CN107817820A - A kind of unmanned plane autonomous flight control method and system based on deep learning - Google Patents
A kind of unmanned plane autonomous flight control method and system based on deep learning Download PDFInfo
- Publication number
- CN107817820A CN107817820A CN201710961301.4A CN201710961301A CN107817820A CN 107817820 A CN107817820 A CN 107817820A CN 201710961301 A CN201710961301 A CN 201710961301A CN 107817820 A CN107817820 A CN 107817820A
- Authority
- CN
- China
- Prior art keywords
- target
- unmanned plane
- deep learning
- control
- flight control
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- RZVHIXYEVGDQDX-UHFFFAOYSA-N 9,10-anthraquinone Chemical compound C1=CC=C2C(=O)C3=CC=CC=C3C(=O)C2=C1 RZVHIXYEVGDQDX-UHFFFAOYSA-N 0.000 title claims abstract description 38
- 238000013135 deep learning Methods 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 title claims abstract description 23
- 238000001514 detection method Methods 0.000 claims abstract description 33
- 230000007935 neutral effect Effects 0.000 claims abstract description 15
- 238000013528 artificial neural network Methods 0.000 claims abstract description 8
- 238000012544 monitoring process Methods 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 6
- 238000011217 control strategy Methods 0.000 claims description 4
- 239000011248 coating agent Substances 0.000 claims description 2
- 238000000576 coating method Methods 0.000 claims description 2
- 238000007689 inspection Methods 0.000 claims description 2
- 230000009467 reduction Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 6
- 238000012360 testing method Methods 0.000 abstract description 5
- 238000004804 winding Methods 0.000 abstract description 3
- 238000003384 imaging method Methods 0.000 description 9
- 238000004891 communication Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000013461 design Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000000007 visual effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 230000035479 physiological effects, processes and functions Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
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
- G05D1/0088—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours
-
- 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
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)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Game Theory and Decision Science (AREA)
- Medical Informatics (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention discloses a kind of unmanned plane autonomous flight control method and system based on deep learning, propose the computer vision target detection of the artificial neural network based on deep learning with towards algorithm for estimating, image is gathered using the multi-rotor unmanned aerial vehicle for carrying head camera, target is detected using the arithmetic element for running neutral net and estimates target direction, and control instruction is finally fed back into UAV Flight Control device with reference to control algolithm.The present invention constructs distributed data acquisition computing and the control winding of " data acquisition unit deep learning arithmetic element flight control units ", test result indicates that, the present invention can realize real-time neural network computing, and then realize that unmanned plane quickly independently finds target, close to target and with the effect of particular pose tracking target with reference to control algolithm.
Description
Technical field
The invention belongs to computer vision and automation field, and in particular to a kind of unmanned plane based on deep learning is certainly
Main flight control method and system.
Background technology
Unmanned plane compared to manned vehicle, have low cost, small volume, it is easy to use, into producing, machine low with maintenance cost
The advantages that dynamic property is by force and survival ability is strong.Because no personnel drive, unmanned plane is not limited by the physiology and life risk of personnel,
It is suitably executed the task of " uninteresting and dangerous " such as information acquisition, geological exploration, low latitude investigation and anti-terrorism strikes.With the hair of science and technology
Exhibition, the production cost of unmanned plane further reduce, and start toward the development of the fields such as civilian and scientific research, such as gas pipeline monitoring, region
Cover monitoring, disaster emergency searching and rescuing, agricultural plant protection, public security fire-fighting and remote sensing are surveyed and drawn etc..
The practical application of unmanned plane is not most high based on artificial remote control or intervention, automaticity under various scenes in early days.
With the continuous expansion of unmanned plane automatically working demand, the target following based on computer vision turns into the heat studied instantly
Point, it is widely used in the field such as flight automatically in security protection rescue, man-machine interaction and unmanned plane.Calculated by target detection and tracking
Method can obtain position and dimensional information of the target on imaging plane, and what is then introduced by filtering out during this makes an uproar
Sound, the control instruction of unmanned plane is obtained according to control algolithm afterwards, and then control the posture of unmanned plane.Said process constantly follows
Ring, it is possible to achieve unmanned plane tracks the purpose of target.
Target detection and the important research content that target direction estimation is machine vision.Traditional target detection flow is first
Target location is oriented over an input image, feature then is extracted to target area, finally with the grader trained to extraction
Feature classified, it is target to judge the region.The flow is primarily present two problems, when time complexity it is high and
Window redundancy, second, feature extraction step extraction is characterized in being characterized as engineer, it is related to task, without universality.Pass
The studying mainly from the target based on various visual angles towards estimation, by the target court of 3D information towards estimation on target of system
To estimation and target based on 2D image informations towards estimate it is several in terms of deploy.Method based on 3D information is by data band
The limitation of width, computing resource, power demands and light etc..Estimation or detection essence are had towards algorithm for estimating based on 2D information
The drawbacks of degree is not high.With the rise of deep learning, the algorithm of target detection based on neutral net and the standard towards algorithm for estimating
True rate and operating rate are obtained for very big lifting, can be widely applied in practical application.
Traditional unmanned plane during flying path clustering is set based on destination, causes unmanned plane by flight controller internal processes
Flown successively along destination, independence is relatively low, and flexibility is poor, and speed is also relatively slow.
The content of the invention
It is an object of the invention to propose a kind of unmanned plane autonomous flight control method and system based on deep learning.
The computer vision target detection and direction that the artificial neural network based on deep learning is proposed in the present invention are estimated
Calculating method, image is gathered using the multi-rotor unmanned aerial vehicle for carrying head camera, is detected using the arithmetic element for running neutral net
Target simultaneously estimates target direction, and control instruction finally is fed back into UAV Flight Control device with reference to control algolithm.
Technical scheme is specifically described as follows.
The present invention provides a kind of unmanned plane independent flight control system based on deep learning, and it is distributed system, bag
Include with lower unit：
Data acquisition unit, it is special airborne slave computer, for gathering unmanned plane during flying state parameter information, airborne equipment and biography
Sensor running state information and its data；
Deep learning arithmetic element, it is the airborne embedded computing platform of embedded high-performance, for performing the inspection of neutral net target
Survey with being calculated towards estimation, it exports the input as control algolithm, implements flight control with reference to control algolithm；
Flight control units, it is onboard flight controller, for running control algolithm to control unmanned plane during flying posture, including it is preceding
To speed, lateral velocity, longitudinal velocity, yaw rate, roll and pitching.
The present invention also provides a kind of above-mentioned unmanned plane autonomous flight control method based on deep learning, including following step
Suddenly：
Target detection step, realized by the artificial neural network based on deep learning to the detection with flying target, obtain target
Corresponding region of interest ROI in the picture；
Target, using the result of target detection step, using the target recognized as input data, passes through base towards estimating step
Realized in the artificial neural network of deep learning to estimation of the target with respect to unmanned plane direction；
Autonomous flight control step, based on target detection and the result towards estimating step, it is used to track target with reference to structure
Control algolithm, implement flight control.
It is specific as follows the step of the target detection in the present invention：
Each angular image of target to be tracked is gathered in advance, as training dataset, utilizes stochastic gradient descent algorithm iterative
Deep learning is carried out, builds target detection neutral net；
Coordinate control unmanned plane during flying is set to target neighborhood according to target a priori location information, gathered by data acquisition unit
Airborne camera image, transmitted as input to deep learning arithmetic element, carry out target detection, obtained target to be tracked and working as
ROI in previous frame.
In the present invention, described target towards estimation the step of it is specific as follows：
Each angular image of target to be tracked is gathered in advance, is carried out towards after demarcation, as training dataset, using under stochastic gradient
Drop algorithm iteration and solve progress deep learning, structure target includes multilayer convolutional layer, Quan Lian towards estimation neutral net, the network
Layer and output layer are connect, with reference to active coating to lift nonlinear fitting, and the dimensionality reduction of characteristic pattern is realized using pond layer and prevents mould
Type over-fitting；
Input using the result that target detection step obtains as neutral net, target is carried out towards estimation, it is relative to obtain target
In unmanned plane it is current towards classification.
It is specific as follows the step of described autonomous flight control in the present invention：
The unmanned plane current pose information and camera current pose information obtained according to data acquisition unit, combining target detection
The target position information and dimensional information that step obtains, the angle and distance of estimation target to unmanned plane；Combining target direction is estimated
The target orientation information that step obtains, the positive minor arc formed with unmanned machine face of estimation target and its angle are counted to determine to be diversion
Direction.
When target range unmanned plane farther out, implement close to control strategy, be respectively adopted control algolithm control unmanned aerial vehicle body
Coordinate ventrocephalad speed and yaw rate, make unmanned plane with face posture close to target；
When target range unmanned plane is nearer, implementation is diversion control strategy, and control algolithm control unmanned aerial vehicle body coordinate is respectively adopted
Ventrocephalad speed, yaw rate and lateral velocity, unmanned plane is set to be diversion to target front.
In the present invention, tracking each angular image of target includes the more monitoring camera tracing figure pictures in ground.
In the present invention, the target a priori location information includes the more monitoring camera tracking informations in ground.
Compared to the prior art, the beneficial effects of the present invention are：
The present invention can realize real-time neural network computing, and then realize that unmanned plane quickly independently finds mesh with reference to control algolithm
Mark, close target and the effect that target is tracked with particular pose.
Brief description of the drawings
Fig. 1：Target position view in image in the present invention.
Fig. 2：Real system structure and functional block diagram in the present invention.
Fig. 3：Real winged test result schematic diagram.
Embodiment
The present invention is further described below in conjunction with the accompanying drawings and by the way that example is embodied, following examples simply describe
Property, it is not limited, it is impossible to which protection scope of the present invention is limited with this.
The key of the present invention is constructing system winding, and image, deep learning arithmetic element are gathered by data acquisition unit
Carry out computing, finally feed back to flight control units implementation control.Fig. 1：Target position view in image in the present invention.Figure
2：Real system structure and functional block diagram in the present invention.
Present disclosure be broadly divided into target detection and towards estimation neutral net design, Flight Control Algorithm design and
Distributed system realizes three parts.
1st, target detection with towards estimation
For insufficient existing for traditional target towards algorithm for estimating based on 2D or 3D information, set forth herein based on convolution god
Target through network is towards algorithm for estimating.In the case of given training set, convolutional neural networks can realize end-to-end study,
The parameter and classifier parameters of automatic learning characteristic extraction, avoid the time-consuming disadvantage low with accuracy rate of engineer's feature link
End.Because the target currently without disclosed unmanned plane visual angle is towards nominal data, therefore data acquisition, Ran Hou are carried out first
The accuracy for carrying algorithm is trained and tested on this data set.By target detection and towards algorithm for estimating, target can be obtained
The information such as position, yardstick and direction on imaging plane.The target of network detection can cover many class scopes, such as people
Body, vehicle, ship, building etc..
2nd, Flight Control Algorithm
The information such as position, yardstick and direction based on target on imaging plane, design the Flight Control Algorithm of unmanned plane.Due to
Unmanned plane distance objective farther out when target orientation information confidence level it is not high enough, so whole process is divided into two ranks by us
Section, i.e., close to stage and revolving process.Close to the stage, unmanned plane is according to position of the target on imaging plane and dimensional information
Oneself posture is adjusted, towards gtoal setting, most the image space of target moves on to the immediate vicinity of imaging plane at last.In rotation rank
Section, unmanned plane adjust oneself posture according to orientation information of target etc., are rotated around target, it is positive finally to move to alignment target
Position.
3rd, system is realized
Finally, on the basis of above-mentioned vision processing algorithm and Flight Control Algorithm, system each unit is united by communications framework
Come together, wherein unmanned plane is responsible for the collection and forwarding of sensor information；High-performance embedded airborne operation board is single as calculating
Member, it is responsible for carrying out visual processes to the image information received, final Flight Control Algorithm generates control instruction, controls unmanned plane
Flight path.
Using human body target tracking, pid algorithm control as example application scenarios, a kind of unmanned plane based on deep learning is certainly
Main flight control method comprises the following steps that：
1) target detection and the structure towards estimation neutral net
By target towards classification problem is modeled as the problem of estimation, the direction of target is divided into 8 directions, tested by allowing
Person does the sample data that different actions gathers each direction in different directions, then by being marked to the data collected
It is fixed, it can obtain the data set of target direction.Using PCA and the method for translation, amplified sample quantity, prevent that over-fitting influences
Generalization ability.Utilize stochastic gradient descent algorithm iterative.It is trained finally by the data set of the picture of collection
And test, show that input up to 97%, is down-sampled to 640 × 360 resolution ratio, to meet by the accuracy rate of test in actual identification
Real-time application demand.
2) autonomous flight control algorithm is built
By taking pid control algorithm as an example, in order to reduce the coupling between control variable, we control the angle of pitch and unmanned plane of head
Constant height.Unmanned plane distance objective farther out when, using close to stage control logic.By controlling unmanned plane to be sat in body
Forward speed under mark systemTo control the distance between unmanned plane and target, so as to realize that the image space by target moves on to
Purpose near imaging plane horizontal center line.When the vertical centerline of the image space migration imaging plane of target, we
Can be by adjusting the yaw angle of unmanned plane, the yaw angle of head or unmanned plane under body axis systemIt is inclined to reduce this
Difference, equally, we select to realize by adjusting the yaw angle of unmanned plane.Variable is controlled to use、、、Represent, point
Not Biao Shi unmanned plane under body axis system along x axles, y axles, the linear velocity of z axles and the angular speed of driftage.According to above-mentioned point
Analysis, the relation table between distance, unmanned plane yaw angle, imaging deviation and the controlled quentity controlled variable of unmanned plane and target can be obtained, such as table 1
It is shown.
The distance of the unmanned plane of table 1 and target, unmanned plane yaw angle, the relation table for being imaged deviation and controlled quentity controlled variable
When unmanned plane and target nearer apart, neutral net can estimate the orientation information of target.Unmanned plane will be according to target
Position, yardstick and towards etc. information around target flight.In order to ensure in flight course target all the time within sweep of the eye and into
Image position, which is tried one's best, maintains the center of imaging plane, and unmanned plane needs to do circular flight around target, introduces lateral velocity device,
Under the collective effect of lateral velocity device and aforementioned controllers, unmanned plane can be flown around target with approximate circle track.
3) working-flow
When system starts, communications framework automatic start, the interior communication winding of constructing system.Data acquisition unit, deep learning simultaneously
Arithmetic element, flight control units automatic start and standby, wait sending for the further interactive instruction in ground.
After system receives start-up trace instruction, flight control units start to aircraft altitude and head angle of declination first
Control, data acquisition unit is started working with deep learning arithmetic element after stabilization.The video that data acquisition unit will collect
Stream is down-sampled to 640*360 resolution ratio, is transmitted by communications framework to deep learning arithmetic element（High-performance embedded computing
Plate）Carry out target detection with towards estimation.Deep learning arithmetic element exports output result to control algolithm, most controls at last
Output, which is transmitted to flight controller, instructs aircraft flight.
Ground G UI can be obtained by communication link the status information of aircraft, the control information of controller, target detection with
Towards the object information and present image of estimation, monitoring display is used during putting into effect, while can be issued and be handed over to communications framework
Mutually instruction, to instruct the start and stop of system autonomous flight control algorithm, flight control can be required by instructing under the state of emergency
Process terminates output, and aircraft enters floating state or returns to Home.
Real winged test result as shown in figure 3, for the 221st in the video recording of ground monitoring Program Screen, 282,343,404,465,
526 frames.
Above-mentioned embodiment is only used as applied field of the present invention by taking the detection to human body classification target and pid control algorithm as an example
The a kind of of scape is implemented, and target can be replaced by other classifications such as vehicle, ship etc. as deep learning in practical application
The training set of neutral net, control algolithm can be changed to optimum control scheduling algorithm to obtain different control effect and performance.
The system all realizes modularized design on software and hardware, and flexibility is strong in structure；Functionally autgmentability is strong, can be automatic
Take photo by plane, man-machine interaction, the security protection of unmanned plane auxiliary, unmanned plane auxiliary traffic control etc. play huge effect.
To sum up, what the present invention can effectively realize unmanned plane follows flight control from major heading.
Claims (8)
- A kind of 1. unmanned plane independent flight control system based on deep learning, it is characterised in that it is distributed system, including With lower unit：Data acquisition unit, it is special airborne slave computer, for gathering unmanned plane during flying state parameter information, airborne equipment and biography Sensor running state information and its data；Deep learning arithmetic element, it is the airborne embedded computing platform of embedded high-performance, for performing the inspection of neutral net target Survey with being calculated towards estimation, it exports the input as control algolithm, implements flight control with reference to control algolithm；Flight control units, it is onboard flight controller, for running control algolithm to control unmanned plane during flying posture, including it is preceding To speed, lateral velocity, longitudinal velocity, yaw rate, roll and pitching.
- A kind of 2. unmanned plane autonomous flight control method according to claim 1 based on deep learning, it is characterised in that Comprise the following steps：Target detection step, realized by the artificial neural network based on deep learning to the detection with flying target, obtain target Corresponding region of interest ROI in the picture；Target, using the result of target detection step, using the target recognized as input data, passes through base towards estimating step Realized in the artificial neural network of deep learning to estimation of the target with respect to unmanned plane direction；Autonomous flight control step, based on target detection and the result towards estimating step, it is used to track target with reference to structure Control algolithm, implement flight control.
- 3. unmanned plane autonomous flight control method according to claim 2, it is characterised in that the step of the target detection It is specific as follows：Each angular image of target to be tracked is gathered in advance, as training dataset, utilizes stochastic gradient descent algorithm iterative Deep learning is carried out, builds target detection neutral net；Coordinate control unmanned plane during flying is set to target neighborhood according to target a priori location information, gathered by data acquisition unit Airborne camera image, transmitted as input to deep learning arithmetic element, carry out target detection, obtained target to be tracked and working as ROI in previous frame.
- 4. the autonomous aircraft control method of unmanned plane according to claim 2, it is characterised in that described target direction estimation The step of it is specific as follows：Each angular image of target to be tracked is gathered in advance, is carried out towards after demarcation, as training dataset, using under stochastic gradient Drop algorithm iteration and solve progress deep learning, structure target includes multilayer convolutional layer, Quan Lian towards estimation neutral net, the network Layer and output layer are connect, with reference to active coating to lift nonlinear fitting, and the dimensionality reduction of characteristic pattern is realized using pond layer and prevents mould Type over-fitting；Input using the result that target detection step obtains as neutral net, target is carried out towards estimation, it is relative to obtain target In unmanned plane it is current towards classification.
- 5. unmanned plane autonomous flight control method according to claim 1, it is characterised in that described autonomous flight controlThe step of it is specific as follows：The unmanned plane current pose information and camera current pose information obtained according to data acquisition unit, combining target detection The target position information and dimensional information that step obtains, the angle and distance of estimation target to unmanned plane；Combining target direction is estimated The target orientation information that step obtains, the positive minor arc formed with unmanned machine face of estimation target and its angle are counted to determine to be diversion Direction.
- 6. working as target range unmanned plane farther out, implement close to control strategy, control algolithm control unmanned aerial vehicle body is respectively adopted and sits Ventrocephalad speed and yaw rate are marked, makes unmanned plane with face posture close to target；When target range unmanned plane is nearer, implementation is diversion control strategy, and control algolithm control unmanned aerial vehicle body coordinate is respectively adopted Ventrocephalad speed, yaw rate and lateral velocity, unmanned plane is set to be diversion to target front.
- 7. the unmanned plane autonomous flight control method according to claim 3 or 4, it is characterised in that the tracking target is each Angular image includes the more monitoring camera tracing figure pictures in ground.
- 8. the autonomous aircraft control method of unmanned plane according to claim 3, it is characterised in that the target priori position letter Breath includes the more monitoring camera tracking informations in ground.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710961301.4A CN107817820A (en) | 2017-10-16 | 2017-10-16 | A kind of unmanned plane autonomous flight control method and system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710961301.4A CN107817820A (en) | 2017-10-16 | 2017-10-16 | A kind of unmanned plane autonomous flight control method and system based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107817820A true CN107817820A (en) | 2018-03-20 |
Family
ID=61608002
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710961301.4A Pending CN107817820A (en) | 2017-10-16 | 2017-10-16 | A kind of unmanned plane autonomous flight control method and system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107817820A (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108615244A (en) * | 2018-03-27 | 2018-10-02 | 中国地质大学（武汉） | A kind of image depth estimation method and system based on CNN and depth filter |
| CN108764456A (en) * | 2018-04-03 | 2018-11-06 | 北京环境特性研究所 | Airborne target identification model construction platform, airborne target recognition methods and equipment |
| CN108985193A (en) * | 2018-06-28 | 2018-12-11 | 电子科技大学 | A kind of unmanned plane portrait alignment methods based on image detection |
| CN109062220A (en) * | 2018-08-31 | 2018-12-21 | 阿里巴巴集团控股有限公司 | The method and apparatus of controlling terminal movement |
| CN109164821A (en) * | 2018-09-26 | 2019-01-08 | 中科物栖（北京）科技有限责任公司 | A kind of UAV Attitude training method and device |
| CN109885091A (en) * | 2019-03-21 | 2019-06-14 | 华北电力大学（保定） | A kind of unmanned plane autonomous flight control method and system |
| CN110104167A (en) * | 2019-04-25 | 2019-08-09 | 江苏大学 | A kind of automation search and rescue UAV system and control method using infrared thermal imaging sensor |
| CN110347035A (en) * | 2018-04-08 | 2019-10-18 | 北京京东尚科信息技术有限公司 | Method for autonomous tracking and device, electronic equipment, storage medium |
| CN110514206A (en) * | 2019-08-02 | 2019-11-29 | 中国航空无线电电子研究所 | A kind of unmanned plane during flying path prediction technique based on deep learning |
| CN110543182A (en) * | 2019-09-11 | 2019-12-06 | 济宁学院 | autonomous landing control method and system for small unmanned gyroplane |
| CN110879602A (en) * | 2019-12-06 | 2020-03-13 | 安阳全丰航空植保科技股份有限公司 | Unmanned aerial vehicle control law parameter adjusting method and system based on deep learning |
| CN111178148A (en) * | 2019-12-06 | 2020-05-19 | 天津大学 | Ground target geographic coordinate positioning method based on unmanned aerial vehicle vision system |
| CN111204452A (en) * | 2020-02-10 | 2020-05-29 | 北京建筑大学 | Target detection system based on miniature aircraft |
| CN111736190A (en) * | 2020-07-24 | 2020-10-02 | 广东电网有限责任公司 | Unmanned aerial vehicle airborne target detection system and method |
| CN111824406A (en) * | 2020-07-17 | 2020-10-27 | 南昌航空大学 | Public safety independently patrols four rotor unmanned aerial vehicle based on machine vision |
| CN112069889A (en) * | 2020-07-31 | 2020-12-11 | 北京信息科技大学 | Civil aircraft trajectory prediction method, electronic device and storage medium |
| CN112179220A (en) * | 2020-09-22 | 2021-01-05 | 上海机电工程研究所 | Vehicle-mounted portable aircraft flying following equipment |
| CN112445238A (en) * | 2020-11-19 | 2021-03-05 | 西南大学 | Unmanned aerial vehicle tracking system based on neural network |
| CN112597799A (en) * | 2020-11-24 | 2021-04-02 | 北京理工大学 | Fixed wing aircraft flight trend judgment method based on vision |
| CN112731918A (en) * | 2020-11-30 | 2021-04-30 | 北京理工大学 | Ground unmanned platform autonomous following system based on deep learning detection tracking |
| CN112925572A (en) * | 2021-03-01 | 2021-06-08 | 联想(北京)有限公司 | Control method and device and electronic equipment |
| CN113306741A (en) * | 2021-04-16 | 2021-08-27 | 西安航空职业技术学院 | External winding inspection unmanned aerial vehicle and method based on deep learning |
| WO2024052928A1 (en) * | 2022-09-10 | 2024-03-14 | Patil Dipti | System for vision-based self-decisive planetary hazard free landing of a space vehicle |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103838244A (en) * | 2014-03-20 | 2014-06-04 | 湖南大学 | Portable target tracking method and system based on four-axis air vehicle |
| CN105487552A (en) * | 2016-01-07 | 2016-04-13 | 深圳一电航空技术有限公司 | Unmanned aerial vehicle tracking shooting method and device |
| CN106339006A (en) * | 2016-09-09 | 2017-01-18 | 腾讯科技（深圳）有限公司 | Object tracking method of aircraft and apparatus thereof |
| CN107087427A (en) * | 2016-11-30 | 2017-08-22 | 深圳市大疆创新科技有限公司 | Control method, device and the equipment and aircraft of aircraft |
-
2017
- 2017-10-16 CN CN201710961301.4A patent/CN107817820A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103838244A (en) * | 2014-03-20 | 2014-06-04 | 湖南大学 | Portable target tracking method and system based on four-axis air vehicle |
| CN105487552A (en) * | 2016-01-07 | 2016-04-13 | 深圳一电航空技术有限公司 | Unmanned aerial vehicle tracking shooting method and device |
| CN106339006A (en) * | 2016-09-09 | 2017-01-18 | 腾讯科技（深圳）有限公司 | Object tracking method of aircraft and apparatus thereof |
| CN107087427A (en) * | 2016-11-30 | 2017-08-22 | 深圳市大疆创新科技有限公司 | Control method, device and the equipment and aircraft of aircraft |
Non-Patent Citations (1)
| Title |
|---|
| 朱宇鑫: "基于神经网络的人脸朝向识别问题研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108615244A (en) * | 2018-03-27 | 2018-10-02 | 中国地质大学（武汉） | A kind of image depth estimation method and system based on CNN and depth filter |
| CN108764456A (en) * | 2018-04-03 | 2018-11-06 | 北京环境特性研究所 | Airborne target identification model construction platform, airborne target recognition methods and equipment |
| CN108764456B (en) * | 2018-04-03 | 2021-06-22 | 北京环境特性研究所 | Airborne target identification model construction platform, airborne target identification method and equipment |
| CN110347035A (en) * | 2018-04-08 | 2019-10-18 | 北京京东尚科信息技术有限公司 | Method for autonomous tracking and device, electronic equipment, storage medium |
| CN108985193A (en) * | 2018-06-28 | 2018-12-11 | 电子科技大学 | A kind of unmanned plane portrait alignment methods based on image detection |
| CN109062220A (en) * | 2018-08-31 | 2018-12-21 | 阿里巴巴集团控股有限公司 | The method and apparatus of controlling terminal movement |
| CN109062220B (en) * | 2018-08-31 | 2021-06-29 | 创新先进技术有限公司 | Method and device for controlling terminal movement |
| CN109164821A (en) * | 2018-09-26 | 2019-01-08 | 中科物栖（北京）科技有限责任公司 | A kind of UAV Attitude training method and device |
| CN109164821B (en) * | 2018-09-26 | 2019-05-07 | 中科物栖（北京）科技有限责任公司 | A kind of UAV Attitude training method and device |
| CN109885091A (en) * | 2019-03-21 | 2019-06-14 | 华北电力大学（保定） | A kind of unmanned plane autonomous flight control method and system |
| CN110104167A (en) * | 2019-04-25 | 2019-08-09 | 江苏大学 | A kind of automation search and rescue UAV system and control method using infrared thermal imaging sensor |
| CN110514206A (en) * | 2019-08-02 | 2019-11-29 | 中国航空无线电电子研究所 | A kind of unmanned plane during flying path prediction technique based on deep learning |
| CN110514206B (en) * | 2019-08-02 | 2023-08-04 | 中国航空无线电电子研究所 | Unmanned aerial vehicle flight path prediction method based on deep learning |
| CN110543182A (en) * | 2019-09-11 | 2019-12-06 | 济宁学院 | autonomous landing control method and system for small unmanned gyroplane |
| CN110543182B (en) * | 2019-09-11 | 2022-03-15 | 济宁学院 | Autonomous landing control method and system for small unmanned gyroplane |
| CN110879602A (en) * | 2019-12-06 | 2020-03-13 | 安阳全丰航空植保科技股份有限公司 | Unmanned aerial vehicle control law parameter adjusting method and system based on deep learning |
| CN111178148A (en) * | 2019-12-06 | 2020-05-19 | 天津大学 | Ground target geographic coordinate positioning method based on unmanned aerial vehicle vision system |
| CN111178148B (en) * | 2019-12-06 | 2023-06-02 | 天津大学 | Ground target geographic coordinate positioning method based on unmanned aerial vehicle vision system |
| CN110879602B (en) * | 2019-12-06 | 2023-04-28 | 安阳全丰航空植保科技股份有限公司 | Unmanned aerial vehicle control law parameter adjusting method and system based on deep learning |
| CN111204452B (en) * | 2020-02-10 | 2021-07-16 | 北京建筑大学 | Target detection system based on miniature aircraft |
| CN111204452A (en) * | 2020-02-10 | 2020-05-29 | 北京建筑大学 | Target detection system based on miniature aircraft |
| CN111824406A (en) * | 2020-07-17 | 2020-10-27 | 南昌航空大学 | Public safety independently patrols four rotor unmanned aerial vehicle based on machine vision |
| CN111736190B (en) * | 2020-07-24 | 2022-01-25 | 广东电网有限责任公司 | Unmanned aerial vehicle airborne target detection system and method |
| CN111736190A (en) * | 2020-07-24 | 2020-10-02 | 广东电网有限责任公司 | Unmanned aerial vehicle airborne target detection system and method |
| CN112069889B (en) * | 2020-07-31 | 2021-08-03 | 北京信息科技大学 | Civil aircraft trajectory prediction method, electronic device and storage medium |
| CN112069889A (en) * | 2020-07-31 | 2020-12-11 | 北京信息科技大学 | Civil aircraft trajectory prediction method, electronic device and storage medium |
| CN112179220A (en) * | 2020-09-22 | 2021-01-05 | 上海机电工程研究所 | Vehicle-mounted portable aircraft flying following equipment |
| CN112445238A (en) * | 2020-11-19 | 2021-03-05 | 西南大学 | Unmanned aerial vehicle tracking system based on neural network |
| CN112445238B (en) * | 2020-11-19 | 2021-09-10 | 西南大学 | Unmanned aerial vehicle tracking system based on neural network |
| CN112597799A (en) * | 2020-11-24 | 2021-04-02 | 北京理工大学 | Fixed wing aircraft flight trend judgment method based on vision |
| CN112731918A (en) * | 2020-11-30 | 2021-04-30 | 北京理工大学 | Ground unmanned platform autonomous following system based on deep learning detection tracking |
| CN112925572B (en) * | 2021-03-01 | 2023-05-23 | 联想(北京)有限公司 | Control method and device and electronic equipment |
| CN112925572A (en) * | 2021-03-01 | 2021-06-08 | 联想(北京)有限公司 | Control method and device and electronic equipment |
| CN113306741A (en) * | 2021-04-16 | 2021-08-27 | 西安航空职业技术学院 | External winding inspection unmanned aerial vehicle and method based on deep learning |
| WO2024052928A1 (en) * | 2022-09-10 | 2024-03-14 | Patil Dipti | System for vision-based self-decisive planetary hazard free landing of a space vehicle |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107817820A (en) | A kind of unmanned plane autonomous flight control method and system based on deep learning | |
| US11635775B2 (en) | Systems and methods for UAV interactive instructions and control | |
| US11748898B2 (en) | Methods and system for infrared tracking | |
| US10650235B2 (en) | Systems and methods for detecting and tracking movable objects | |
| CN103365297B (en) | Based on four rotor wing unmanned aerial vehicle flight control methods of light stream | |
| CN105929850B (en) | A kind of UAV system and method with lasting locking and tracking target capability | |
| Mademlis et al. | Autonomous UAV cinematography: A tutorial and a formalized shot-type taxonomy | |
| CN103587708B (en) | The microminiature unmanned rotary wing aircraft field zero autonomous soft landing method in blind area of fixing a point | |
| KR102254491B1 (en) | Automatic fly drone embedded with intelligent image analysis function | |
| CN110494360A (en) | For providing the autonomous system and method photographed and image | |
| CN112130579A (en) | Tunnel unmanned aerial vehicle inspection method and system | |
| CN103438869B (en) | A kind of large scale vegetation coverage aviation Dynamic Acquisition system | |
| US20220137647A1 (en) | System and method for operating a movable object based on human body indications | |
| CN107209514A (en) | The selectivity processing of sensing data | |
| CN201217501Y (en) | Suspending type aviation camera shooting self-determination aircraft system | |
| CN106155082B (en) | A kind of unmanned plane bionic intelligence barrier-avoiding method based on light stream | |
| CN105182992A (en) | Unmanned aerial vehicle control method and device | |
| CN111824406A (en) | Public safety independently patrols four rotor unmanned aerial vehicle based on machine vision | |
| CN104777847A (en) | Unmanned aerial vehicle target tracking system based on machine vision and ultra-wideband positioning technology | |
| CN102190081A (en) | Vision-based fixed point robust control method for airship | |
| CN105334347A (en) | Particle image velocimetry system and method based on unmanned plane | |
| CN108848348A (en) | A kind of crowd's abnormal behaviour monitoring device and method based on unmanned plane | |
| CN202879792U (en) | Suspension type aviation camera shooting and tracing autonomous aircraft system | |
| CN109709975A (en) | A kind of quadrotor indoor security system and method for view-based access control model SLAM | |
| CN108803652A (en) | A kind of autonomous tracking control method of rotor craft |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180320 |