CN107025802A - A kind of method and unmanned plane that parking stall is found based on unmanned plane - Google Patents
A kind of method and unmanned plane that parking stall is found based on unmanned plane Download PDFInfo
- Publication number
- CN107025802A CN107025802A CN201710317680.3A CN201710317680A CN107025802A CN 107025802 A CN107025802 A CN 107025802A CN 201710317680 A CN201710317680 A CN 201710317680A CN 107025802 A CN107025802 A CN 107025802A
- Authority
- CN
- China
- Prior art keywords
- parking
- parking stall
- unmanned plane
- user
- automobile
- 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
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/14—Traffic control systems for road vehicles indicating individual free spaces in parking areas
- G08G1/141—Traffic control systems for road vehicles indicating individual free spaces in parking areas with means giving the indication of available parking spaces
- G08G1/142—Traffic control systems for road vehicles indicating individual free spaces in parking areas with means giving the indication of available parking spaces external to the vehicles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Traffic Control Systems (AREA)
Abstract
The present invention provides a kind of method that parking stall is found based on unmanned plane, comprises the following steps：Directly over unmanned plane during flying to target area center；The multiple cameras set on unmanned aerial vehicle body are opened and gather the image of the whole target area below unmanned plane；Unmanned machine testing described image, selectes out the parking stall of suitable user's automobile parking from described image；Unmanned plane during flying is moved to directly over the parking stall and hovered；Unmanned plane feeds back itself GPS location to user terminal.A kind of method that parking stall is found based on unmanned plane that the present invention is provided, it is only necessary to which user rests on parking can know in an open parking ground whether also there is free parking stall outside the venue, it is to avoid the situation for occurring making a journey for nothing by parking lot Nei Wukong parking stalls occurs.
Description
Technical field
The present invention relates to unmanned plane during flying field, more particularly to a kind of method that parking stall is found based on unmanned plane.
Background technology
User is in one open parking ground of arrival, and normal conditions are not aware that whether the open parking ground has empty parking stall
And position of the empty parking stall in the open parking ground, user can only drive to enter the searching of open parking ground line by line,
Seeing can just park in the case of being free parking stall, if it find that parking lot Nei Wukong parking stalls, can only drive to roll the dew away from
Its parking lot, finds the place that elsewhere can be parked.Above-mentioned user drives to enter the method that empty parking stall is found in parking lot in person,
Obviously it is not convenient enough.In the case of suddenly lacking on parking stall, often occur what is made a journey for nothing by parking lot Nei Wukong parking stalls
Situation, or in the case of having parking stall in parking lot, nearest optimal parking stall can not be searched out with existing method, these are all
Urgent problem to be solved.
The content of the invention
In view of the above-mentioned problems, the present invention provides a kind of method that parking stall is found based on unmanned plane.
A kind of method that parking stall is found based on unmanned plane that the present invention is provided, is comprised the following steps：S1：Unmanned plane during flying
Directly over to target area center；S2：The multiple cameras set on unmanned aerial vehicle body are opened and gathered below unmanned plane
Whole target area image；S3：Unmanned machine testing described image, selectes out suitable user's automobile parking from described image
Parking stall；S4：Unmanned plane during flying is moved to directly over the parking stall and hovered；S5：Unmanned plane feeds back itself GPS location extremely
User terminal.
Preferably, the step S3 further comprises：A1：Unmanned machine testing simultaneously filters out and can accommodate the sky of user's automobile
White parking area；A2：Unmanned plane selects one apart from user from the blank parking area that can accommodate user's automobile
Nearest blank parking area.
Preferably, the step S3 includes again：A3：It is one using the svm classifier model inspection recognized based on automobile
Whether the blank parking area nearest apart from user be close to parked car；If one blank nearest apart from user
Parking area is then directly chosen to be the parking stall of suitable user's automobile parking close to parked car；If it is one away from
The blank parking area nearest from user be not close to parked car, then in one blank parking nearest apart from user
The nearest parked car of outside detecting distance around region, and by the blank neighbouring with the closest parked car
Parking area is chosen to be the parking stall of suitable user's automobile parking.
Preferably, the generation step of the svm classifier model recognized based on automobile, including：Collect in advance for training
The first positive sample collection and the first negative sample collection, wherein the first positive sample collection is the multiple samples for including parked car
Image, the negative sample collection is not comprising the multiple sample images for having parked car；Specification according to setting cuts out described
Each sample image in one positive sample collection so that the complete vapour parked only is distributed with each sample image
Car；The characteristic vector of each sample image in the first positive sample collection after extracting the first negative sample collection respectively and cutting out；Using
The maps feature vectors are turned into the characteristic vector in higher dimensional space by sigmoid kernel functions；According in the higher dimensional space
Characteristic vector, obtains the svm classifier model recognized based on automobile with Optimal Parameters.
Preferably, the step A1 further comprises：An a moment corresponding set of frames is chosen, and extracts institute
The characteristic vector per two field picture in a set of frames is stated, wherein one set of frames is set on unmanned aerial vehicle body
The image construction that the multiple camera put is gathered respectively in synchronization.
Preferably, the step A1 further comprises：Using the svm classifier model recognized based on empty parking stall to described one
The characteristic vector per two field picture is detected in individual set of frames, in the one set of frames of detection whether
Include blank parking area；If not including blank parking area in one set of frames, described one is abandoned
Individual set of frames a, set of frames for choosing subsequent time according to the order of physical storage address re-starts detection；
If including blank parking area in one set of frames, mark what is included in one set of frames
All blank parking areas.
Preferably, the generation step of the svm classifier model recognized based on empty parking stall, including：Collecting in advance is used for
The the second positive sample collection and the second negative sample collection of training, wherein the second positive sample collection is to include the multiple samples for being free parking stall
This image, the second negative sample collection is not comprising the multiple sample images for being free parking stall；Specification according to setting cuts out institute
State each sample image in the second positive sample collection so that a complete empty parking is only distributed with each sample image
Position；The characteristic vector of each sample image in the second positive sample collection after extracting the second negative sample collection respectively and cutting out；Using
The maps feature vectors are turned into the characteristic vector in higher dimensional space by sigmoid kernel functions；According in the higher dimensional space
Characteristic vector, obtains the svm classifier model recognized based on empty parking stall with Optimal Parameters.
Preferably, the step A1 further comprises：The blank parking area that magnifying tags go out, by the institute after amplification
State blank parking area area and user needed for parking stall area do difference operation and obtain difference amount, analyze the difference amount with
The relation of the threshold value pre-set；If the difference amount is less than the threshold value, abandon that the difference amount is corresponding to be marked
Blank parking area；If the difference amount is more than the threshold value, the corresponding blank parking marked of the difference amount
Region is directly stored for that can accommodate the blank parking area of user's automobile.
Preferably, the step A2 includes：Filter out the maximum blank parking that can accommodate user's automobile of area
Region is simultaneously stored, and wherein the maximum blank parking area of area includes one or more；According to the suitable of physical storage address
Sequence chooses the maximum blank parking area of an area, is used as the blank parking area nearest apart from user.
Preferably, the step A3 includes again：A detection blank parking area corresponding frame nearest apart from user
Whether image is to be gathered by main camera；If not what is gathered by main camera, then the main camera and described one are kept
The corresponding auxiliary camera of two field picture is opened, and closes remaining auxiliary camera；If being gathered by the main camera
, then keep the main camera to open, close all auxiliary cameras.
Preferably, the step S4 includes：The unlatching situation of camera first on detection unmanned aerial vehicle body；When only master takes the photograph
When being opened as head and an auxiliary camera：The orientation movement of unmanned plane towards the auxiliary camera direction opened is fitted up to described
The parking stall for sharing family automobile parking is begun to appear in the range of the field of view of main camera；The auxiliary for closing the unlatching is taken the photograph
As head；According to positional distance image center location of the parking stall of suitable user's automobile parking in the image of main camera
Miss distance, unmanned plane during flying is moved to the surface on the parking stall of suitable user's automobile parking.
Preferably, the step S4 includes：The unlatching situation of camera first on detection unmanned aerial vehicle body；When only master takes the photograph
When being opened as head：According in positional distance image of the parking stall of suitable user's automobile parking in the image of main camera
The miss distance of heart position, unmanned plane during flying is moved to the surface on the parking stall of suitable user's automobile parking.
Preferably, the multiple camera includes being arranged on four auxiliary cameras of unmanned plane surrounding side and is arranged on
The main camera of unmanned plane bottom surface, the front lower place of the region overlay unmanned plane of four auxiliary cameras collection image, it is rear under
Side, lower left and lower right, the underface of the region overlay unmanned plane of the main camera collection image.
Preferably, the multiple camera uses varifocal wide-angle camera, and the focal length of the camera is taken the photograph by controlling
As the supersonic motor of head is controlled.
The present invention provides a kind of unmanned plane of any one of application above method.
A kind of method that parking stall is found based on unmanned plane that the present invention is provided, it is only necessary to which user rests on parking outside the venue just
Understand whether also there is free parking stall into an open parking ground, it is to avoid appearance is made a journey for nothing because of parking lot Nei Wukong parking stalls
Situation occur.
Brief description of the drawings
Fig. 1 represents a kind of schematic flow sheet that parking stall is found based on unmanned plane described in the embodiment of the present invention；
Fig. 2, which represents that being found based on unmanned plane described in the embodiment of the present invention is detected and filtered out in parking stall, can accommodate user's vapour
The schematic flow sheet on the empty parking stall of car；
Fig. 3 represents that being found based on unmanned plane in parking stall described in the embodiment of the present invention selects closest empty parking stall
Schematic flow sheet.
Embodiment
Technical scheme is described in further detail below in conjunction with the accompanying drawings, the reality described by reference to accompanying drawing
It is exemplary to apply example, it is intended to for explaining the present invention, and be not considered as limiting the invention.Based on the reality in the present invention
Example is applied, the every other embodiment that those skilled in the art are obtained under the premise of creative work is not made belongs to this
Invent the scope of protection.
One embodiment of the present of invention be unmanned plane on open parking ground automatic searching blank parking stall, it is handled below
Flow is described in detail.
User searches the empty wagons in some open parking ground when reaching some large-scale open parking ground using unmanned plane
The handling process of position, user's operation and unmanned plane is as shown in fig. 1.
User obtains the region area in parking lot center position coordinates and parking lot by map, then by terminal
APP the region area of parking lot center position coordinates and parking lot is wirelessly transmitted to unmanned plane, root after unmanned plane takes off
Coordinate and area generation flight path, fly and are moved to the surface of parking lot center position and in vertical direction distance accordingly
Center position certain altitude（Step 10）.The coverage energy of 5 cameras on above-mentioned certain altitude, unmanned aerial vehicle body
Enough cover whole car park areas.
Wherein, the area s in the parking lot determines the height of unmanned plane lift-off, and the area s in parking lot is bigger, unmanned plane liter
Empty height h is relative to be raised, and the height h of unmanned plane lift-off sets exemplary as follows：At oneStop
Parking lot, height h=20m of unmanned plane lift-off；At oneParking lot, unmanned plane
Height h=30m of lift-off, it is exemplary only herein to enumerate, also can using it is other it is suitable by the way of carry out height setting.
Unmanned plane is reached behind the position of setting, is fed back to user terminal up to set point information, and then user passes through
APP in terminal is opened after function of searching, unmanned plane wireless receiving to searching instruction, unmanned plane internal processor control unmanned plane
Main camera and four auxiliary cameras on fuselage start simultaneously and concurrently to whole car park areas carry out IMAQ（Step
Rapid 20）.The parking field picture of main camera and four auxiliary camera collections on a certain moment, unmanned aerial vehicle body is through conventional place
Formed after reason in a set of frames corresponding with the moment, each set of frames and include 5 two field pictures, i.e., at certain for the moment
Carve, 5 camera collection 5 two field pictures of generation, form a set of frames at the moment on unmanned aerial vehicle body.
Wherein, on the unmanned aerial vehicle body, remove outside the main camera being located at immediately below unmanned plane, the front side end of unmanned plane
Face, back end surface, all it is separately installed with an auxiliary camera at the center position of left end face and its right end face, the master takes the photograph
The image-region that can be gathered as head and four auxiliary cameras covers underface, front lower place, the back lower place, the lower-left of unmanned plane
Five directions such as side, lower right.When unmanned plane is located at the suitable height in parking lot overhead, the main camera on unmanned aerial vehicle body
Whole target area can be covered with the image-region that four auxiliary cameras can be gathered.Four installed on unmanned aerial vehicle body
Auxiliary camera is all varifocal wide-angle camera, and the focal length of camera can be controlled by the supersonic motor of control camera
System, while the control of supersonic motor is realized by the processor of unmanned plane by drive and control of electric machine chip.The main camera
The light that object reflects is converted into by original optical information and storage, Ran Houtong by photo-sensitive cell with four auxiliary cameras
Cross internal transmission interface such as mobile Industry Processor Interface（MIPI, Mobile Industry Processor Interface）
Or serial ports, using general Data Transport Protocol is by the image transmitting of collection to unmanned plane and is stored in memory, so as to
It is further processed.
Unmanned plane, which is detected and filtered out to the image of above-mentioned collection, can accommodate the empty parking stall of user's automobile（Step
30）.
In above-mentioned steps, specifically, as shown in Fig. 2 unmanned plane internal processor to choose a certain moment corresponding one
Set of frames, and the extraction of characteristic vector is carried out to selecting 5 two field pictures included in the one set of frames come
（Step 301）, it is directed to that to the extracting method that every two field picture carries out characteristic vector existing conventional method can be used.
Specifically, choosing the corresponding set of frames of a certain moment and to being carried out in the set of frames per two field picture
The extraction of characteristic vector, herein only with histograms of oriented gradients（Histogram of Oriented Gradient, HOG）Algorithm
It is illustratively illustrative.The step of HOG feature extracting methods, is as follows：
Will be per two field picture gray processing；
The normalization of color space is carried out to every two field picture using Gamma orthosises, the Mathematical representation of Gamma corrections is：, wherein I（X, y）For the gray value of each pixel, gamma value is selected according to the situation of adaptation, example
It such as can use 0.5；It is that evolution is carried out to the value of each pixel as gamma=0.5；
The gradient of each pixel in every two field picture, including size and Orientation are calculated, the Mathematical representation of its gradient calculation is：,, whereinRepresent per frame figure
The pixel as in（X, y）The horizontal direction gradient at place,Pixel in representing per two field picture（X, y）The vertical direction ladder at place
Degree, H（X, y）Pixel in representing per two field picture（X, y）The gray value at place；Pixel（X, y）The gradient magnitude at place and gradient side
To respectively：,；
Several junior unit cell will be divided into per two field picture, the regional window of such as 6*6 pixel composition counts each unit
Cell histogram of gradients, forms each unit cell descriptor；
Choose several said units cell and constitute a block block, by the feature of all cell in a block block
Descriptor, which is together in series, just obtains block block HOG features, can constitute several block block per two field picture, obtain
To several block block HOG features；
The HOG features of all block block in every two field picture are together in series and obtain the characteristic vector of every two field picture.Then utilize
Same method extracts characteristic vector of the said one set of frames per two field picture successively.
Then unmanned plane internal processor utilizes the SVMs recognized based on empty parking stall（SVM, support
vector machine）The characteristic vector that disaggregated model is extracted to said one set of frames per two field picture is examined
Survey, detect in above-mentioned set of frames whether there is free parking stall（Step 302）：If do not examined in said one set of frames
Sky parking stall has been measured, has illustrated that detection mistake occurs or at the time of above-mentioned set of frames correspondence, the parking lot has been fully parked with
Car, then abandon said one set of frames, return to step 20, restarts to carry out IMAQ（Step 303）；If
Sky parking stall has been detected in said one set of frames, then has marked the sky in above-mentioned set of frames in every two field picture and stops
Parking stall（Step 304）；
Wherein, the svm classifier model recognized based on empty parking stall is in advance by the sample image to a large amount of empty parking stalls
Carry out feature extraction and training is obtained and is stored in the memory in unmanned plane, the SVM recognized based on empty parking stall
The mistake that the generation of disaggregated model and the characteristic vector extracted to said one set of frames per two field picture are detected
Journey is specific as follows：
First, substantial amounts of parking lot Nei Kong parking stalls sample image and substantial amounts of non-NULL parking stall sample image are collected in advance, its
In comprising be free parking stall sample image be the second positive sample collection, not comprising the sample image for being free parking stall be the second negative sample
This collection；Each sample image in above-mentioned second positive sample collection is cut out successively according to the specification of setting, it is ensured that after cutting out
Each sample image in a complete empty parking stall is only distributed with；Above-mentioned second negative sample collection does not cut out processing；
Secondly, the characteristic vector of the second positive sample collection after above-mentioned cut out and the second negative sample collection is extracted；
Again, using sigmoid kernel functions, the characteristic vector data of above-mentioned two classes sample is mapped to higher dimensional space, it is to avoid go out
It can not find the function parameter of the condition of satisfaction in present luv space two class sample datas are divided into situation about being distinguished；
Then, using the characteristic vector of above-mentioned two classes sample as the SVM classifier with initiation parameter input, it is trained after
The SVM classifier with Optimal Parameters is obtained, the SVM classifier with Optimal Parameters is exactly required based on the knowledge of empty parking stall
Other svm classifier model；
Finally, the svm classifier model recognized using above-mentioned generation based on empty parking stall is in higher dimensional space to said one figure
The characteristic vector extracted as frame set per two field picture carries out key words sorting, so that it is every to mark said one set of frames
The hollow parking bit position of two field picture and non-NULL parking bit position.
Process to above-mentioned key words sorting in the way of mathematical modeling carries out example description, i.e.,：Said one picture frame collection
Closing the characteristic vector extracted per two field picture includes belonging to the characteristic vector of sky parking stall classification and belongs to non-NULL parking stall
The characteristic vector of classification, if representing characteristic vector data with X, y=1 or -1 classification for representing characteristic vector data for example can be with
It is that y=1 represents to belong to the characteristic vector of sky parking stall classification, y=- 1 represents to belong to the characteristic vector of non-NULL parking stall classification, then belonged to
It may be expressed as in the characteristic vector of empty parking stall classification（X, 1）, the characteristic vector for belonging to non-NULL parking stall classification can
It is expressed as（X, -1）, this two category features DUAL PROBLEMS OF VECTOR MAPPING to higher dimensional space is to represent two class higher-dimension discrete coordinates, and above-mentioned is based on
The purpose that the svm classifier model of empty parking stall identification needs to carry out sample training in advance is to find one in higher dimensional space to surpass
Plane z=f（x, y）, above-mentioned two classes coordinate points can be made a distinction, specially f（x, y）>The institute of 0 above-mentioned y=1 of point correspondence
There are coordinate points, f（x, y）<All coordinate points of 0 above-mentioned y=- 1 of point correspondence, i.e. f（X, y）>0 point belongs to sky parking stall class
Another characteristic vector, f（X, y）<0 point belongs to the characteristic vector of non-NULL parking stall classification.In addition, an also class f（x, y）=
0 point belongs to the unconspicuous point of feature, it is impossible to be clearly attributed to the characteristic vector or non-NULL parking stall class of sky parking stall classification
Another characteristic vector, for f（x, y）=0 point, then directly abandon, advantageously reduce error rate.
Therefore it is above-mentioned it is trained obtain have Optimal Parameters based on empty parking stall recognize SVM classifier as we
The required such a hyperplane function that two class coordinate points can be made a distinction, using such a hyperplane function with regard to energy
Enough characteristic vectors extracted to a set of frames per two field picture carry out key words sorting.
Return in Fig. 2, after marking in above-mentioned set of frames per the empty parking stall in two field picture, then actually to stop
The equivalent scalar multiple in parking stall amplifies the empty parking stall of above-mentioned mark, calculates the empty parking stall area surface of the mark after above-mentioned amplification
Product, and do difference operation with the actual parking stall region area needed for user's automobile（Step 305）；Whether judge above-mentioned difference amount
More than the threshold value pre-set in system（Step 306）：If above-mentioned difference amount is less than threshold value, illustrate the mark that participation is calculated
Actual parking stall region in parking lot corresponding to empty parking stall is not enough to for parking user's automobile, then directly abandons this poor
The empty parking stall of the corresponding mark of value amount（Step 307）；If above-mentioned difference amount is more than threshold value, illustrate the mark that participation is calculated
Actual parking stall region in parking lot corresponding to empty parking stall can park user's automobile, then storing above-mentioned can accommodate use
The empty parking space number evidence of family automobile（Step 308）.
Return in Fig. 1, used filtering out to accommodate behind the empty parking stall of user's automobile, it is necessary to further select distance
The nearest empty parking stall in family（Step 40）.
Specifically, in above-mentioned steps, stopping as shown in figure 3, calculating the above-mentioned sky that can accommodate user's automobile filtered out
The area of parking stall, and the size of relatively more each area, filter out the maximum empty parking stall of area and store（Step 401）, wherein sieving
The empty parking stall for the area maximum selected is nearest apart from user, and the empty parking stall of the area maximum filtered out may be comprising number
It is individual, then therefrom choose the maximum empty parking stall of an area filtered out according to the order of storage address（Step 402）；Find
A two field picture corresponding to the above-mentioned empty parking stall selected（Step 403）；The area selected in detection above-mentioned steps is maximum
The corresponding two field picture of parking facility whether gathered by main camera（Step 404）：Stop if above-mentioned area is maximum
A two field picture corresponding to parking lot is gathered by main camera, then keeps the main camera to open and close four second cameras
Head（Step 405；If the maximum corresponding two field picture of parking facility of above-mentioned area is not gathered by main camera, it is
By any one collection in four auxiliary cameras, then main camera auxiliary camera corresponding with the two field picture is kept to open
Open, close remaining auxiliary camera（Step 406）.
Continue back in Fig. 1, it is necessary to determine whether what is selected after the empty parking stall nearest apart from user is selected
Whether empty parking stall is close to the other automobiles parked（Step 50）：If the above-mentioned empty parking stall selected is not close to parking
Other automobiles, illustrating may be also close to other empty parking stall, in order to avoid interval around the above-mentioned empty parking stall selected
Park, it is necessary to detect the other automobiles parked apart from the above-mentioned empty parking stall selected recently, it is then other by what is parked
Automobile close to empty parking stall be defined as the parking stall that user can park（Step 60）；If examined around the tentative empty parking stall
Measure and parked automobile, illustrate that the above-mentioned empty parking selected is that the automobile of user can be straight close to parked car around
Connect and be parked in the above-mentioned empty parking stall selected, then will be defined as close to the above-mentioned empty parking stall selected of parked car
The parking stall that user's automobile can be parked（Step 70）, unmanned plane during flying, which is moved to, is defined as the parking stall that user's automobile can be parked
Simultaneously hover surface（Step 80）.
In above-mentioned steps 50, specifically, whether other close to what is parked on the empty parking stall for determining whether to select
, it is necessary to use the svm classifier model recognized based on automobile to being selected in corresponding two field picture in empty parking stall for selecting during automobile
Image around the empty parking stall of taking-up is detected with the presence or absence of automobile, wherein the SVM recognized based on automobile used herein
Disaggregated model is to carry out feature extraction by the sample image to parked car in a large amount of parking lots in advance and train to obtain
And be stored in the memory in unmanned plane, the generation of the svm classifier model recognized based on automobile and to selecting
Whether the image around empty parking stall selected in the corresponding two field picture in empty parking stall parks the tool that automobile is detected
Body process is as follows：
First, the sample image of parked car and substantial amounts of empty parking stall sample image in substantial amounts of parking lot are collected in advance,
The sample image for wherein including parked car is the first positive sample collection, is the not comprising the sample image for having parked car
One negative sample collection；Each sample image in above-mentioned first positive sample collection is cut out successively according to the specification of setting so that described
A complete parked car is only distributed with each sample image；Above-mentioned first negative sample collection does not cut out processing；
Secondly, the characteristic vector of the first positive sample collection after above-mentioned cut out and the first negative sample collection is extracted；
Again, using sigmoid kernel functions, the characteristic vector data of above-mentioned two classes sample is mapped to higher dimensional space, it is to avoid
The parametric function for occurring can not find the condition of satisfaction in luv space divides two class sample datas situation about being distinguished；
Then, using the characteristic vector of above-mentioned two classes sample as the SVM classifier with initiation parameter input, it is trained after
The Optimal Parameters of SVM classifier are obtained, the SVM classifier with Optimal Parameters is exactly the required SVM recognized based on automobile
Disaggregated model；
Finally, being stopped based on the svm classifier model that automobile is recognized in higher dimensional space to the sky selected using above-mentioned generation
Image around position carries out whether having parked automobile around key words sorting, the empty parking stall that detection identification is selected.
Process to above-mentioned key words sorting in the way of mathematical modeling carries out example description, i.e.,：The above-mentioned sky selected stops
The characteristic vector of image around parking stall includes belonging to the characteristic vector of class of vehicles and belongs to sky parking stall class another characteristic
Vector, if representing characteristic vector data with X, the classification of y=2 or -2 expression characteristic vector datas for example can be that y=2 are represented
Belong to the characteristic vector of class of vehicles, y=- 2 represent to belong to the characteristic vector of sky parking stall classification, then belong to the spy of class of vehicles
Vector is levied to may be expressed as（X, 2）, the characteristic vector for belonging to sky parking stall classification may be expressed as（X, -2）, this two class spy
Levy DUAL PROBLEMS OF VECTOR MAPPING and represent two class higher-dimension discrete coordinates to higher dimensional space, and the above-mentioned svm classifier model recognized based on automobile
It is that a hyperplane z=f is found in higher dimensional space to need the purpose for carrying out sample training in advance（x, y）, can will be above-mentioned
Two class coordinate points make a distinction, specially f（x, y）>All coordinate points of 0 above-mentioned y=2 of point correspondence, f（x, y）<0 point pair
Answer all coordinate points of above-mentioned y=- 2, i.e. f（x, y）>0 point belongs to the characteristic vector of class of vehicles, f（x, y）<0 point
Belong to the characteristic vector of sky parking stall classification.In addition, an also class f（x, y）=0 point belongs to the unconspicuous point of feature, no
The characteristic vector of sky parking stall classification or the characteristic vector of class of vehicles can be clearly attributed to, for f（x, y）=0 point, then
Directly abandon, advantageously reduce error rate.
Therefore it is above-mentioned it is trained obtain have Optimal Parameters the SVM classifier recognized based on automobile be us needed for
Such a hyperplane function that two class coordinate points can be made a distinction, just can be right using such a hyperplane function
The characteristic vector of image around the above-mentioned empty parking stall selected carries out key words sorting.
The step 60 in Fig. 1 is back to, is detecting that the empty parking stall that distance is selected parked vapour around nearby
Che Hou, then performs step 70, will can be parked as user's automobile close to the empty parking stall of the parked car detected
Parking stall.
In step 80, unmanned plane during flying is moved to the process for the surface for being defined as the parking stall that user's automobile can be parked
As described in following step：
The unlatching situation of each camera first on monitoring unmanned aerial vehicle body, for the main camera of above-mentioned unlatching and a second camera
The situation of head, the auxiliary that then processor of unmanned plane drives the flight control units of unmanned plane make it that unmanned plane direction is opened is taken the photograph
As the side surface direction movement of head institute direction is until the above-mentioned parking stall that is defined as user's automobile and can park at a time starts
In the image of main camera collection now, when the above-mentioned parking stall that is defined as user's automobile and can park occurs in main camera
Afterwards, the auxiliary camera of unlatching can be closed, and then the processor of unmanned plane can be parked according to the above-mentioned user's automobile that is defined as
Positional distance image center location of the parking stall in the image of main camera miss distance, the flight control of driving unmanned plane is single
Member control unmanned plane towards be defined as the parking stall flight that user's automobile can park move so that unmanned plane be moved to it is above-mentioned really
It is set to the surface for the parking stall that user's automobile can be parked, the parking stall that now user's automobile can be parked should be located at unmanned plane
Center position in the image for the main camera collection opened, i.e., the parking stall that now user's automobile can be parked is in main camera
Image in the miss distance of positional distance image center location be zero；When opening main camera for above-mentioned, nobody
The processor of machine is according to the above-mentioned position for being defined as parking stall that user's automobile can park in the image that the main camera is gathered
The miss distance of range image center, the flight control units control unmanned plane of driving unmanned plane is defined as user towards above-mentioned
The parking stall flight that automobile can be parked moves so that unmanned plane is moved to the surface of the above-mentioned destination object selected.
During side surface direction movement of the unmanned plane towards the auxiliary camera institute direction opened, in the auxiliary of unlatching
In the image of camera collection, it should be gradually toward the side of image to detect the above-mentioned parking stall that is defined as user's automobile and can park
Edge is moved, while in the image that gathers of main camera of unlatching, the above-mentioned parking stall that is defined as user's automobile and can park be from
Without to the image border for appearing in collection, then gradually toward the process of the picture centre movement gathered.If do not occurred above-mentioned
Phenomenon, then explanation this time detection failure, then abandon this moment corresponding set of frames, chooses next according to storage address order
Moment corresponding set of frames simultaneously restarts detection blank parking stall.
After unmanned plane hovering, the GPS position information of unmanned plane can be wirelessly transmitted to user terminal（Step 90）, Yong Hugen
According to the GPS location of unmanned plane, with regard to the parking stall that will be parked can be found.
Those skilled in the art will readily occur to its of the present invention after considering specification and putting into practice invention disclosed herein
Its embodiment.The application be intended to the present invention any modification, purposes or adaptations, these modifications, purposes or
Person's adaptations follow the general principle of the present invention and including undocumented common knowledge or usual in the art
Technological means.Description and embodiments be considered only as it is exemplary, true scope and spirit of the invention by following right will
Ask and point out.
It should be appreciated that the invention is not limited in the precision architecture for being described above and being shown in the drawings, and
And various modifications and changes can be being carried out without departing from the scope.The scope of the present invention is only limited by appended claim.
Claims (15)
1. a kind of method that parking stall is found based on unmanned plane, it is characterised in that comprise the following steps：
S1：Directly over unmanned plane during flying to target area center；
S2：The multiple cameras set on unmanned aerial vehicle body are opened and gather the image of the whole target area below unmanned plane；
S3：Unmanned machine testing described image, selectes out the parking stall of suitable user's automobile parking from described image；
S4：Unmanned plane during flying is moved to directly over the parking stall and hovered；
S5：Unmanned plane feeds back itself GPS location to user terminal.
2. according to the method described in claim 1, it is characterised in that the step S3 further comprises：
A1：Unmanned machine testing simultaneously filters out and can accommodate the blank parking area of user's automobile；
A2：Unmanned plane selects a sky nearest apart from user from the blank parking area that can accommodate user's automobile
White parking area.
3. method according to claim 2, it is characterised in that the step S3 includes again：
A3：It is using the one blank parking area nearest apart from user of the svm classifier model inspection recognized based on automobile
It is no close to parked car；
If one blank parking area nearest apart from user is directly chosen to be suitable use close to parked car
The parking stall of family automobile parking；
If one blank parking area nearest apart from user be not close to parked car, in one distance
The outside nearest parked car of detecting distance around the nearest blank parking area of user, and will stop with described closest
The neighbouring blank parking area of the automobile put is chosen to be the parking stall of suitable user's automobile parking.
4. method according to claim 3, it is characterised in that
The generation step of the svm classifier model recognized based on automobile, including：
The the first positive sample collection and the first negative sample collection for training are collected in advance, wherein the first positive sample collection is to include
Multiple sample images of parked car, the negative sample collection is not comprising the multiple sample images for having parked car；
Specification according to setting cuts out each sample image in the first positive sample collection so that in each sample image
A complete parked car is only distributed with；
The characteristic vector of each sample image in the first positive sample collection after extracting the first negative sample collection respectively and cutting out；
Using sigmoid kernel functions, the maps feature vectors are turned into the characteristic vector in higher dimensional space；
According to the characteristic vector in the higher dimensional space, the svm classifier model recognized based on automobile with Optimal Parameters is obtained.
5. method according to claim 2, it is characterised in that the step A1 further comprises：
An a moment corresponding set of frames is chosen, and extracts the feature in one set of frames per two field picture
Vector, distinguishes wherein one set of frames is the multiple camera set on unmanned aerial vehicle body in synchronization
The image construction of collection.
6. method according to claim 5, it is characterised in that the step A1 further comprises：
Using the svm classifier model recognized based on empty parking stall to described in every two field picture in one set of frames
Characteristic vector detected, detects in one set of frames whether include blank parking area；
If not including blank parking area in one set of frames, one set of frames is abandoned, is pressed
A set of frames according to the order selection subsequent time of physical storage address re-starts detection；
If including blank parking area in one set of frames, bag in one set of frames is marked
All blank parking areas contained.
7. method according to claim 6, it is characterised in that the svm classifier model recognized based on empty parking stall
Generation step, including：
The the second positive sample collection and the second negative sample collection for training are collected in advance, wherein the second positive sample collection is to include
Multiple sample images on empty parking stall, the second negative sample collection is not comprising the multiple sample images for being free parking stall；
Specification according to setting cuts out each sample image in the second positive sample collection so that in each sample image
A complete empty parking stall is only distributed with；
The characteristic vector of each sample image in the second positive sample collection after extracting the second negative sample collection respectively and cutting out；
Using sigmoid kernel functions, the maps feature vectors are turned into the characteristic vector in higher dimensional space；
According to the characteristic vector in the higher dimensional space, the svm classifier recognized based on empty parking stall with Optimal Parameters is obtained
Model.
8. method according to claim 6, it is characterised in that the step A1 further comprises：
The blank parking area that magnifying tags go out, will stop needed for the area of the blank parking area after amplification and user
Parking stall area does difference operation and obtains difference amount, analyzes the relation of the difference amount and the threshold value pre-set；
If the difference amount is less than the threshold value, the corresponding blank parking area marked of the difference amount is abandoned；
If the difference amount is more than the threshold value, the corresponding blank parking area marked of the difference amount is to hold
Receive the blank parking area of user's automobile, directly stored.
9. method according to claim 2, it is characterised in that the step A2 includes：
Filter out the maximum blank parking area that can accommodate user's automobile of area and store, wherein the maximum institute of area
Blank parking area is stated including one or more；
Order according to physical storage address chooses the maximum blank parking area of an area, as nearest apart from user
Blank parking area.
10. the method according to claim 2 or 9, it is characterised in that the step A2 further comprises：Detection it is described away from
Whether a blank parking area corresponding two field picture nearest from user is to be gathered by main camera；
If not what is gathered by main camera, then the main camera and the corresponding auxiliary of a two field picture is kept to take the photograph
As head is opened, remaining auxiliary camera is closed；
If by the main camera collection, then keeping the main camera to open, closing all auxiliary cameras.
11. according to the method described in claim 1, it is characterised in that the step S4 includes：First on detection unmanned aerial vehicle body
The unlatching situation of camera；
When only main camera and an auxiliary camera are opened：
Unmanned plane moves the parking stall until suitable user's automobile parking towards the orientation for the auxiliary camera direction opened
In the range of the field of view for beginning to appear in main camera；
Close the auxiliary camera of the unlatching；
According to positional distance image center location of the parking stall of suitable user's automobile parking in the image of main camera
Miss distance, unmanned plane during flying is moved to the surface on the parking stall of suitable user's automobile parking.
12. according to the method described in claim 1, it is characterised in that the step S4 includes：First on detection unmanned aerial vehicle body
The unlatching situation of camera；
When only main camera is opened：
According to positional distance image center location of the parking stall of suitable user's automobile parking in the image of main camera
Miss distance, unmanned plane during flying is moved to the surface on the parking stall of suitable user's automobile parking.
13. according to the method described in claim 1, it is characterised in that：The multiple camera includes being arranged on unmanned plane surrounding
Four auxiliary cameras and the main camera installed in unmanned plane bottom surface of side, four auxiliary cameras collection image
Front lower place, the back lower place, lower left and the lower right of region overlay unmanned plane, the region overlay of the main camera collection image without
Man-machine underface.
14. according to the method described in claim 1, it is characterised in that：The multiple camera uses varifocal wide-angle imaging
Head, the focal length of the camera is controlled by the supersonic motor of control camera.
15. a kind of unmanned plane of any one of application claim 1-14 methods described.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710317680.3A CN107025802B (en) | 2017-05-08 | 2017-05-08 | Unmanned aerial vehicle-based parking space searching method and unmanned aerial vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710317680.3A CN107025802B (en) | 2017-05-08 | 2017-05-08 | Unmanned aerial vehicle-based parking space searching method and unmanned aerial vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107025802A true CN107025802A (en) | 2017-08-08 |
| CN107025802B CN107025802B (en) | 2020-10-13 |
Family
ID=59529492
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710317680.3A Active CN107025802B (en) | 2017-05-08 | 2017-05-08 | Unmanned aerial vehicle-based parking space searching method and unmanned aerial vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107025802B (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107610523A (en) * | 2017-10-18 | 2018-01-19 | 冯迎安 | A kind of parking stall automatic monitoring alarm method |
| CN107644545A (en) * | 2017-10-18 | 2018-01-30 | 冯迎安 | A kind of unmanned plane parking position monitoring system and its monitoring method |
| CN107730985A (en) * | 2017-10-18 | 2018-02-23 | 冯迎安 | A kind of unmanned plane seeks stall system and its control method |
| CN107862893A (en) * | 2017-10-31 | 2018-03-30 | 西安科锐盛创新科技有限公司 | Intelligent stereo parking space |
| CN107886761A (en) * | 2017-11-14 | 2018-04-06 | 金陵科技学院 | A kind of parking lot monitoring method based on unmanned plane |
| CN109598978A (en) * | 2019-01-07 | 2019-04-09 | 哈尔滨理工大学 | Parking guide method based on unmanned plane and the unmanned plane for stopping guide |
| CN109859524A (en) * | 2019-03-29 | 2019-06-07 | 中国铁塔股份有限公司上海市分公司 | System, method and apparatus for parking position monitoring |
| CN110135360A (en) * | 2019-05-17 | 2019-08-16 | 北京深醒科技有限公司 | A kind of parking stall recognition methods based on local binary patterns and support vector machines |
| CN110606071A (en) * | 2019-09-06 | 2019-12-24 | 中国第一汽车股份有限公司 | Parking method, parking device, vehicle and storage medium |
| CN110758246A (en) * | 2018-07-25 | 2020-02-07 | 广州小鹏汽车科技有限公司 | Automatic parking method and device |
| CN110956846A (en) * | 2019-12-11 | 2020-04-03 | 济宁市众帮来袭信息科技有限公司 | Parking service method, device and system and storage medium |
| CN114927006A (en) * | 2022-05-23 | 2022-08-19 | 东风汽车集团股份有限公司 | Indoor passenger-replacing parking system based on unmanned aerial vehicle |
| CN115140022A (en) * | 2022-06-24 | 2022-10-04 | 重庆金康赛力斯新能源汽车设计院有限公司 | Automatic parking debugging method and device, computer equipment and storage medium |
| CN115762234A (en) * | 2023-01-10 | 2023-03-07 | 广东广宇科技发展有限公司 | Intelligent community parking management method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663357A (en) * | 2012-03-28 | 2012-09-12 | 北京工业大学 | Color characteristic-based detection algorithm for stall at parking lot |
| CN104112370A (en) * | 2014-07-30 | 2014-10-22 | 哈尔滨工业大学深圳研究生院 | Monitoring image based intelligent parking lot parking place identification method and system |
| CN105427661A (en) * | 2015-12-14 | 2016-03-23 | 山东科技大学 | Large-sized parking space detection system based on solar unmanned aerial vehicle group and working method thereof |
| CN105654773A (en) * | 2016-01-07 | 2016-06-08 | 张文忠 | Intelligent guide system of vehicle-mounted flight-accompanying unmanned plane |
| WO2016177545A1 (en) * | 2015-05-04 | 2016-11-10 | Siemens Aktiengesellschaft | System for automatic mobile parking space monitoring |
| CN106251697A (en) * | 2016-10-18 | 2016-12-21 | 珠海格力电器股份有限公司 | Method, device and system for searching parking space |
-
2017
- 2017-05-08 CN CN201710317680.3A patent/CN107025802B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663357A (en) * | 2012-03-28 | 2012-09-12 | 北京工业大学 | Color characteristic-based detection algorithm for stall at parking lot |
| CN104112370A (en) * | 2014-07-30 | 2014-10-22 | 哈尔滨工业大学深圳研究生院 | Monitoring image based intelligent parking lot parking place identification method and system |
| WO2016177545A1 (en) * | 2015-05-04 | 2016-11-10 | Siemens Aktiengesellschaft | System for automatic mobile parking space monitoring |
| CN105427661A (en) * | 2015-12-14 | 2016-03-23 | 山东科技大学 | Large-sized parking space detection system based on solar unmanned aerial vehicle group and working method thereof |
| CN105654773A (en) * | 2016-01-07 | 2016-06-08 | 张文忠 | Intelligent guide system of vehicle-mounted flight-accompanying unmanned plane |
| CN106251697A (en) * | 2016-10-18 | 2016-12-21 | 珠海格力电器股份有限公司 | Method, device and system for searching parking space |
Non-Patent Citations (1)
| Title |
|---|
| 李星等: "基于HOG特征和SVM的前向车辆识别方法", 《计算机科学》 * |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107644545A (en) * | 2017-10-18 | 2018-01-30 | 冯迎安 | A kind of unmanned plane parking position monitoring system and its monitoring method |
| CN107730985A (en) * | 2017-10-18 | 2018-02-23 | 冯迎安 | A kind of unmanned plane seeks stall system and its control method |
| CN107610523A (en) * | 2017-10-18 | 2018-01-19 | 冯迎安 | A kind of parking stall automatic monitoring alarm method |
| CN107862893A (en) * | 2017-10-31 | 2018-03-30 | 西安科锐盛创新科技有限公司 | Intelligent stereo parking space |
| CN107886761A (en) * | 2017-11-14 | 2018-04-06 | 金陵科技学院 | A kind of parking lot monitoring method based on unmanned plane |
| CN110758246B (en) * | 2018-07-25 | 2021-06-04 | 广州小鹏汽车科技有限公司 | Automatic parking method and device |
| CN110758246A (en) * | 2018-07-25 | 2020-02-07 | 广州小鹏汽车科技有限公司 | Automatic parking method and device |
| CN109598978A (en) * | 2019-01-07 | 2019-04-09 | 哈尔滨理工大学 | Parking guide method based on unmanned plane and the unmanned plane for stopping guide |
| CN109859524A (en) * | 2019-03-29 | 2019-06-07 | 中国铁塔股份有限公司上海市分公司 | System, method and apparatus for parking position monitoring |
| CN110135360A (en) * | 2019-05-17 | 2019-08-16 | 北京深醒科技有限公司 | A kind of parking stall recognition methods based on local binary patterns and support vector machines |
| CN110606071A (en) * | 2019-09-06 | 2019-12-24 | 中国第一汽车股份有限公司 | Parking method, parking device, vehicle and storage medium |
| CN110956846A (en) * | 2019-12-11 | 2020-04-03 | 济宁市众帮来袭信息科技有限公司 | Parking service method, device and system and storage medium |
| CN114927006A (en) * | 2022-05-23 | 2022-08-19 | 东风汽车集团股份有限公司 | Indoor passenger-replacing parking system based on unmanned aerial vehicle |
| CN114927006B (en) * | 2022-05-23 | 2023-03-14 | 东风汽车集团股份有限公司 | Indoor passenger-replacing parking system based on unmanned aerial vehicle |
| CN115140022A (en) * | 2022-06-24 | 2022-10-04 | 重庆金康赛力斯新能源汽车设计院有限公司 | Automatic parking debugging method and device, computer equipment and storage medium |
| CN115140022B (en) * | 2022-06-24 | 2024-05-24 | 重庆赛力斯新能源汽车设计院有限公司 | Automatic parking debugging method and device, computer equipment and storage medium |
| CN115762234A (en) * | 2023-01-10 | 2023-03-07 | 广东广宇科技发展有限公司 | Intelligent community parking management method |
| CN115762234B (en) * | 2023-01-10 | 2023-06-20 | 广东广宇科技发展有限公司 | Smart community parking management method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107025802B (en) | 2020-10-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107025802A (en) | A kind of method and unmanned plane that parking stall is found based on unmanned plane | |
| CN109460709B (en) | RTG visual barrier detection method based on RGB and D information fusion | |
| CN110988912B (en) | Road target and distance detection method, system and device for automatic driving vehicle | |
| CN104537360B (en) | Vehicle does not give way peccancy detection method and its detecting system | |
| CN103824066B (en) | A kind of licence plate recognition method based on video flowing | |
| CN110223511A (en) | A kind of automobile roadside is separated to stop intelligent monitoring method and system | |
| CN106203265A (en) | A kind of Construction Fugitive Dust Pollution based on unmanned plane collection image is derived from dynamic monitoring and coverage prognoses system and method | |
| CN106251695A (en) | Destination's parking stall intelligent recommendation system and method based on parking space state monitoring | |
| CN106971185B (en) | License plate positioning method and device based on full convolution network | |
| CN109949593A (en) | A kind of traffic lights recognition methods and system based on crossing priori knowledge | |
| CN109740484A (en) | The method, apparatus and system of road barrier identification | |
| CN106570439B (en) | Vehicle detection method and device | |
| CN113808098A (en) | Road disease identification method and device, electronic equipment and readable storage medium | |
| CN110458050A (en) | Vehicle based on Vehicular video cuts detection method and device | |
| WO2021036559A1 (en) | Automatic vehicle window cleaning method and apparatus | |
| CN112418081B (en) | Method and system for quickly surveying traffic accidents by air-ground combination | |
| CN108550274A (en) | A kind of unmanned auxiliary device of traffic lights based on Faster RCNN and method | |
| CN107909012A (en) | A kind of real-time vehicle tracking detection method and device based on disparity map | |
| CN115273023A (en) | Vehicle-mounted road pothole identification method and system and automobile | |
| CN111753610A (en) | Weather identification method and device | |
| CN111098850A (en) | Automatic parking auxiliary system and automatic parking method | |
| CN110733416B (en) | Lane departure early warning method based on inverse perspective transformation | |
| CN114140396A (en) | Road surface damage detection method, system, device and medium based on unmanned aerial vehicle image | |
| CN111739332B (en) | Parking lot management system | |
| CN105574490A (en) | Vehicle brand identification method and system based on headlight image characteristics |
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 | ||
| CP03 | Change of name, title or address |
Address after: 518066 Room 201, building A, No. 1, Qian Wan Road, Qianhai Shenzhen Hong Kong cooperation zone, Shenzhen, Guangdong (Shenzhen Qianhai business secretary Co., Ltd.) Patentee after: Puzhou Technology (Shenzhen) Co.,Ltd. Address before: 518000 Room 201, building A, 1 front Bay Road, Shenzhen Qianhai cooperation zone, Shenzhen, Guangdong Patentee before: PRODRONE TECHNOLOGY (SHENZHEN) Co.,Ltd. |
|
| CP03 | Change of name, title or address |