CN108089695A - A kind of method and apparatus for controlling movable equipment - Google Patents
A kind of method and apparatus for controlling movable equipment Download PDFInfo
- Publication number
- CN108089695A CN108089695A CN201611051269.8A CN201611051269A CN108089695A CN 108089695 A CN108089695 A CN 108089695A CN 201611051269 A CN201611051269 A CN 201611051269A CN 108089695 A CN108089695 A CN 108089695A
- Authority
- CN
- China
- Prior art keywords
- human body
- target
- movable equipment
- body attitude
- motion vector
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/011—Arrangements for interaction with the human body, e.g. for user immersion in virtual reality
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/017—Gesture based interaction, e.g. based on a set of recognized hand gestures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/28—Recognition of hand or arm movements, e.g. recognition of deaf sign language
Abstract
An embodiment of the present invention provides it is a kind of control movable equipment method and apparatus, the described method includes:Image is gathered by image acquisition units, the image collected is identified, obtains recognition result;It is the first human body attitude that the recognition result, which is represented in described the image collected there are human body and human body attitude,;The corresponding first sports immunology information of first human body attitude and the currently performed second sports immunology information of the movable equipment are overlapped, obtain the 3rd sports immunology information;The corresponding target moving parameter of the 3rd sports immunology information is calculated, and the movable equipment is controlled to be moved with the target moving parameter.The present invention is used to implement the technique effect that movable equipment is controlled by human body attitude, into without additional controller.
Description
Technical field
The present invention relates to electronic technology field more particularly to a kind of method and apparatus for controlling movable equipment.
Background technology
At present, the movement of correlation technique control movable equipment, one of which is the controller held by receiving user,
Corresponding control is performed such as remote controler or the control signal of mobile phone transmission.For example, user presses left key on a remote control, then
Movable equipment exports according to received signal and responds left shift instruction, and then is moved to the left.
However, above-mentioned control mode needs to increase additional controller.
The content of the invention
An embodiment of the present invention provides a kind of method and apparatus for controlling movable equipment, are used to implement and pass through human body attitude
The technique effect of movable equipment is controlled, into without additional controller.
In a first aspect, the present invention provides it is a kind of control movable equipment method, including:
Image is gathered by image acquisition units, the image collected is identified, obtains recognition result;The identification
As a result represent in described the image collected there are human body and human body attitude to be the first human body attitude;
By the corresponding first sports immunology information of first human body attitude and the movable equipment currently performed
Two sports immunology information are overlapped, and obtain the 3rd sports immunology information;
The corresponding target moving parameter of the 3rd sports immunology information is calculated, and controls the movable equipment with institute
State the movement of target moving parameter.
Optionally, the image collected is identified, obtains recognition result, including:
Human testing is carried out in described the image collected, obtains testing result;
When the testing result is represented in described the image collected there are during human body, by the target where the human body with
Track region inputs multiple human body attitude graders and is matched, and obtains the first matching of each human body attitude grader output
Fraction;
Determine that corresponding first human body attitude of the highest first matching fraction is the human body attitude.
Optionally, the second sports immunology information tracks the sports immunology information of a target for the movable equipment,
The target is the human body, and the method further includes:
Judge whether the movable equipment loses target;
When the movable equipment loses target, the target is found again.
Optionally, judge whether the movable equipment loses target, including:
Utilize the target following regional training object module of the first two field picture in described the image collected;
Based on the object module, to the target following region of the second two field picture after first two field picture into
Row matching obtains the second matching fraction;
Judge whether the second matching fraction reaches threshold value;Wherein, when the described second matching fraction is not up to the threshold
During value, represent that the movable equipment loses target.
Optionally, after the target is found again, further include:
When searching out suspected target again, based on the object module, to the suspected target where the suspected target
Tracing area is matched, and obtains the 3rd matching fraction;
When the described 3rd matching fraction reaches the threshold value, it is the target to determine the suspected target.
Optionally, the sports immunology information is motion vector, and corresponding first movement of first human body attitude is retouched
It states information and the currently performed second sports immunology information of the movable equipment is overlapped, obtain the 3rd sports immunology letter
Breath, including:
By corresponding first motion vector of first human body attitude and the currently performed institute of the movable equipment
The superposition of the second motion vector is stated, obtains the 3rd motion vector;
Wherein, the size and Orientation of first motion vector is determined according to first human body attitude;Described second
The size of motion vector determines that direction is determined according to the direction of motion according to speed.
Second aspect, the present invention provides it is a kind of control movable equipment device, including:
Identification module for gathering image by image acquisition units, is identified the image collected, is identified
As a result;It is the first human body attitude that the recognition result, which is represented in described the image collected there are human body and human body attitude,;
Laminating module, for by the corresponding first sports immunology information of first human body attitude and the movable equipment
Currently performed second sports immunology information is overlapped, and obtains the 3rd sports immunology information;
Control module, for calculating the corresponding target moving parameter of the 3rd sports immunology information, and described in control
Movable equipment is moved with the target moving parameter.
Optionally, the identification module obtains testing result for carrying out human testing in described the image collected;
It is when the testing result is represented in described the image collected there are during human body, the target following region where the human body is defeated
Enter multiple human body attitude graders to be matched, obtain the first matching fraction of each human body attitude grader output;Really
It is the human body attitude to determine corresponding first human body attitude of the highest first matching fraction.
Optionally, the second sports immunology information tracks the sports immunology information of a target for the movable equipment,
The target is the human body, and described device further includes:
Judgment module, for judging whether the movable equipment loses target;
Module is found, for when the movable equipment loses target, finding the target again.
Optionally, the judgment module be used for the target using the first two field picture in described the image collected with
Track regional training object module;Based on the object module, to the mesh of the second two field picture after first two field picture
Mark tracing area is matched, and obtains the second matching fraction;Judge whether the second matching fraction reaches threshold value;Wherein, when
When the second matching fraction is not up to the threshold value, represent that the movable equipment loses target.
Optionally, described device further includes:
Module is obtained, for after the target is found again, when searching out suspected target again, based on the mesh
Model is marked, the suspected target tracing area where the suspected target is matched, obtains the 3rd matching fraction;
Determining module, for when the described 3rd matching fraction reaches the threshold value, it to be described to determine the suspected target
Target.
Optionally, the sports immunology information is motion vector, and the laminating module is used for first human body attitude
Corresponding first motion vector is superimposed with currently performed second motion vector of the movable equipment, obtains the 3rd
Motion vector;Wherein, the size and Orientation of first motion vector is determined according to first human body attitude;Described second
The size of motion vector determines that direction is determined according to the direction of motion according to speed.
Said one or multiple technical solutions in the embodiment of the present application at least have following one or more technology effects
Fruit:
In the inventive solutions, image is gathered by image acquisition units first, the image collected is carried out
Identification obtains recognition result, wherein, it is first that recognition result, which is represented in the image collected there are human body and human body attitude,
Human body attitude, then, by the second current movement of the corresponding first sports immunology information of the first human body attitude and movable equipment
Description information is overlapped, and obtains the 3rd sports immunology information, finally, calculates the corresponding target moving parameter of the 3rd movement,
And movable equipment is controlled to be moved with target moving parameter.As it can be seen that the present invention is by identifying human body attitude, so will with it is the first
The corresponding first sports immunology information superposition of body posture is set on the second sports immunology information of itself, then control is removable
It is standby to be moved with the corresponding target moving parameter of the 3rd sports immunology information after superposition.It is set so the present invention is removable
Standby accordingly to be controlled according to human body attitude, user can control the movement of movable equipment by making different postures,
And additional controller need not be increased.
Description of the drawings
Fig. 1 is that the removable method flow diagram set is controlled in the embodiment of the present invention;
Fig. 2 is one second motion vector in the embodiment of the present inventionWith schematic diagram of the target in earth coordinates;
Fig. 3 a- Fig. 3 e are that motion vector is superimposed schematic diagram in the embodiment of the present invention;
Fig. 4 is the schematic device that movable equipment is controlled in the embodiment of the present invention.
Specific embodiment
An embodiment of the present invention provides a kind of method and apparatus for controlling movable equipment, are used to implement and pass through human body attitude
The technique effect of movable equipment is controlled, into without additional controller.
In order to solve the above technical problem, the present invention provides technical solution thinking it is as follows:
In the inventive solutions, image is gathered by image acquisition units first, the image collected is carried out
Identification obtains recognition result, wherein, it is first that recognition result, which is represented in the image collected there are human body and human body attitude,
Human body attitude, then, by the second current movement of the corresponding first sports immunology information of the first human body attitude and movable equipment
Description information is overlapped, and obtains the 3rd sports immunology information, finally, calculates the corresponding target moving parameter of the 3rd movement,
And movable equipment is controlled to be moved with target moving parameter.As it can be seen that the present invention is by identifying human body attitude, so will with it is the first
The corresponding first sports immunology information superposition of body posture is set on the second sports immunology information of itself, then control is removable
It is standby to be moved with the corresponding target moving parameter of the 3rd sports immunology information after superposition.It is set so the present invention is removable
Standby accordingly to be controlled according to human body attitude, user can control the movement of movable equipment by making different postures,
And additional controller need not be increased.
Technical solution of the present invention is described in detail below by attached drawing and specific embodiment, it should be understood that the application
Specific features in embodiment and embodiment are to the detailed description of technical scheme rather than to present techniques
The restriction of scheme, in the case where there is no conflict, the technical characteristic in the embodiment of the present application and embodiment can be mutually combined.
The terms "and/or", is only a kind of incidence relation for describing affiliated partner, and expression may have three kinds of passes
System, for example, A and/or B, can represent:Individualism A exists simultaneously A and B, these three situations of individualism B.In addition, herein
Middle character "/", it is a kind of relation of "or" to typically represent forward-backward correlation object.
First aspect present invention provides a kind of method for controlling movable equipment.Wherein, movable equipment is, for example, flat
Mobile equipment can occur for weighing apparatus vehicle, robot or unmanned plane etc., and the present invention is not particularly limited.It please refers to Fig.1, for the present invention
The method flow diagram of movable equipment is controlled in embodiment.This method includes:
S101:Image is gathered by image acquisition units, the image collected is identified, obtains recognition result;
S102:By the corresponding first sports immunology information of first human body attitude and the movable equipment it is current the
Two sports immunology information are overlapped, and obtain the 3rd sports immunology information;
S103:The corresponding target moving parameter of the 3rd sports immunology information is calculated, and described move is controlled to set
It is standby to be moved with the target moving parameter.
Specifically, the movable equipment in the embodiment of the present invention is equipped with image acquisition units, in specific implementation process
In, image acquisition units can be specially RGB (RGB, Red Green Blue) image acquisition units, depth image acquisition
Any one or more in unit, infrared image acquisition unit and binocular image collecting unit, the present invention do not do specific limit
System.The image in acquisition range is acquired by image acquisition units, and by the image collected be transferred to control it is removable
The device of dynamic equipment is identified.
In S101, the image of collecting unit is identified, obtains recognition result.Due to during specific implementation,
User may be not present in acquisition range, it is also possible to there are user, and at different moments the posture of human body be also possible to it is different,
Therefore recognition result, which is just specifically as follows, represents that there are do not deposited in human body or expression the image collected in the image collected
In human body.In order to facilitate description, it will be assumed that recognition result specifically represents that there are people in the image collected in the embodiment of the present invention
Body, and human body attitude is the first human body attitude.Wherein, the first human body attitude is, for example, that user waves to body interior direction
To posture of body lateral direction pushing hands etc., the present invention is not particularly limited by posture, user.
Next, in S102, the corresponding first sports immunology information of the first human body attitude and movable equipment is current
Second sports immunology information is overlapped, the 3rd sports immunology information after being superimposed.Specifically, in the embodiment of the present invention
Sports immunology information be intended to indicate that the information of movement, such as linear velocity on motion vector, angular speed, horizontal plane and vertical
Linear velocity on face etc..First sports immunology information and the second sports immunology information are overlapped, and then after being superimposed
Sports immunology information, i.e. the 3rd sports immunology information.
Next, in S103, the corresponding target moving parameter of the 3rd sports immunology information is calculated.Specifically, originally
Moving parameter in inventive embodiments includes but not limited to linear velocity, angular speed, in linear velocity and vertical plane on horizontal plane
Linear velocity etc., the present invention is not particularly limited.Target moving parameter is gone out by the 3rd sports immunology information backwards calculation, and then is controlled
Movable equipment processed is moved according to target moving parameter.
Since target moving parameter is obtained by the 3rd sports immunology information backwards calculation, and the 3rd movable information is
The corresponding first sports immunology superposition of the first human body attitude of user, so, control movable equipment is moved according to target
Parameter moves, and is achieved that the technique effect of response user's the first human body gesture stability movable equipment movement.So that with
When family needs to control movable equipment, it is no longer necessary to additional controller, by making different human body attitudes.
Specifically, the image collected is identified in S101 of the embodiment of the present invention, can specifically passes through following mistake
Cheng Shixian:
Human testing is carried out in described the image collected, obtains testing result;
When the testing result is represented in described the image collected there are during human body, by the target where the human body with
Track region inputs multiple human body attitude graders and is matched, and obtains the first matching of each human body attitude grader output
Fraction;
Determine that corresponding first human body attitude of the highest first matching fraction is the human body attitude.
Specifically, in order to carry out human testing or pedestrian detection in the image collected, there are two types of sides by the present invention
Formula.
The first:Human testing model is stored in memory space by movable equipment in advance, when needing to be identified
It is read out from memory space.During specific implementation, the human testing model of movable equipment storage can be specially
Any one or more in Face datection model, human body half body detection model and Whole Body detection model, the present invention are not done
Concrete restriction.Movable equipment has detected whether to meet human body in the image collected according to the feature of human testing model
The feature of detection model.If detecting the feature for meeting human testing model, obtain and represent exist in the image collected
The testing result of human body;, whereas if the feature for meeting human testing model is not detected, then expression is obtained in the figure collected
The testing result of human body is not present as in.
Second:In embodiments of the present invention, in order to carry out human testing to the image collected, can first be gathered
The ACF (converging channels feature, Aggregate Channel Feature) of the image arrived.ACF features are generally by 10 passage groups
Into, be respectively LUV colors 3 passages, 1 gradient magnitude passage and 6 gradient orientation histogram passages.Adjust gradient side
To quantization step can change gradient orientation histogram number of channels.LUV (L represents that brightness, U and V represent colourity) color is led to
Road can also add, delete or replace with other Color Channels, as RGB or HSV (colourity, purity, lightness, Hue,
Saturation, Value) etc..
During specific implementation, ACF features can be obtained according to universal method, i.e., different scale first are carried out to image
Scaling, then ACF features are calculated respectively.Alternatively, in order to improve efficiency, shortening processing time, and excessive precision is not lost, in this hair
In bright embodiment, accurate ACF features can also be calculated by shortening, other scales are then according to neighbouring on the scale of part
ACF features are estimated.
For example, it is necessary to which the scale of scaling is specially 1,7/8,6/8,5/8,4/8,3/8,2/8 and 1/8, the present invention is real
It applies in example and is zoomed in and out according to 1,5/8 and 2/8 scale, and the corresponding ACF channel characteristics of 1,5/8 and 2/8 scale are calculated.
Then, the corresponding ACF channel characteristics of 7/8 and 6/8 scale are estimated with 7/8 and 6/8 proportionate relationship according to 1, according to 5/8 and 4/
8 and 3/8 proportionate relationship estimates the corresponding ACF channel characteristics of 4/8 and 3/8 scale, estimates according to 2/8 with 1/8 proportionate relationship
The corresponding ACF channel characteristics of 1/8 scale are counted out, it is final to obtain whole ACF features.
After ACF features are calculated, judge to whether there is human body in the image collected using grader.In the present invention
In embodiment, the grader that may be employed include but not limited to adaptive enhancing Adaboost, soft cascade Soft Cascade and
Support vector machines Structural SVM (support vector machines, Support Vector Machine) of structuring etc., the present invention
It is not particularly limited.By taking Adaboost graders as an example, judge that the process in the image collected with the presence or absence of human body is specially:
Positive negative sample is extracted first and calculates ACF features, then trains Adaboost graders, is classified then according to existing Adaboost
Device extracts new negative sample and calculates new ACF features, then the new Adaboost graders of retraining.To each Weak Classifier
Output weights add up, when accumulator be less than -1 when be directly determined as negative sample, after sample is by all Weak Classifiers
Then it is determined as positive sample.Non-maxima suppression finally is done to all positive samples and obtains testing result.
During specific implementation, what those skilled in the art can be in actual selection above two realization method appoints
Meaning one kind detects human body, and the present invention is not particularly limited.
During specific implementation, movable equipment can also may be used using any human body detected as the target of tracking
With only to certain customers, such as owner of movable equipment etc. possesses the user of control movable equipment permission into line trace,
And the control of the human body attitude only in response to the user for possessing control movable equipment permission.So, further, people is being detected
After body, Authority Verification can also be carried out to the human body that detects, such as verify whether the face of the human body detected is advance
Face of user for the limit that is possessed of control power of storage etc..If the human body detected by Authority Verification, movable equipment with
Track the user, and further identify the human body attitude of the user, and then calculate the 3rd sports immunology information., whereas if
For the human body detected not by Authority Verification, then movable equipment will not follow the user, will not further identify the user
Human body attitude and calculate corresponding 3rd sports immunology information.
Further, if identifying multiple human bodies from the image collected, it is necessary to carry out permission to each human body
Verification.If only there are one by Authority Verification in multiple human bodies, movable equipment only keeps track the people by Authority Verification
Body, and further identify the human body attitude of the human body, and then calculate the 3rd sports immunology information.If have in multiple human bodies
Two or more human bodies are by Authority Verification, then movable equipment only keeps track the highest human body of permission or tracking is used
The human body of family selection, and further identify the human body attitude of the human body, and then calculate the 3rd sports immunology information.
Next, when testing result represent in the image collected there are during human body or detect with control it is removable
During the user of dynamic equipment permission, the target following region where human body is inputted into multiple human body attitude graders and is matched, is obtained
Obtain the first matching fraction of each human body attitude grader output.
Specifically, target following region is the region comprising human body, such as the boundary rectangle of human body or includes human body
Non- external rectangle etc., the present invention is not particularly limited.Multiple human body attitude graders are previously stored in movable equipment, often
A grader corresponds to a different human body attitude, for example, wave to body interior direction posture, push away to body lateral direction
The posture of hand, arm are directed toward the posture of movable equipment left, arm is directed toward the posture and two arms of movable equipment right
Be connected posture of cyclization etc..Target following region is inputted into each human body attitude grader, obtain target following region with it is each
First matching fraction of human body attitude grader.Wherein, the first matching fraction is higher, represents target following region and human body attitude
Grader more matches, and then represents the human body attitude in target following region closer to the people represented by the human body attitude grader
Body posture.
Therefore, in embodiments of the present invention, the highest first matching fraction is corresponded into the first human body attitude and is determined as target
The human body attitude of user in tracing area.
In addition, in embodiments of the present invention, S102 can be realized by following process:
By corresponding first motion vector of first human body attitude and the currently performed institute of the movable equipment
The superposition of the second motion vector is stated, obtains the 3rd motion vector;
Wherein, the size and Orientation of first motion vector is determined according to first human body attitude;Described second
The size of motion vector determines that direction is determined according to the direction of motion according to speed.
After the human body attitude for determining user is the first human body attitude, corresponding first movement of the first human body attitude is read
Vector.For the second motion vector, then gather the current movement of movable equipment direction and speed and movable equipment with
Then the distance between target and direction synthesize the second motion vector.Then, by the first motion vector and the second motion vector into
Row superposition, obtains the 3rd motion vector.Wherein, the 3rd motion vector is exactly the 3rd sports immunology information.
Finally, in S103, by the 3rd motion vector backwards calculation into target moving parameter, and movable equipment is controlled with mesh
Mark moving parameter is moved.
For example, movable equipment is specially robot, and the human body attitude being stored in robot is specially to body
Posture that interior direction is waved, to body lateral direction pushing hands posture, arm is directed toward the posture of robot left, arm is directed toward
The posture of robot right be connected with arm cyclization posture.Wherein, the posture waved to body interior direction corresponds to robot
The direction of human body is directed toward, speed is the motion vector of 2m/s;Human body family, which is corresponded to, to the posture of body lateral direction pushing hands is directed toward machine
The direction of device people, speed are the motion vector of 4m/s;Arm is directed toward the corresponding direction robot left of posture of robot left,
Speed is the motion vector of 2m/s;The posture that arm is directed toward robot right corresponds to direction robot right, and speed is 2m/s's
Motion vector;Arm be connected cyclization the corresponding motion vector of posture there are two, one to be identical with the second motion vector size,
And the first sub- motion vector that direction is opposite, another is using human body as the center of circle, using the tangent line with human body distance as the circle of radius
Direction, speed be 2m/s the second sub- motion vector.
Robot is located at tracking user behind, is moved with the speed tracing user of 1m/s, then the second motion vector
Human body, velocity magnitude 1m/s are directed toward for robot in direction.User turns round, and makes the posture waved to body interior direction.
Robot determines to identify that human body attitude is the posture waved to body interior direction by identifying the image collected, and then
Read the first motion vector corresponding with the posture waved to body interior direction.Then, the first motion vector and second are transported
Dynamic vector is overlapped, and is obtained direction and is directed toward human body, the 3rd motion vector of speed 3m/s for robot.Then, backwards calculation
Go out the corresponding target moving parameter of the 3rd motion vector for linear velocity+3m/s and angular speed 0rad/s.So when user make to
The posture that body interior direction is waved, robot will accelerate close to user.
Alternatively, user turns round, the posture to body lateral direction pushing hands is made.Robot is by identifying the figure collected
Picture, it is the posture to body lateral direction pushing hands to determine to identify human body attitude, so read with to body lateral direction pushing hands
Corresponding first motion vector of posture.Then, the first motion vector and the second motion vector are overlapped, obtaining direction is
Human body is directed toward robot, the 3rd motion vector of speed 3m/s.Then, backwards calculation goes out the corresponding target of the 3rd motion vector and moves
Dynamic parameter is linear velocity -3m/s and angular speed 0rad/s.So when user makes the posture to body lateral direction pushing hands, machine
Device people will slow down close to user or even retreat.
Alternatively, user turns round, the posture that arm is directed toward robot left is made.Robot is by identifying the figure collected
Picture determines to identify the posture that human body attitude is directed toward robot left for arm, and then reads and be directed toward robot left with arm
Corresponding first motion vector of posture.Then, the first motion vector and the second motion vector are overlapped, obtain the side that moves ahead
To 63.5 ° to the left, the 3rd motion vector of speed 2.23m/s.Then, backwards calculation goes out the corresponding target of the 3rd motion vector and moves
Dynamic parameter is linear velocity+2.23m/s and angular speed+0.46rad/s.So it is directed toward robot left when user makes arm
Posture, robot will be moved into user left back into line trace.
Alternatively, user turns round, the posture that arm is directed toward robot right is made.Robot is by identifying the figure collected
Picture determines to identify the posture that human body attitude is directed toward robot right for arm, and then reads and be directed toward robot right with arm
Corresponding first motion vector of posture.Then, the first motion vector and the second motion vector are overlapped, obtain the side that moves ahead
To 63.5 ° to the right, the 3rd motion vector of speed 2.23m/s.Then, backwards calculation goes out the corresponding target of the 3rd motion vector and moves
Dynamic parameter is linear velocity+2.23m/s and angular speed -0.46rad/s.So it is directed toward robot right when user makes arm
Posture, robot will be moved into user right back into line trace.
Alternatively, user turns round, make two arms and be connected the posture of cyclization.Robot is by identifying the figure collected
Picture determines to identify that human body attitude is connected the posture of cyclization for arm, and then it is corresponding to read the cyclic posture that is connected with arm
First sub- motion vector and the second sub- motion vector.Then, the first sub- dynamic vector, the second sub- motion vector and the second movement are sweared
Amount is overlapped, and is obtained to be the tangential direction of the circle of radius for direction with human body distance, the 3rd movement arrow that speed is 2m/s
Amount.Then, backwards calculation goes out the corresponding target moving parameter of the 3rd motion vector for linear velocity -2m/s and angular speed+1.5rad/
s.So when user make arm be connected cyclization posture after, robot will surround tracking user rotate.
It should be appreciated that above-mentioned movable equipment is specially in several examples cited by robot, linear velocity is specific
For the speed in the horizontal plane parallel to the earth, the "+" before linear velocity represents the driving device of driving machine human hair life linear velocity
Forward drive, "-" represent the driving device reverse drive of driving machine human hair life linear velocity.Similar, the "+" table before angular speed
Show the driving device forward drive of driving machine human hair the male character types in Chinese operas, usu. referring tov the bearded character speed, "-" represents the driving dress of driving machine human hair the male character types in Chinese operas, usu. referring tov the bearded character speed
Put reverse drive.
Again for example, movable equipment is specially unmanned plane, and the human body attitude being stored in unmanned plane is specially to body
Posture that internal side direction is waved, to body lateral direction pushing hands posture, arm is directed toward the posture of unmanned plane left, arm refers to
Be connected to the posture of unmanned plane right with two arms the posture of cyclization.Wherein, corresponded to the posture that body interior direction is waved
Unmanned plane is directed toward human body direction, and speed is the motion vector of 2m/s;Human body is corresponded to the posture of body lateral direction pushing hands to be directed toward
Unmanned plane direction, speed are the motion vector of 4m/s;Arm is directed toward the corresponding direction of posture of unmanned plane left (in horizontal plane)
Unmanned plane left, speed are the motion vector of 2m/s (in horizontal plane);The posture that arm is directed toward unmanned plane right corresponds to direction nothing
Man-machine right, speed are the motion vector of 2m/s (in horizontal plane);Arm be connected cyclization the corresponding motion vector of posture have two
It is a, one to be identical with the second motion vector size, and the first sub- motion vector that direction is opposite, another is using human body to be round
The heart, using with human body distance be radius and height for the circle of flying height tangential direction as direction, speed is the second sub of 2m/s
Motion vector.
Unmanned plane is currently located at the positive 2m height of user, is moved with the speed tracing user of 1m/s.It, please for convenience of expression
It is the second motion vector with reference to figure 2With schematic diagram of the human body in earth coordinates.User makes to body interior direction and recruiting
The posture of hand, unmanned plane determine to identify that human body attitude is waved to body interior direction by identifying the image collected
Posture, and then read the first motion vector corresponding with the posture waved to body interior directionThen, the first movement is sweared
AmountWith the second motion vectorIt is overlapped, obtains the 3rd motion vector as shown in Figure 3aThen, backwards calculation goes out
Three motion vectorsCorresponding target moving parameter for linear velocity (in horizontal plane)+3.2m/s, linear velocity (in vertical plane)-
1.5m/s and angular speed 0rad/s.So when user makes the posture waved to body interior direction, unmanned plane will be close to use
Family.
Alternatively, user makes the posture to body lateral direction pushing hands, unmanned plane is determined by identifying the image collected
It is the posture to body lateral direction pushing hands to identify human body attitude, and then is read and the posture pair to body lateral direction pushing hands
The first motion vector answeredThen, by the first motion vectorWith the second motion vectorIt is overlapped, obtains such as Fig. 3 b institutes
The 3rd motion vector shownThen, backwards calculation goes out the 3rd motion vectorCorresponding target moving parameter is linear velocity (water
In plane) -1.2m/s, linear velocity (in vertical plane)+1.5m/s and angular speed 0rad/s.So when user is made outside body
The posture of side direction pushing hands, unmanned plane will be reduced close to user, even remotely from user.
Alternatively, user makes the posture that arm is directed toward unmanned plane left, unmanned plane is determined by identifying the image collected
It identifies that human body attitude is directed toward the posture of unmanned plane left for arm, and then reads the posture pair that unmanned plane left is directed toward with arm
The first motion vector answeredThen, by the first motion vectorWith the second motion vectorIt is overlapped, obtains such as Fig. 3 c institutes
The 3rd motion vector shownThen, backwards calculation goes out the 3rd motion vectorCorresponding target moving parameter is linear velocity (water
In plane)+2.23m/s, linear velocity (in vertical plane) 0m/s and angular speed+0.46rad/s.So refer to when user makes arm
To the posture of unmanned plane left, unmanned plane will fly to the left.
Alternatively, for making the posture that arm is directed toward unmanned plane right, unmanned plane is determined by identifying the image collected
It identifies that human body attitude is directed toward the posture of unmanned plane right for arm, and then reads the posture pair that unmanned plane right is directed toward with arm
The first motion vector answeredThen, by the first motion vectorWith the second motion vectorIt is overlapped, obtains such as Fig. 3 d institutes
The 3rd motion vector shownThen, backwards calculation goes out the 3rd motion vectorCorresponding target moving parameter is linear velocity (water
In plane)+2.23m/s, linear velocity (in vertical plane) 0m/s and angular speed -0.46rad/s.So refer to when user makes arm
To the posture of unmanned plane right, unmanned plane will fly to the right.
It is connected the posture of cyclization alternatively, user makes two arms, unmanned plane determines to know by identifying the image collected
Do not go out human body attitude for arm be connected cyclization posture, and then read be connected with arm cyclization posture it is corresponding first son move
VectorWith the second sub- motion vectorThen, by the first sub- motion vectorSecond sub- motion vectorWith the second movement
VectorIt is overlapped, obtains the 3rd motion vector as shown in Figure 3 eThen, backwards calculation goes out the 3rd motion vectorIt is right
The target moving parameter answered is linear velocity (in horizontal plane)+2m/s, linear velocity (vertical direction) is 0m/s and angular speed+
1.5rad/s.So when user make arm be connected cyclization posture after, unmanned plane is in current flight height, around user
Corresponding position flight.
It should be appreciated that above-mentioned movable equipment is specially in several examples cited by unmanned plane, linear velocity is (horizontal
In face) it is specially parallel to the speed in the horizontal plane of the earth, the "+" before linear velocity (in horizontal plane) represents that driving unmanned plane exists
The driving device forward drive of linear velocity occurs in horizontal plane, "-" represents that linear velocity occurs in the horizontal plane for driving unmanned plane
Driving device reverse drive.Speed of the linear velocity (in vertical plane) specially in the face of the earth, linear velocity (vertical plane
It is interior) before "+" represent that the driving device forward drive of linear velocity occurs in vertical plane for driving unmanned plane, unmanned plane during flying at this time
Highly increase, "-" represents that the driving device reverse drive of linear velocity occurs in vertical plane for driving unmanned plane, and unmanned plane flies at this time
Row height reduces.Similar, the "+" before angular speed represents the driving device forward drive of driving machine human hair the male character types in Chinese operas, usu. referring tov the bearded character speed, "-"
Represent the driving device reverse drive of driving machine human hair the male character types in Chinese operas, usu. referring tov the bearded character speed.
User passes through different human body attitudes it can be seen from foregoing description, it is possible to control the movement of movable equipment
Direction and speed, so that when user needs to control movable equipment, it is no longer necessary to additional controller, by doing
Go out different human body attitudes.
As a kind of optional embodiment, the method in the embodiment of the present invention can further include:
Judge whether the movable equipment loses target;
When the movable equipment loses target, the target is found again.
Specifically, the description that the second sports immunology information moves for movable equipment tracking target in the embodiment of the present invention
Information, and the target of the tracking is human body, i.e. user or the user with control movable equipment permission.Work as movable equipment
After identifying human body in S101, using the human body as target into line trace.In tracking, movable equipment is tracked according to vision
Algorithm comes to target into line trace.Wherein, the Vision Tracking of the embodiment of the present invention can be DSST (metric space track,
Discriminative Scale Space Tracker) algorithm or SRDCF (normalization correlation filtering in space tracks,
Spatially Regularization Discriminative Correlation Filter) algorithm etc., belonging to the present invention
The those of ordinary skill in field can be according to actually making choice, and the present invention is not particularly limited.
Since human motion has erratic behavior and unpredictability, so during movable equipment tracks human body,
It is possible to lose target.Therefore, in embodiments of the present invention, it is also necessary to judge whether movable equipment loses target.For
How to judge whether movable equipment loses target, can be realized by following process:
Utilize the target following regional training object module of the first two field picture in described the image collected;
Based on the object module, to the target following region of the second two field picture after first two field picture into
Row matching obtains the second matching fraction;
Judge whether the second matching fraction reaches threshold value;Wherein, when the described second matching fraction is not up to the threshold
During value, represent that the movable equipment loses target.
Specifically, the first two field picture in the embodiment of the present invention refers to this any that process is followed to recognize human body
Two field picture or any two field picture for recognizing the face with control movable equipment permission.Second two field picture is then the
The next two field picture collected after one two field picture.The target following region of first two field picture is inputted into Vision Tracking, instruction
Practice object module.
For by taking Vision Tracking is DSST as an example.DSST be a kind of online target following based on correlation filter with
Model training method.By the filter model for training the first two field picture Input Online, according to output in response to determining that in image
The position of target and scale.The filter model parameters of on-line training are by minimizing training sample output and desired output
Determined by error.Wherein, target following region is training sample.Extract FHoG (the direction ladders of fusion in target following region
Spend histogram, Fused Histogram of Gradient) feature, and establish high as the two dimension of peak value using FHoG eigencenters
This function is as desired output.Then, calculated by least square method and minimum wave filter between response and desired output
Error, so that it is determined that the parameters of the filter model of on-line training.And then the parameters of filter model are changed
For the value determined, and the on-line training wave filter after the completion of parameter is changed is as object module.
Then, FHoG features, the filter that region of search Input Online is trained are extracted in the region of search of the second two field picture
Ripple device model obtains the target following regional location and scale of the second two field picture.Then by the target following area of the second two field picture
Domain adds in training sample, updates the filter model of on-line training.
Next, based on object module, i.e. the filter model for the on-line training that training obtains and for the second time for the first time
The matching degree of the filter model for the on-line training that training obtains is calculated, and obtains the second matching fraction.Alternatively, calculate the
The distance of the position in the target following region of one two field picture and the position in the target following region of the second two field picture, by two targets
The distance of tracing area is determined as the second matching fraction.Alternatively, the average value by the characteristic peaks of the second two field picture and non-peak
Ratio be determined as the second matching fraction.Next, it is judged that whether the second matching fraction reaches threshold value.If the second matching fraction
Reach threshold value, then it represents that the target found twice is more similar, i.e., detects human body twice or detect same tool
There is the user of control movable equipment permission, it is determined that movable equipment does not lose target;, whereas if the second matching fraction is not
Reach threshold value, then it represents that the target found twice is dissimilar, i.e. human body or the second two field picture is not detected in the second two field picture
In be not detected with having the identical user of user of control movable equipment permission in the first two field picture, it is determined that it is removable to set
It is standby to lose target.
In embodiments of the present invention, after movable equipment loses target, S101 is performed again, so as to find mesh again
Mark.
Judge whether movable equipment is lost during movable equipment tracks human body it can be seen from foregoing description
Target, and target is found again when losing target, so as to provide movable equipment tracking reliability.
Further, after target is searched out again, can also include:
When searching out suspected target again, based on the object module, to the suspected target where the suspected target
Tracing area is matched, and obtains the 3rd matching fraction;
When the described 3rd matching fraction reaches the threshold value, it is the target to determine the suspected target.
Specifically, the suspected target in the embodiment of the present invention is to determine that movable equipment re-recognizes after losing target
Possibility be target human body.After suspected target is searched out, in order to confirm whether the suspected target searched out at this time is needs
The target of tracking, movable equipment are based further on object module, are carried out in the suspected target tracing area where doubtful mark
Matching, and then obtain the 3rd matching fraction.In the present invention is implemented, the method for obtaining the 3rd matching fraction is matched with obtaining second
Fraction is similar, therefore just it is no longer repeated herein.
When the 3rd matching fraction reaches threshold value, represent that the suspected target currently searched out is high with target degree of closeness, because
Suspected target is determined as target by this, restarts to track target.When the 3rd matching fraction is not up to threshold value, expression is currently sought
The suspected target found is low with target degree of closeness, it is thus determined that suspected target is not target, abandons the suspected target, seeks again
Look for other suspected targets.
Seen from the above description, when searching out suspected target again, judge whether suspected target is target, is further carried
The reliability that movable equipment has been supplied to track.
Based on the same inventive concept of method with controlling movable equipment in previous embodiment, second aspect of the present invention is also
A kind of device for controlling movable equipment is provided, as shown in figure 4, including:
Identification module 101 for gathering image by image acquisition units, is identified the image collected, obtains
Recognition result;It is the first human body appearance that the recognition result, which is represented in described the image collected there are human body and human body attitude,
State;
Laminating module 102, for by the corresponding first sports immunology information of first human body attitude and described removable
The currently performed second sports immunology information of equipment is overlapped, and obtains the 3rd sports immunology information;
Control module 103 for calculating the corresponding target moving parameter of the 3rd sports immunology information, and controls institute
Movable equipment is stated to move with the target moving parameter.
Specifically, identification module 101 obtains detection knot for carrying out human testing in described the image collected
Fruit;When the testing result is represented in described the image collected there are during human body, by the target following area where the human body
Domain inputs multiple human body attitude graders and is matched, and obtains the first matching point of each human body attitude grader output
Number;Determine that corresponding first human body attitude of the highest first matching fraction is the human body attitude.
Wherein, the second sports immunology information tracks the sports immunology information of a target for movable equipment, and target is human body,
Further, the device in the embodiment of the present invention further includes:
Judgment module, for judging whether the movable equipment loses target;
Module is found, for when the movable equipment loses target, finding the target again.
Wherein, judgment module is used for the target following region using the first two field picture in described the image collected
Training objective model;Based on the object module, the target following to the second two field picture after first two field picture
Region is matched, and obtains the second matching fraction;Judge whether the second matching fraction reaches threshold value;Wherein, when described
When two matching fractions are not up to the threshold value, represent that the movable equipment loses target.
Further, the device in the embodiment of the present invention further includes:
Module is obtained, for after the target is found again, when searching out suspected target again, based on the mesh
Model is marked, the suspected target tracing area where the suspected target is matched, obtains the 3rd matching fraction;
Determining module, for when the described 3rd matching fraction reaches the threshold value, it to be described to determine the suspected target
Target.
In embodiments of the present invention, sports immunology information is motion vector, and laminating module 102 is used for first human body
Corresponding first motion vector of posture is superimposed with currently performed second motion vector of the movable equipment, is obtained
3rd motion vector;Wherein, the size and Orientation of first motion vector is determined according to first human body attitude;It is described
The size of second motion vector determines that direction is determined according to the direction of motion according to speed.
The various change mode and specific example of the method for control movable equipment in earlier figures 1- Fig. 3 e embodiments are same
Sample is suitable for the device of the control movable equipment of the present embodiment, is retouched by foregoing the detailed of method to controlling movable equipment
It stating, those skilled in the art are clear that the implementation for the device that movable equipment is controlled in the present embodiment, so
In order to illustrate the succinct of book, this will not be detailed here.
Said one or multiple technical solutions in the embodiment of the present application at least have following one or more technology effects
Fruit:
In the inventive solutions, image is gathered by image acquisition units first, the image collected is carried out
Identification obtains recognition result, wherein, it is first that recognition result, which is represented in the image collected there are human body and human body attitude,
Human body attitude, then, by the second current movement of the corresponding first sports immunology information of the first human body attitude and movable equipment
Description information is overlapped, and obtains the 3rd sports immunology information, finally, calculates the corresponding target moving parameter of the 3rd movement,
And movable equipment is controlled to be moved with target moving parameter.As it can be seen that the present invention is by identifying human body attitude, so will with it is the first
The corresponding first sports immunology information superposition of body posture is set on the second sports immunology information of itself, then control is removable
It is standby to be moved with the corresponding target moving parameter of the 3rd sports immunology information after superposition.It is set so the present invention is removable
Standby accordingly to be controlled according to human body attitude, user can control the movement of movable equipment by making different postures,
And additional controller need not be increased.
It should be understood by those skilled in the art that, the embodiment of the present invention can be provided as method, system or computer program
Product.Therefore, the reality in terms of complete hardware embodiment, complete software embodiment or combination software and hardware can be used in the present invention
Apply the form of example.Moreover, the computer for wherein including computer usable program code in one or more can be used in the present invention
The computer program production that usable storage medium is implemented on (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
The form of product.
The present invention be with reference to according to the method for the embodiment of the present invention, the flow of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that it can be realized by computer program instructions every first-class in flowchart and/or the block diagram
The combination of flow and/or box in journey and/or box and flowchart and/or the block diagram.These computer programs can be provided
The processor of all-purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices is instructed to produce
A raw machine so that the instruction performed by computer or the processor of other programmable data processing devices is generated for real
The device for the function of being specified in present one flow of flow chart or one box of multiple flows and/or block diagram or multiple boxes.
These computer program instructions, which may also be stored in, can guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works so that the instruction generation being stored in the computer-readable memory includes referring to
Make the manufacture of device, the command device realize in one flow of flow chart or multiple flows and/or one box of block diagram or
The function of being specified in multiple boxes.
These computer program instructions can be also loaded into computer or other programmable data processing devices so that counted
Series of operation steps is performed on calculation machine or other programmable devices to generate computer implemented processing, so as in computer or
The instruction offer performed on other programmable devices is used to implement in one flow of flow chart or multiple flows and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art
God and scope.In this way, if these modifications and changes of the present invention belongs to the scope of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to comprising including these modification and variations.
Claims (12)
- A kind of 1. method for controlling movable equipment, which is characterized in that including:Image is gathered by image acquisition units, the image collected is identified, obtains recognition result;The recognition result Represent in described the image collected there are human body and human body attitude to be the first human body attitude;By currently performed second fortune of the corresponding first sports immunology information of first human body attitude and the movable equipment Dynamic description information is overlapped, and obtains the 3rd sports immunology information;The corresponding target moving parameter of the 3rd sports immunology information is calculated, and controls the movable equipment with the mesh Mark moving parameter movement.
- 2. the method as described in claim 1, which is characterized in that be identified to the image collected, obtain recognition result, bag It includes:Human testing is carried out in described the image collected, obtains testing result;When the testing result is represented in described the image collected there are during human body, by the target following area where the human body Domain inputs multiple human body attitude graders and is matched, and obtains the first matching point of each human body attitude grader output Number;Determine that corresponding first human body attitude of the highest first matching fraction is the human body attitude.
- 3. method as claimed in claim 2, which is characterized in that the second sports immunology information for the movable equipment with The sports immunology information of one target of track, the target are the human body, and the method further includes:Judge whether the movable equipment loses target;When the movable equipment loses target, the target is found again.
- 4. method as claimed in claim 3, which is characterized in that judge whether the movable equipment loses target, including:Utilize the target following regional training object module of the first two field picture in described the image collected;Based on the object module, the target following region progress to the second two field picture after first two field picture Match somebody with somebody, obtain the second matching fraction;Judge whether the second matching fraction reaches threshold value;Wherein, when the described second matching fraction is not up to the threshold value, Represent that the movable equipment loses target.
- 5. method as claimed in claim 4, which is characterized in that after the target is found again, further include:When searching out suspected target again, based on the object module, the suspected target where the suspected target is tracked Region is matched, and obtains the 3rd matching fraction;When the described 3rd matching fraction reaches the threshold value, it is the target to determine the suspected target.
- 6. the method as described in claim 1, which is characterized in that the sports immunology information is motion vector, by described first The corresponding first sports immunology information of human body attitude and the currently performed second sports immunology information of the movable equipment carry out Superposition obtains the 3rd sports immunology information, including:By corresponding first motion vector of first human body attitude and the movable equipment currently performed described the Two motion vectors are superimposed, and obtain the 3rd motion vector;Wherein, the size and Orientation of first motion vector is determined according to first human body attitude;Second movement The size of vector determines that direction is determined according to the direction of motion according to speed.
- 7. a kind of device for controlling movable equipment, which is characterized in that including:Identification module for gathering image by image acquisition units, is identified the image collected, obtains identification knot Fruit;It is the first human body attitude that the recognition result, which is represented in described the image collected there are human body and human body attitude,;Laminating module, for the corresponding first sports immunology information of first human body attitude and the movable equipment is current The the second sports immunology information performed is overlapped, and obtains the 3rd sports immunology information;Control module for calculating the corresponding target moving parameter of the 3rd sports immunology information, and controls described removable Dynamic equipment is moved with the target moving parameter.
- 8. device as claimed in claim 7, which is characterized in that the identification module be used in described the image collected into Row human testing obtains testing result;When the testing result is represented in described the image collected there are during human body, will described in Target following region where human body inputs multiple human body attitude graders and is matched, and obtains each human body attitude classification First matching fraction of device output;Determine that corresponding first human body attitude of the highest first matching fraction is the people Body posture.
- 9. device as claimed in claim 8, which is characterized in that the second sports immunology information for the movable equipment with The sports immunology information of one target of track, the target are the human body, and described device further includes:Judgment module, for judging whether the movable equipment loses target;Module is found, for when the movable equipment loses target, finding the target again.
- 10. device as claimed in claim 9, which is characterized in that the judgment module is used to utilize described the image collected In the first two field picture the target following regional training object module;Based on the object module, to the first frame figure The target following region of the second two field picture as after is matched, and obtains the second matching fraction;Judge described second Whether reach threshold value with fraction;Wherein, when the described second matching fraction is not up to the threshold value, the movable equipment is represented Lose target.
- 11. device as claimed in claim 10, which is characterized in that described device further includes:Module is obtained, for after the target is found again, when searching out suspected target again, based on the target mould Type matches the suspected target tracing area where the suspected target, obtains the 3rd matching fraction;Determining module, for when the described 3rd matching fraction reaches the threshold value, it to be the target to determine the suspected target.
- 12. device as claimed in claim 7, which is characterized in that the sports immunology information be motion vector, the superposition mould Block is used for corresponding first motion vector of first human body attitude and the movable equipment is currently performed described Second motion vector is superimposed, and obtains the 3rd motion vector;Wherein, the size and Orientation of first motion vector is according to described One human body attitude determines;The size of second motion vector determines that direction is determined according to the direction of motion according to speed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611051269.8A CN108089695B (en) | 2016-11-23 | 2016-11-23 | Method and device for controlling movable equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611051269.8A CN108089695B (en) | 2016-11-23 | 2016-11-23 | Method and device for controlling movable equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108089695A true CN108089695A (en) | 2018-05-29 |
CN108089695B CN108089695B (en) | 2021-05-18 |
Family
ID=62171145
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611051269.8A Active CN108089695B (en) | 2016-11-23 | 2016-11-23 | Method and device for controlling movable equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108089695B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108830219A (en) * | 2018-06-15 | 2018-11-16 | 北京小米移动软件有限公司 | Method for tracking target, device and storage medium based on human-computer interaction |
CN109032039A (en) * | 2018-09-05 | 2018-12-18 | 北京羽扇智信息科技有限公司 | A kind of method and device of voice control |
CN109460031A (en) * | 2018-11-28 | 2019-03-12 | 科大智能机器人技术有限公司 | A kind of system for tracking of the automatic tractor based on human bioequivalence |
CN109558835A (en) * | 2018-11-28 | 2019-04-02 | 科大智能机器人技术有限公司 | A kind of control method and its system of the automatic tractor based on human bioequivalence |
CN115690194A (en) * | 2022-10-17 | 2023-02-03 | 广州赤兔宸行科技有限公司 | Vehicle-mounted XR equipment positioning method, device, equipment and storage medium |
CN116524569A (en) * | 2023-05-10 | 2023-08-01 | 深圳大器时代科技有限公司 | Multi-concurrency face recognition system and method based on classification algorithm |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104808799A (en) * | 2015-05-20 | 2015-07-29 | 成都通甲优博科技有限责任公司 | Unmanned aerial vehicle capable of indentifying gesture and identifying method thereof |
CN104820998A (en) * | 2015-05-27 | 2015-08-05 | 成都通甲优博科技有限责任公司 | Human body detection and tracking method and device based on unmanned aerial vehicle mobile platform |
CN105159452A (en) * | 2015-08-28 | 2015-12-16 | 成都通甲优博科技有限责任公司 | Control method and system based on estimation of human face posture |
CN105607740A (en) * | 2015-12-29 | 2016-05-25 | 清华大学深圳研究生院 | Unmanned aerial vehicle control method and device based on computer vision |
CN105892493A (en) * | 2016-03-31 | 2016-08-24 | 纳恩博(北京)科技有限公司 | Information processing method and mobile device |
CN105929838A (en) * | 2016-05-20 | 2016-09-07 | 腾讯科技(深圳)有限公司 | Flight control method of aircraft, mobile terminal and flight control end |
US20160313743A1 (en) * | 2015-04-24 | 2016-10-27 | Samsung Display Co., Ltd. | Flying display device |
-
2016
- 2016-11-23 CN CN201611051269.8A patent/CN108089695B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160313743A1 (en) * | 2015-04-24 | 2016-10-27 | Samsung Display Co., Ltd. | Flying display device |
CN104808799A (en) * | 2015-05-20 | 2015-07-29 | 成都通甲优博科技有限责任公司 | Unmanned aerial vehicle capable of indentifying gesture and identifying method thereof |
CN104820998A (en) * | 2015-05-27 | 2015-08-05 | 成都通甲优博科技有限责任公司 | Human body detection and tracking method and device based on unmanned aerial vehicle mobile platform |
CN105159452A (en) * | 2015-08-28 | 2015-12-16 | 成都通甲优博科技有限责任公司 | Control method and system based on estimation of human face posture |
CN105607740A (en) * | 2015-12-29 | 2016-05-25 | 清华大学深圳研究生院 | Unmanned aerial vehicle control method and device based on computer vision |
CN105892493A (en) * | 2016-03-31 | 2016-08-24 | 纳恩博(北京)科技有限公司 | Information processing method and mobile device |
CN105929838A (en) * | 2016-05-20 | 2016-09-07 | 腾讯科技(深圳)有限公司 | Flight control method of aircraft, mobile terminal and flight control end |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108830219A (en) * | 2018-06-15 | 2018-11-16 | 北京小米移动软件有限公司 | Method for tracking target, device and storage medium based on human-computer interaction |
CN109032039A (en) * | 2018-09-05 | 2018-12-18 | 北京羽扇智信息科技有限公司 | A kind of method and device of voice control |
CN109460031A (en) * | 2018-11-28 | 2019-03-12 | 科大智能机器人技术有限公司 | A kind of system for tracking of the automatic tractor based on human bioequivalence |
CN109558835A (en) * | 2018-11-28 | 2019-04-02 | 科大智能机器人技术有限公司 | A kind of control method and its system of the automatic tractor based on human bioequivalence |
CN115690194A (en) * | 2022-10-17 | 2023-02-03 | 广州赤兔宸行科技有限公司 | Vehicle-mounted XR equipment positioning method, device, equipment and storage medium |
CN115690194B (en) * | 2022-10-17 | 2023-09-19 | 广州赤兔宸行科技有限公司 | Vehicle-mounted XR equipment positioning method, device, equipment and storage medium |
CN116524569A (en) * | 2023-05-10 | 2023-08-01 | 深圳大器时代科技有限公司 | Multi-concurrency face recognition system and method based on classification algorithm |
Also Published As
Publication number | Publication date |
---|---|
CN108089695B (en) | 2021-05-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108089695A (en) | A kind of method and apparatus for controlling movable equipment | |
CN109800689B (en) | Target tracking method based on space-time feature fusion learning | |
CN109102547A (en) | Robot based on object identification deep learning model grabs position and orientation estimation method | |
CN109597087A (en) | A kind of 3D object detection method based on point cloud data | |
CN108256421A (en) | A kind of dynamic gesture sequence real-time identification method, system and device | |
CN104615986B (en) | The method that pedestrian detection is carried out to the video image of scene changes using multi-detector | |
CN109191497A (en) | A kind of real-time online multi-object tracking method based on much information fusion | |
CN106780484A (en) | Robot interframe position and orientation estimation method based on convolutional neural networks Feature Descriptor | |
CN104281839A (en) | Body posture identification method and device | |
CN110427797B (en) | Three-dimensional vehicle detection method based on geometric condition limitation | |
CN104537689B (en) | Method for tracking target based on local contrast conspicuousness union feature | |
CN104392223A (en) | Method for recognizing human postures in two-dimensional video images | |
CN111178170B (en) | Gesture recognition method and electronic equipment | |
CN109087337B (en) | Long-time target tracking method and system based on hierarchical convolution characteristics | |
Ali et al. | Visual tree detection for autonomous navigation in forest environment | |
CN109740454A (en) | A kind of human body posture recognition methods based on YOLO-V3 | |
CN106327528A (en) | Moving object tracking method and operation method of unmanned aerial vehicle | |
Pham et al. | Pencilnet: Zero-shot sim-to-real transfer learning for robust gate perception in autonomous drone racing | |
Morita et al. | Panoramic view-based navigation in outdoor environments based on support vector learning | |
CN117011378A (en) | Mobile robot target positioning and tracking method and related equipment | |
CN110111358B (en) | Target tracking method based on multilayer time sequence filtering | |
CN116665097A (en) | Self-adaptive target tracking method combining context awareness | |
Togo et al. | Gesture recognition using hand region estimation in robot manipulation | |
CN109977884A (en) | Target follower method and device | |
CN110197123A (en) | A kind of human posture recognition method based on Mask R-CNN |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |