimensional code marker and the second motion velocity vector of the camera relative to the two-dimensional code marker;

Wherein, the camera is located at the bottom of the drone body, and the field of view of the camera is vertically downward.

Among them, the visual processing module includes an image grayscale module, an image binarization module, a binary image processing module, a two-dimensional code information extraction module, and a position and posture acquisition module,

The above-mentioned image grayscale module is used to convert the image of the two-dimensional code marker into a single-channel grayscale image;

The above image binarization module is used to set a fixed threshold value according to the single-channel grayscale image, and convert the grayscale image into a binary image;

The above-mentioned binary image processing module is used to perform contour detection on the binary image, traverse all polygons with 4 sides in the binary image, and eliminate polygons whose area is smaller than the preset threshold, and then convert the remaining polygons with 4 sides Orthogonal projection of the polygon to obtain a standard square image;

The two-dimensional code information extraction module is used to extract binary coded information and corner information in a square image according to preset coded information rules;

The above-mentioned position and posture acquisition module is used to obtain the deviation distance vector of the camera relative to the two-dimensional code marker and the second motion velocity vector of the camera relative to the two-dimensional code marker according to the extracted binary code information and corner point information.

Among them, the size of the two-dimensional code marker is m cm × m cm, and the obtained corner information of the two-dimensional code marker represents the position information under the image coordinates of the camera. Since the follow-up is mainly to compensate the flight error of the waypoint, Therefore, it is uniformly stipulated that the real world coordinates of the four corners of the two-dimensional code marker are (m, m, 0), (m, 0, 0), (0, m, 0), (0, 0, 0), Camera imaging principle: s·m'=A·[R|T]·M, where A is the internal reference matrix of the camera, which can be obtained through experimental calibration, m' is the position of the camera in the camera coordinate system, and M is the position of the camera in the real world The position in the coordinate system, [R|T] is the rotation and translation matrix, that is, the position and attitude of the camera relative to a certain point in the real world coordinate system, and the deviation distance of the camera relative to the two-dimensional code marker can be calculated vector and the second motion velocity vector of the camera relative to the two-dimensional code marker.

The above sensor update module is used to use the position information of the UAV body, the first motion velocity vector, the deviation distance vector and the second motion velocity vector to perform multi-sensor information fusion through the Kalman filter algorithm, and obtain the Kalman filter algorithm filtered The target position information of the UAV body, the first target motion speed vector, the target deviation distance vector and the target second motion speed vector.

Among them, the measurement accuracy can be improved by designing a Kalman filter to fuse multi-sensor information. The state update formula of the Kalman filter is:

Among them, θ and γ are the pitch and roll angles in the rotation matrix R, V is the velocity vector of the drone in the real world coordinates, a is the acceleration vector of the drone in the real world coordinates, a ^b is the drone The acceleration vector under the coordinates can be measured by the accelerometer in the drone, w ^b is the angular velocity vector under the coordinates of the drone, which can be measured by the gyroscope in the drone, and Δt is the filter update interval.

The above-mentioned flight control module is used to use the expected position of the target expected waypoint, the expected speed vector of the target expected waypoint, the target position information, the first target motion speed vector, the target deviation distance vector and the target second motion speed vector to pass the deviation self-adaptation Compensation generates a guidance command, and sends the guidance command to the command output module, wherein the desired waypoint of the target is the waypoint that the UAV is currently going to, and the guidance command includes a roll angle and a pitch angle;

Among them, the control goal of high-precision flight is to make the position of the UAV converge to a sufficiently small neighborhood of the desired waypoint set of the target. In the high-precision flight stage with visual aids, since the measurement accuracy of the visual equipment is better than that of the traditional navigation equipment, it is necessary to compensate the input quantity in the controller at this time:

V _err (t)＝[V ^d (t)-w ₃ ·V(t)]-w ₄ ·V _vision (t)

Among them, P ^d (t), V ^d (t) are the desired position and speed input vectors of the UAV, respectively, P _err (t), _Verr (t) are the position outer loop controller and the speed inner loop control The error input vector of the device, P(t), V(t) are the position vector and motion velocity vector of the UAV calculated by the traditional GPS integrated navigation system, respectively, T(t), V _vision (t) are the vision processing The deviation distance vector and motion velocity vector of the UAV relative to the target two-dimensional code marker calculated by the module, w ₁ , w ₂ , w ₃ , and w ₄ are the compensation weight coefficients, generally w ₁ = w ₂ , w ₃ = w ₄ , the compensation weight coefficient can take a fixed value, or it can be determined in an adaptive way:

w ₂ =1-w ₁

During the flight of the UAV, the limitation of the field of view of the camera and the interference of the actual flight environment will affect the accuracy of the UAV marker extraction information. The traditional UAV control system does not recognize the ground markers. Therefore, the flight control module will frequently switch between the two states of ground marker recognition and non-recognition, resulting in unstable control. The present invention adopts the same cascaded flight control module in the ground marker recognition state and the unrecognized state, that is, one flight control module is used, and a deviation-based self-adaptive compensation control method is proposed. Compensate for the acquired additional position and motion velocity information. It can realize the smooth transition between the ground marker recognition state and the unrecognized state, improve the stability of flight control, and ensure that the UAV can achieve high-precision flight in various environments.

The above-mentioned instruction output module is used for outputting the above-mentioned guidance instruction.

Fig. 4 is a schematic flowchart of a method for visually aided positioning and flight control of a drone based on two-dimensional code landmark recognition disclosed in an embodiment of the present invention, wherein the method shown in Fig. 4 includes the following steps:

S1: Obtain the position information of the UAV and the first motion velocity vector of the UAV;

S2: Generate route tracking trajectories according to preset mission waypoint information, and perform discrete processing on the route tracking trajectories to obtain N expected waypoints, where N is a positive integer;

S3: Accurately identify images of pre-arranged two-dimensional code markers acquired by the camera;

Wherein, step S3 is implemented in the following manner: according to the image of the pre-arranged two-dimensional code marker obtained by the camera, the position information, attitude information and coding information of the two-dimensional code marker are obtained, and the above position information, attitude information and coding information The information obtains the deviation distance vector of the camera relative to the two-dimensional code marker and the second motion velocity vector of the camera relative to the two-dimensional code marker;

S4: Using the information of the identified two-dimensional code marker, the position information of the UAV, and the first motion velocity vector to perform multi-sensor information fusion through the Kalman filter algorithm;

Among them, the specific implementation of step S4 is as follows: using the position information of the UAV, the first motion velocity vector, the deviation distance vector and the second motion velocity vector to perform multi-sensor information fusion through the Kalman filter algorithm, and obtain the Kalman filter algorithm The filtered UAV's target position information, target first motion speed vector, target deviation distance vector and target second motion speed vector.

S5: Use the information obtained after Kalman filtering to generate guidance instructions, where the guidance instructions include roll angle and pitch angle;

Wherein, the specific implementation of step S5 is: using the expected position of the target desired waypoint, the desired speed vector of the target desired waypoint, the target position information, the first target motion speed vector, the target deviation distance vector and the target second motion speed vector Guidance instructions are generated through bias adaptive compensation, where the target desired waypoint is the waypoint the UAV is currently heading to, and the guidance instructions include roll angle and pitch angle.

S6: output the above-mentioned guidance instruction.

Fig. 5 is a schematic flow chart of a high-precision autonomous flight of a drone disclosed in an embodiment of the present invention. In Figure 5, there are three mission waypoints, among which mission waypoints (n) and (n+1) are key waypoints, and markers are arranged on the ground: QR code 1 and QR code 2 . When the UAV flies through the waypoint (n-1), it only uses the traditional GPS integrated navigation system. When flying through the mission waypoints (n) and (n+1), the position, attitude and coding information of the ground markers will be extracted to assist the traditional GPS integrated navigation system.

It is easy for those skilled in the art to understand that the above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention, All should be included within the protection scope of the present invention.

Claims (6)

1. it is a kind of based on Quick Response Code terrestrial reference recognize unmanned plane vision auxiliary positioning and flight control system, it is characterised in that include：Nothing Man-machine body, sensor assembly, pursuit path generation module, vision processing module, sensor update module, flight control mould Block, command output module, vision auxiliary control handover module and camera：

The sensor assembly is used to obtain the first fortune of the positional information of the unmanned plane body and the unmanned plane body Dynamic velocity；

The pursuit path generation module is used to generate course line pursuit path according to default task way point information, and to the boat Line pursuit path carries out discrete processes and obtains N number of expectation destination, and N is positive integer；

The vision processing module is used for the image of the Quick Response Code mark acquired according to the camera and obtains described two Positional information, attitude information and the coding information of code mark thing are tieed up, by the positional information, attitude information and coding information The deviation distance vector and the camera that the camera is obtained relative to the Quick Response Code mark is relative to described two Second movement velocity vector of dimension code mark thing；

The sensor update module be used for using the positional information of the unmanned plane body, the first movement velocity vector, The deviation distance vector and the second movement velocity vector carry out multi-sensor information and melt by Kalman filtering algorithm Close, obtain the target position information of the filtered unmanned plane body of Jing Kalman filtering algorithms, target the first movement velocity vector, Target deviation distance vector and target the second movement velocity vector；

The flight control modules are used to expect that the desired locations of destination, target expect the desired speed arrow of destination using target Amount, the target position information, the target the first movement velocity vector, the target deviation distance vector and the target Second movement velocity vector by deviation adaptive equalization generate guidance command, by it is described guidance command be sent to it is described instruction it is defeated Go out module, wherein, the target expects the destination that destination is currently being directed to for unmanned plane, described to guidance command including roll angle And the angle of pitch；

The vision auxiliary control handover module, for when the Quick Response Code mark is in identification state, controlling described flying Row control module is calculated according to the information that the sensor assembly and the vision processing module are obtained and guidanceed command, described two When dimension code mark thing is in unidentified state, the information that the flight control modules are obtained according only to the sensor assembly is controlled Calculating is guidanceed command；

The command output module is used to export described guidanceing command.

2. system according to claim 1, it is characterised in that the camera is located at the bottom of the unmanned plane body, And the visual field direction of the camera is vertically downward.

3. system according to claim 1, it is characterised in that the vision processing module include image gray processing module, Image binaryzation module, binary map processing module, 2 D code information extraction module and position and attitude acquisition module,

Described image gray processing module is used to for the image of the Quick Response Code mark to be converted into single channel gray-scale map；

Described image binarization block is used to set a fixed threshold values according to single channel gray-scale map, and gray-scale map is converted into into two-value Figure；

The binary map processing module is used to carry out the binary map contour detecting, travels through all side numbers in the binary map For 4 polygon, and polygon of the area less than predetermined threshold value is rejected, then carry out on the polygon that remaining side number is 4 Rectangular projection, obtains the square-shaped image of standard；

The 2 D code information extraction module is used for according to two in square-shaped image described in default coding information Rule Extraction Scale coding information and angle point information；

The position and attitude acquisition module is used to obtain the camera according to the binary-coded information and angle point information extracted Relative to the Quick Response Code mark deviation distance vector and the camera relative to the of the Quick Response Code mark Two movement velocity vectors.

4. it is a kind of based on Quick Response Code terrestrial reference recognize unmanned plane vision auxiliary positioning and winged prosecutor method, it is characterised in that include：

S1：Obtain the positional information of unmanned plane and the first movement velocity vector of unmanned plane；

S2：Course line pursuit path is generated according to default task way point information, and discrete place is carried out to the course line pursuit path Reason obtains N number of expectation destination, and N is positive integer；

S3：The positional information of the image acquisition Quick Response Code mark of the Quick Response Code mark acquired according to camera, Attitude information and coding information, by the positional information, attitude information and coding information obtain the camera relative to The deviation distance vector and the camera of the Quick Response Code mark is moved relative to the second of the Quick Response Code mark Velocity；

S4：Using the positional information of the unmanned plane, the first movement velocity vector, the deviation distance vector and described Second movement velocity vector carries out multi-sensor information fusion by Kalman filtering algorithm, obtains the filter of Jing Kalman filtering algorithms The target position information of the unmanned plane after ripple, target the first movement velocity vector, target deviation distance vector and target second Movement velocity vector；

S5：Expect that the desired locations of destination, target expect the desired speed vector of destination, target location letter using target Breath, the target the first movement velocity vector, the target deviation distance vector and the target the second movement velocity vector Generated by deviation adaptive equalization and guidanceed command, wherein, the target expects the boat that destination is currently being directed to for unmanned plane Point, it is described to guidance command including roll angle and the angle of pitch；

S6：Guidance command described in output.

5. method according to claim 4, it is characterised in that the camera is located at the bottom of the unmanned plane, and The visual field direction of the camera is vertically downward.

6. method according to claim 4, it is characterised in that step S3 specifically includes following sub-step：

S301：The image of the Quick Response Code mark is converted into into single channel gray-scale map；

S302：One fixed threshold values is set according to single channel gray-scale map, gray-scale map is converted into into binary map；

S303：Contour detecting is carried out to the binary map, it is 4 polygon to travel through all side numbers in the binary map, and is picked Except area is less than the polygon of predetermined threshold value, then the polygon that remaining side number is 4 is carried out into rectangular projection, obtain standard Square-shaped image；

S304：Believe according to the binary-coded information in square-shaped image described in default coding information Rule Extraction and angle point Breath；

S305：Binary-coded information and angle point information according to extracting obtains the camera relative to the two-dimentional code mark Second movement velocity vector of the deviation distance vector and the camera of thing relative to the Quick Response Code mark.