CN110211421A - Parking space information intelligent identification Method and system based on cloud - Google Patents
Parking space information intelligent identification Method and system based on cloud Download PDFInfo
- Publication number
- CN110211421A CN110211421A CN201910565499.3A CN201910565499A CN110211421A CN 110211421 A CN110211421 A CN 110211421A CN 201910565499 A CN201910565499 A CN 201910565499A CN 110211421 A CN110211421 A CN 110211421A
- Authority
- CN
- China
- Prior art keywords
- data
- cloud
- cell phone
- phone client
- processing data
- 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
- G08G1/144—Traffic control systems for road vehicles indicating individual free spaces in parking areas with means giving the indication of available parking spaces on portable or mobile units, e.g. personal digital assistant [PDA]
-
- 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
- G08G1/145—Traffic control systems for road vehicles indicating individual free spaces in parking areas where the indication depends on the parking areas
- G08G1/148—Management of a network of parking areas
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
Abstract
The present invention discloses a kind of parking space information intelligent identification Method and system based on cloud, and method includes: that the central processing unit in bicycle recording apparatus carries out processing to driving video and formed just to handle data, and just handling data includes the picture with parking stall number；First processing data are sent to cell phone client by the bluetooth port in bicycle recording apparatus；Cell phone client receives and just handles data, and first processing data are uploaded to cloud server；Cloud server, which receives, just handles data；Operational model in cloud server is trained study and verification based on initial data in the cloud database in first processing data and cloud server, and first processing data are identified to obtain recognition result, cell phone client is sent to by matching degree in recognition result is highest, while will just handle data and cloud database is added；Cell phone client receives the highest recognition result of matching degree.The present invention can identify that the vehicle for finding oneself for user provides help to parking space information.
Description
Technical field
The present invention relates to parking space informations to identify field, intelligently knows more particularly, to a kind of parking space information based on cloud
Other method and system.
Background technique
With the increasingly raising of people 's material life, the number of car owners is being continuously increased, so that corresponding trolley stops
Parking lot also the more is repaired the more big, found in the case where no auxiliary among large parking lot oneself automobile undoubtedly have it is certain
Difficulty.In a kind of automobile data recorder (application number 2013100584910) based on block chain technology that the prior art provides, by pair
Camera and master record instrument composition.Main camera is also had on master record instrument, inside includes PCB main board, can be stored, be located
Reason and sensing.Master record instrument has WIFI module, and can be wireless each automobile data recorder networking, constitutes cloud data platform.
There are the following problems for automobile data recorder in the prior art:
1, the building of cloud data platform is needed by internet database technology, although open and clearization may be implemented,
It is that its operation needs to provide preferable network condition and power supply, is difficult to run in the case where no connection network；
2, source is closed, only developed, safeguarded by fixed incorporated business, exchange of technology, technology development and technology are unfavorable for
Accumulation, be also unfavorable for forming unified professional standard, be unfavorable for the sound development of the industry on long terms；
3, function is more single, is confined to writing function, can not achieve the identification of parking space information.
Summary of the invention
In view of this, the method and system of the present invention provides a kind of parking space information intelligent recognition based on cloud, it can
Parking space information is identified, the vehicle for finding oneself for user provides help.
In order to solve the above-mentioned technical problem, the present invention proposes a kind of parking space information intelligent identification Method based on cloud, packet
Include: the central processing unit in bicycle recording apparatus carries out processing to driving video and forms just processing data, wherein just handles data
Including the picture with parking stall number；First processing data are sent to cell phone client by the bluetooth port in bicycle recording apparatus；Mobile phone
Client receives and just handles data, and first processing data are uploaded to cloud server；Cloud server, which receives, just handles data；
Operational model in cloud server is carried out based on initial data in the cloud database in first processing data and cloud server
Training study and verification, and first processing data are identified to obtain recognition result, by the highest hair of matching degree in recognition result
It send to cell phone client, while will just handle data and cloud database is added；Wherein, the operational model in cloud server is based on
Just initial data carries out two-way training study and verification in the cloud database in processing data and cloud server, comprising: cloud
The operational model in server is held to read in initial data in the cloud database in just processing data and cloud server；To just it locate
Reason data and initial data are integrally divided into learning data and verification two parts of data；It is created using support vector machine algorithm
Binary class model builds the classification mould of multiclass first from one group of binary classification model point on the basis of binary class model
Learning data the first disaggregated model of training is added in type, and verification data the first disaggregated model resulting to training is added and scores；
The second disaggregated model of multiclass is created using decision clump calculating method, learning data the second disaggregated model of training is added, check number is added
It scores according to resulting second disaggregated model of training；The first disaggregated model and the second classification mould are assessed using standard index
Type；Cell phone client receives the highest recognition result of matching degree.
Preferably, wherein further include: cell phone client judge whether obtain permission, permission include at least bluetooth, file and
Mobile phone positioning；When not obtaining permission, user is reminded to authorize；When obtaining permission, cell phone client receives just processing
Data, and first processing data are uploaded to cloud server.
Preferably, wherein in the initial state, initial data includes ideal data and experimental data to cloud database,
In, ideal data is the data formed according to experiment condition, and experimental data is according to experiment condition and practical factor to be added and is formed
Data, practical factor include at least light and angle.
Preferably, wherein the central processing unit in bicycle recording apparatus carries out processing to driving video and forms just processing number
According to, comprising: the central processing unit in bicycle recording apparatus carries out picture interception to driving video according to first time interval, obtains
Picture is intercepted, wherein first time interval is set according to the processing speed of central processing unit；Central processing unit will intercept in picture
Picture with parking stall number, which is picked out to be formed, just handles data.
Preferably, wherein further include: the GPS module in bicycle recording apparatus records road location；Driving recording
Road location is sent to cell phone client by the bluetooth port in device；Cell phone client receives road location.
The present invention also proposes a kind of parking space information intelligent identifying system based on cloud, comprising: bicycle recording apparatus, mobile phone
Client and cloud server；Bicycle recording apparatus includes central processing unit and bluetooth port；Wherein, central processing unit is used for
Processing is carried out to driving video and forms just processing data, just handling data includes the picture with parking stall number；Bluetooth port, being used for will
Just processing data are sent to cell phone client；Cell phone client includes file module and cloud module；Wherein, file module is used
Data are just handled in receiving；Cloud module, for first processing data to be uploaded to cloud server；Cloud server is also used to
Study and verification are trained to operational model by just handling initial data in data and cloud database, and pass through operation mould
Type is identified to obtain recognition result to first processing data, is sent to cell phone client for matching degree in recognition result is highest,
Data will be just handled simultaneously, and cloud database is added；Wherein, operational model includes: reading module, just handles data for reading in
With initial data in the cloud database in cloud server；Division module, it is whole for will just handle data and initial data
It is divided into learning data and verification two parts of data；First study correction verification module, for using support vector machine algorithm to create
Binary class model is built, builds the classification mould of multiclass first from one group of binary classification model point on the basis of binary class model
Learning data the first disaggregated model of training is added in type, and verification data the first disaggregated model resulting to training is added and scores；
Learning data training the is added for creating the second disaggregated model of multiclass using decision clump calculating method in second study correction verification module
Two disaggregated models are added verification data the second disaggregated model resulting to training and score；Evaluation module, for using standard
The first disaggregated model of index evaluation and the second disaggregated model；Identification module class is known for being identified to first processing data
Other result；Cloud module is also used to receive the highest recognition result of matching degree.
Preferably, cell phone client further includes authority module；Authority module, for judging whether cell phone client is weighed
Limit, permission include at least bluetooth, file and mobile phone positioning；Wherein, when not obtaining permission, authority module reminds user to carry out
Authorization；When obtaining the permission, cell phone client receives and just handles data, and first processing data are uploaded to cloud service
Device.
Preferably, wherein in the initial state, initial data includes ideal data and experimental data to cloud database,
In, ideal data is the data formed according to experiment condition, and experimental data is according to experiment condition and practical factor to be added and is formed
Data, practical factor include at least light and angle.
Preferably, wherein central processing unit is obtained for carrying out picture interception to driving video according to first time interval
To interception picture, wherein first time interval is set according to the processing speed of central processing unit；Central processing unit is also used to cut
The picture in picture with parking stall number is taken to pick out to form just processing data.
Preferably, wherein bicycle recording apparatus further includes GPS module；GPS module, for being recorded to road location；
Bluetooth port is also used to road location being sent to cell phone client；File module is also used to receive road location.
Compared with prior art, the parking space information intelligent identification Method and system provided by the invention based on cloud is realized
It is following the utility model has the advantages that
(1) the parking space information intelligent identification Method and system of the present invention based on cloud, in bicycle recording apparatus
Central processing unit first carries out processing formation to driving video and includes the first processing data of the picture with parking stall number, and will just handle number
According to cell phone client is sent to, cell phone client, which receives, just handles data, only needs between bicycle recording apparatus and cell phone client
Transmission just handles data, reduces the transmission informational capacity between bicycle recording apparatus and cell phone client, using electricity wisely and subtracts
The occupied space of few cell phone client.
(2) the parking space information intelligent identification Method and system of the present invention based on cloud, cell phone client receive just
Data are handled, and first processing data are uploaded to cloud server, using the operational model in cloud server to first processing number
According to being identified to obtain recognition result, cell phone client is sent to by matching degree in recognition result is highest, passes through cloud service
Operational model in device can carry out intelligent recognition to parking space information, while recognition result is sent to cell phone client, can be effective
Auxiliary user searches out the vehicle of oneself in parking lot.
(3) the parking space information intelligent identification Method and system of the present invention based on cloud, the fortune in cloud server
It calculates model and study and verification is trained based on initial data in the cloud database in first processing data and cloud server, it can
It pushes operational model to continue to optimize raising model performance by user feedback data, avoids excessively managing due to undersampling and sampling
Wanting bring error, and continuing to optimize by operational model, can independently adapt to various novel scene conditions, reduce the later period
Maintenance cost.
(4) the parking space information intelligent identification Method and system of the present invention based on cloud, operational model are located at cloud
In server, reduce the operation demand of cell phone client, reduce and be applicable in threshold, the system that can also reduce integrally consumes energy, and improves
System cruising ability.
(5) the parking space information intelligent identification Method and system of the present invention based on cloud, the fortune in cloud server
It calculates model and two-way training study and school is carried out based on initial data in the cloud database in first processing data and cloud server
It tests, effectively improves the performance of operational model, to improve the accuracy identified to parking space information.
(6) the parking space information intelligent identification Method and system of the present invention based on cloud, bicycle recording apparatus record
Road location, is transferred to cell phone client for road location, and user is further assisted to search out in parking lot the vehicle of oneself
?.
Certainly, implementing any of the products of the present invention specific needs while need not reach all the above technical effect.
By referring to the drawings to the detailed description of exemplary embodiment of the present invention, other feature of the invention and its
Advantage will become apparent.
Detailed description of the invention
It is combined in the description and the attached drawing for constituting part of specification shows the embodiment of the present invention, and even
With its explanation together principle for explaining the present invention.
Fig. 1 is a kind of flow chart of parking space information intelligent identification Method based on cloud proposed by the present invention；
Fig. 2 is the flow chart for the method that a kind of operational model proposed by the present invention is trained study and verification；
Fig. 3 is the flow chart of another parking space information intelligent identification Method based on cloud proposed by the present invention；
Fig. 4 is a kind of composition schematic diagram of parking space information intelligent identifying system based on cloud proposed by the present invention；
Fig. 5 is a kind of structural schematic diagram of bicycle recording apparatus proposed by the present invention；
Fig. 6 is a kind of structural schematic diagram of cell phone client proposed by the present invention；
Fig. 7 is a kind of structural schematic diagram of cloud server proposed by the present invention.
Specific embodiment
Carry out the various exemplary embodiments of detailed description of the present invention now with reference to attached drawing.It should also be noted that unless in addition having
Body explanation, the unlimited system of component and the positioned opposite of step, numerical expression and the numerical value otherwise illustrated in these embodiments is originally
The range of invention.
Be to the description only actually of at least one exemplary embodiment below it is illustrative, never as to the present invention
And its application or any restrictions used.
Technology, method and apparatus known to person of ordinary skill in the relevant may be not discussed in detail, but suitable
In the case of, the technology, method and apparatus should be considered as part of specification.
It is shown here and discuss all examples in, any occurrence should be construed as merely illustratively, without
It is as limitation.Therefore, other examples of exemplary embodiment can have different values.
It should also be noted that similar label and letter indicate similar terms in following attached drawing, therefore, once a certain Xiang Yi
It is defined in a attached drawing, then in subsequent attached drawing does not need that it is further discussed.
Embodiment 1
Fig. 1 is a kind of flow chart of parking space information intelligent identification Method based on cloud proposed by the present invention, and Fig. 2 is this hair
A kind of operational model of bright proposition is trained the flow chart of the method for study and verification, with reference to Fig. 1 and Fig. 2, this method comprises:
Central processing unit carries out processing to driving video and forms just processing data in step 101, bicycle recording apparatus,
In, just handling data includes the picture with parking stall number；
First processing data are sent to cell phone client by bluetooth port in step 102, bicycle recording apparatus；
Step 103, cell phone client receive and just handle data, and first processing data are uploaded to cloud server；
Step 104, cloud server, which receive, just handles data；
Operational model in step 105, cloud server is based on the cloud data in first processing data and cloud server
Initial data is trained study and verification in library, and identifies to first processing data, by matching degree highest in recognition result
Be sent to cell phone client, while will just handle data be added cloud database；
Step 106, cell phone client receive the highest recognition result of matching degree.
Wherein, the operational model in step 105 in cloud server is based on the cloud in first processing data and cloud server
Initial data carries out two-way training study and verification in client database, comprising:
Operational model in step 201, cloud server reads in the cloud data in just processing data and cloud server
Initial data in library；
First processing data and initial data are integrally divided into learning data and verification two parts of data by step 202；
Wherein, illustratively, will just processing data and initial data it is whole 80% be divided into learning data, by its 20%
It is divided into verification data；
Step 203 creates binary class model using support vector machine algorithm, on the basis of binary class model
On from one group of binary classification model point build the first disaggregated model of multiclass, learning data the first disaggregated model of training is added, school is added
Data the first disaggregated model resulting to training is tested to score；
Wherein, binary class model for realizing based on classification production forecast possible outcome, can only there are two prediction knot
Fruit, in initial designs, this part the number of iterations is 1, has the function of standardization；When result depends on predictive variable, first
Disaggregated model predicts more possible outcomes for creating；
Step 204, decision clump calculating method create the second disaggregated model of multiclass, and learning data training the second classification mould is added
Type is added verification data the second disaggregated model resulting to training and scores；
Wherein, the second disaggregated model is used for Definition Model and parameter, and then the training dataset of linkage flag is to use it
In a training pattern, the target with multiple probable values can be trained, in initial designs, this part method for resampling is to be packaged
Operation, maximum decision number are set as 8, and maximum decision depth is 32, and maximum decision width is 128, and the optimization of each decision-making level walks
Rapid number is 2048；
Step 205 assesses the first disaggregated model and the second disaggregated model using standard index.
Middle rolling car recording device of the present invention first carries out the first place that processing formation includes the picture with parking stall number to driving video
Manage data, and just processing data will be sent to cell phone client, cell phone client receives and just handles data, bicycle recording apparatus and
It need to only be transmitted between cell phone client and just handle data, reduce the transmission information between bicycle recording apparatus and cell phone client
Total amount, using electricity wisely and the occupied space for reducing cell phone client.
Cell phone client receives and just handles data, and first processing data are uploaded to cloud server, uses cloud service
Operational model in device is identified to obtain recognition result to first processing data, is sent to matching degree in recognition result is highest
Cell phone client can carry out intelligent recognition by the operational model in cloud server to parking space information, while by recognition result
It is sent to cell phone client, user can effectively be assisted to search out the vehicle of oneself in parking lot.Operation in cloud server
Model is trained study and verification based on initial data in the cloud database in first processing data and cloud server, can lead to
Crossing user feedback data pushes operational model to continue to optimize raising model performance, avoids since undersampling and sampling are excessively ideal
Change bring error, and continuing to optimize by operational model, can independently adapt to various novel scene conditions, reduces later period dimension
Protect cost.Operational model is located in cloud server, reduces the operation demand of cell phone client, reduces and is applicable in threshold, may be used also
It is integrally consumed energy with reducing system, improves system cruising ability.
First processing data are sent to cell phone client by blue-teeth data sending function by bluetooth port, and transmission speed is suitable
In, transmission channel can be commonly used to avoid other are occupied, and still can carry out information transmission in no network condition, improved
The adaptive capacity to environment and the scope of application of device.
Operational model in cloud server is based on original in the cloud database in first processing data and cloud server
Data carry out two-way training study and verification, effectively improve the performance of operational model, identify to improve to parking space information
Accuracy.
Embodiment 2
Fig. 3 is the flow chart of another parking space information intelligent identification Method based on cloud proposed by the present invention, with reference to Fig. 2
And Fig. 3, this method comprises:
Central processing unit in step 301, bicycle recording apparatus carries out processing to driving video and forms just processing data,
In, just handling data includes the picture with parking stall number；
First processing data are sent to cell phone client by the bluetooth port in step 302, bicycle recording apparatus；
Step 303, cell phone client judge whether that acquisition permission, permission include at least bluetooth, file and mobile phone positioning；
When not obtaining the permission, step 304 is executed；
When obtaining the permission, step 305 is executed；
Step 304 reminds user to authorize；
Step 305, cell phone client receive and just handle data, and first processing data are uploaded to cloud server；
Step 306, cloud server, which receive, just handles data；
Operational model in step 307, cloud server is based on the cloud data in first processing data and cloud server
Initial data is trained study and verification in library, and identifies to first processing data, by matching degree highest in recognition result
Be sent to cell phone client, while will just handle data be added cloud database；
Step 308, cell phone client receive the highest recognition result of matching degree.
Wherein, the operational model in step 307 in cloud server is based on the cloud in first processing data and cloud server
Initial data carries out two-way training study and verification in client database, comprising:
Operational model in step 201, cloud server reads in the cloud data in just processing data and cloud server
Initial data in library；
First processing data and initial data are integrally divided into learning data and verification two parts of data by step 202；
Wherein, illustratively, will just processing data and initial data it is whole 80% be divided into learning data, by its 20%
It is divided into verification data；
Step 203 creates binary class model using support vector machine algorithm, on the basis of binary class model
On from one group of binary classification model point build the first disaggregated model of multiclass, learning data the first disaggregated model of training is added, school is added
Data the first disaggregated model resulting to training is tested to score；
Wherein, binary class model for realizing based on classification production forecast possible outcome, can only there are two prediction knot
Fruit, in initial designs, this part the number of iterations is 1, has the function of standardization；When result depends on predictive variable, first
Disaggregated model predicts more possible outcomes for creating；
Step 204, decision clump calculating method create the second disaggregated model of multiclass, and learning data training the second classification mould is added
Type is added verification data the second disaggregated model resulting to training and scores；
Wherein, the second disaggregated model is used for Definition Model and parameter, and then the training dataset of linkage flag is to use it
In a training pattern, the target with multiple probable values can be trained, in initial designs, this part method for resampling is to be packaged
Operation, maximum decision number are set as 8, and maximum decision depth is 32, and maximum decision width is 128, and the optimization of each decision-making level walks
Rapid number is 2048；
Step 205 assesses the first disaggregated model and the second disaggregated model using standard index.
Cell phone client judges whether to obtain a variety of permissions such as bluetooth, file and mobile phone positioning, mention when not obtaining permission
Awake user authorizes, and after being authorized, cell phone client receives and just handles data, and first processing data are uploaded to cloud
Hold server.
Wherein, in the initial state, initial data includes ideal data and experimental data to cloud database, wherein ideal
Data are the data formed according to experiment condition, and experimental data is and the data of practical factor formation to be added according to experiment condition,
Practical factor includes at least light and angle.
Operational model in cloud server is based on original in the cloud database in first processing data and cloud server
Data are trained study and verification, and in the initial state, initial data includes ideal data and experimental data to cloud database,
Initial data is added in user feedback data, and pushes operational model to continue to optimize by user feedback data and improves model
Energy is avoided excessively being idealized bring error due to undersampling and sampling, and by the gradually abundant of initial data and transported
Continuing to optimize for model is calculated, various novel scene conditions can be independently adapted to, reduces later maintenance cost.
Wherein, step 301, the central processing unit in bicycle recording apparatus carry out processing to driving video and are formed just to handle number
According to, comprising:
Central processing unit in bicycle recording apparatus carries out picture interception to driving video according to first time interval, obtains
Picture is intercepted, wherein first time interval is set according to the processing speed of central processing unit；
Central processing unit will intercept the picture with parking stall number in picture and pick out to form just processing data.
Wherein, the parking space information intelligent identification Method proposed by the present invention based on cloud further include:
GPS module in bicycle recording apparatus records road location；
Road location is sent to cell phone client by the bluetooth port in bicycle recording apparatus；
Cell phone client receives road location.
GPS module has recorded road location, and road location is transferred to cell phone client, further user is assisted to stop
The vehicle of oneself is searched out in.
Wherein, the parking space information intelligent identification Method proposed by the present invention based on cloud further include:
Camera shooting driving video in bicycle recording apparatus；
Display in bicycle recording apparatus carries out real-time display to driving video；
Storage card in bicycle recording apparatus will drive a vehicle video and road location stores.
Illustratively, cell phone client is developed based on android system；Operational model be trained BP neural network with
And the cloud learning model of initial training, service can be provided in networking and failed cluster；Operational model is using Microsoft
Microsoft azure machine learning studio carries out modeling operation.
Embodiment 3
Fig. 4 is a kind of composition schematic diagram of parking space information intelligent identifying system based on cloud proposed by the present invention, and Fig. 5 is
A kind of structural schematic diagram of bicycle recording apparatus proposed by the present invention, Fig. 6 are a kind of knots of cell phone client proposed by the present invention
Structure schematic diagram, Fig. 7 are a kind of structural schematic diagrams of cloud server proposed by the present invention, and with reference to Fig. 4-Fig. 7, the present invention provides one
Parking space information intelligent identifying system of the kind based on cloud, comprising: bicycle recording apparatus 10, cell phone client 20 and cloud server
30；
Bicycle recording apparatus 10 includes central processing unit 11 and bluetooth port 12；Wherein,
Central processing unit 11 forms just processing data for carrying out processing to driving video, and just handling data includes band vehicle
The picture of position number；
Bluetooth port 12, for first processing data to be sent to cell phone client 20；
Cell phone client 20 includes file module 21 and cloud module 22；Wherein,
File module 21 just handles data for receiving；
Cloud module 22, for first processing data to be uploaded to cloud server 30；
Cloud server 30 just handles data for receiving；
Cloud server 30, be also used to by just handle in data and cloud database 31 initial data to operational model into
Row training study and verification, and first processing data are identified to obtain recognition result by operational model 32, by recognition result
Middle matching degree is highest to be sent to the cell phone client 20, while will just handle data and cloud database 31 is added；
Cloud module 22 is also used to receive the highest recognition result of matching degree.
Central processing unit 11 first carries out processing to driving video to be formed including band vehicle in middle rolling car recording device 10 of the present invention
The first processing data of the picture of position number, then first processing data are sent to by cell phone client 20, mobile phone visitor by bluetooth port 12
Family end 20, which receives, just handles data, need to only transmit between bicycle recording apparatus 10 and cell phone client 20 and just handle data, reduce
Transmission informational capacity between bicycle recording apparatus 10 and cell phone client 20, using electricity wisely and reduces cell phone client 10
Occupied space.First processing data are sent to cell phone client 20, transmission speed by blue-teeth data sending function by bluetooth port 12
Spend it is moderate, can be to avoid occupying other common transmission channels, and information transmission still can be carried out in no network condition, mention
The high adaptive capacity to environment and the scope of application of device.
File module 21 receives and just handles data in cell phone client 20, and will just be handled in data by cloud module 22
Cloud server 30 is reached, first processing data are identified to obtain identification knot using the operational model 32 in cloud server 30
Fruit is sent to cell phone client 20 for matching degree in recognition result is highest, by the operational model 32 in cloud server 30,
Intelligent recognition can be carried out to parking space information, while recognition result is sent to cell phone client 20, can effectively assist user stopping
The vehicle of oneself is searched out in parking lot.Operational model 32 in cloud server 30 is based on just processing data and cloud server 30
Initial data is trained study and verification in interior cloud database 31, can push operational model 32 by user feedback data
Raising model performance is continued to optimize, avoids excessively idealizing bring error due to undersampling and sampling, and pass through operation mould
Type 32 is continued to optimize, and various novel scene conditions can be independently adapted to, and reduces later maintenance cost.Operational model 32 is located at cloud
It holds in server 30, reduces the operation demand of cell phone client 20, reduce and be applicable in threshold, device can also be reduced and integrally consumed
Can, improve the cruising ability of device.
Wherein, the operational model 32 in cloud server 30 includes:
Module 321 is read in, for reading in original number in the cloud database 31 in just processing data and cloud server 30
According to；
Division module 322, for first processing data and initial data to be integrally divided into learning data and verification data two
A part；
First study correction verification module 323, for using support vector machine algorithm create binary class model, two into
The first disaggregated model of multiclass is built from one group of binary classification model point on the basis of disaggregated model processed, learning data training first is added
Disaggregated model is added verification data the first disaggregated model resulting to training and scores；
Study is added for creating the second disaggregated model of multiclass using decision clump calculating method in second study correction verification module 324
Data train the second disaggregated model, and verification data the second disaggregated model resulting to training is added and scores；
Evaluation module 325, for assessing the first disaggregated model and the second disaggregated model using standard index；
Identification module 326, for being identified to obtain recognition result to first processing data.
Operational model 32 in cloud server 30 provided by the invention can be based on just processing data and cloud server 31
Initial data carries out two-way training study and verification in interior cloud database, effectively improves the performance of operational model 32, thus
Improve the accuracy identified to parking space information.
Wherein, cell phone client 20 further includes authority module 23；
Authority module 23, for judging whether cell phone client 20 obtains permission, permission includes at least bluetooth, file and hand
Machine positioning；
When not obtaining permission, authority module 23 reminds user to authorize；
When obtaining permission, cell phone client 20 receives and just handles data, and will just handle data by cloud module 22
It is uploaded to cloud server 30.
Wherein, in the initial state, initial data includes ideal data and experimental data to cloud database 31, wherein reason
Think that data are the data formed according to experiment condition, experimental data is and the number of practical factor formation to be added according to experiment condition
According to practical factor includes at least light and angle.
Operational model 32 in cloud server 30 is based on the cloud database 31 in first processing data and cloud server
Middle initial data is trained study and verification, cloud database 31 in the initial state, initial data include ideal data and
User feedback data is added initial data, and pushes operational model 32 to continue to optimize by user feedback data by experimental data
Model performance is improved, avoids excessively idealizing bring error due to undersampling and sampling, and gradually by initial data
It is abundant and operational model 32 to continue to optimize, various novel scene conditions can be independently adapted to, later maintenance cost is reduced.
Wherein, central processing unit 11 are intercepted for carrying out picture interception to driving video according to first time interval
Picture, first time interval are set according to the processing speed of central processing unit 11；
Central processing unit 11 is also used to pick out the picture with parking stall number in interception picture to form just processing data.
Wherein, file module 21 is also used to store data, call, handles, shows, uploads in cell phone client 20
With receive etc..
Wherein, bicycle recording apparatus 10 further includes display 13, storage card 14, power supply 15, camera 16 and GPS module
17；
Camera 16, for shooting driving video；GPS module 17, for recording road location；Display 13, for pair
Video of driving a vehicle carries out real-time display；Storage card 14, for that will drive a vehicle video and road location stores；Bluetooth port 12, also
For road location to be sent to cell phone client 20；Power supply 15, for providing energy consumption for each component in bicycle recording apparatus 10.
It is equipped with GPS module 17 in bicycle recording apparatus 10, its position can be recorded while recording driving conditions and be used for
Navigation, while help is provided for the vehicle that user finds oneself.
Illustratively, the model STA32F767 of central processing unit 11, the model ATK-HC05- of bluetooth port 12
V13, the model ATK-7RGB TFTLC-V13 of display 13, the model SD-K08G of storage card 14, power supply 15 are 5V direct current
Power supply, the model ATK-OV5640AF-V11 of camera 16, the model ATK-S126-V1.1 of GPS module 17.
By above each embodiment it is found that parking space information intelligent identification Method provided by the invention based on cloud and being
Existing beneficial effect of uniting is:
(1) the parking space information intelligent identification Method and system of the present invention based on cloud, in bicycle recording apparatus
Central processing unit first carries out processing formation to driving video and includes the first processing data of the picture with parking stall number, and will just handle number
According to cell phone client is sent to, cell phone client, which receives, just handles data, only needs between bicycle recording apparatus and cell phone client
Transmission just handles data, reduces the transmission informational capacity between bicycle recording apparatus and cell phone client, using electricity wisely and subtracts
The occupied space of few cell phone client.
(2) the parking space information intelligent identification Method and system of the present invention based on cloud, cell phone client receive just
Data are handled, and first processing data are uploaded to cloud server, using the operational model in cloud server to first processing number
According to being identified to obtain recognition result, cell phone client is sent to by matching degree in recognition result is highest, passes through cloud service
Operational model in device can carry out intelligent recognition to parking space information, while recognition result is sent to cell phone client, can be effective
Auxiliary user searches out the vehicle of oneself in parking lot.
(3) the parking space information intelligent identification Method and system of the present invention based on cloud, the fortune in cloud server
It calculates model and study and verification is trained based on initial data in the cloud database in first processing data and cloud server, it can
It pushes operational model to continue to optimize raising model performance by user feedback data, avoids excessively managing due to undersampling and sampling
Wanting bring error, and continuing to optimize by operational model, can independently adapt to various novel scene conditions, reduce the later period
Maintenance cost.
(4) the parking space information intelligent identification Method and system of the present invention based on cloud, operational model are located at cloud
In server, reduce the operation demand of cell phone client, reduce and be applicable in threshold, the system that can also reduce integrally consumes energy, and improves
System cruising ability.
(5) the parking space information intelligent identification Method and system of the present invention based on cloud, the fortune in cloud server
It calculates model and two-way training study and school is carried out based on initial data in the cloud database in first processing data and cloud server
It tests, effectively improves the performance of operational model, to improve the accuracy identified to parking space information.
(6) the parking space information intelligent identification Method and system of the present invention based on cloud, bicycle recording apparatus record
Road location, is transferred to cell phone client for road location, and user is further assisted to search out in parking lot the vehicle of oneself
?.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, apparatus or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
Although some specific embodiments of the invention are described in detail by example, the skill of this field
Art personnel it should be understood that example above merely to being illustrated, the range being not intended to be limiting of the invention.The skill of this field
Art personnel are it should be understood that can without departing from the scope and spirit of the present invention modify to above embodiments.This hair
Bright range is defined by the following claims.
Claims (10)
1. a kind of parking space information intelligent identification Method based on cloud characterized by comprising
Central processing unit in bicycle recording apparatus carries out processing to driving video and forms just processing data, wherein the just place
Managing data includes the picture with parking stall number；
The just processing data are sent to cell phone client by the bluetooth port in the bicycle recording apparatus；
The cell phone client receives the just processing data, and the just processing data are uploaded to cloud server；
The cloud server receives the just processing data；
Operational model in the cloud server is based on the cloud data in just the processing data and the cloud server
Initial data is trained study and verification in library, and is identified to obtain recognition result to the just processing data, will be described
Matching degree is highest in recognition result is sent to the cell phone client, while the cloud number is added in the just processing data
According to library；Wherein,
Operational model in the cloud server is based on the cloud data in just the processing data and the cloud server
Initial data carries out two-way training study and verification in library, comprising:
The operational model in the cloud server reads in the cloud in just the processing data and the cloud server
Initial data in database；
Just processing data and the initial data are integrally divided into learning data and verification two parts of data；
Binary class model is created using support vector machine algorithm, from one group on the basis of the binary class model
Binary classification model point builds the first disaggregated model of multiclass, and the learning data training first disaggregated model is added, institute is added
Verification data first disaggregated model resulting to training is stated to score；
The second disaggregated model of multiclass is created using decision clump calculating method, the learning data training the second classification mould is added
Type is added verification data second disaggregated model resulting to training and scores；
First disaggregated model and second disaggregated model are assessed using standard index；
The cell phone client receives the highest recognition result of matching degree.
2. the method for the parking space information intelligent recognition according to claim 1 based on cloud, which is characterized in that further include:
The cell phone client judges whether that acquisition permission, the permission include at least bluetooth, file and mobile phone positioning；
When not obtaining the permission, user is reminded to authorize；
When obtaining the permission, the cell phone client receives the just processing data, and the just processing data are uploaded
To cloud server.
3. the method for the parking space information intelligent recognition according to claim 1 based on cloud, which is characterized in that
In the initial state, the initial data includes ideal data and experimental data to the cloud database, wherein the reason
Think that data are the data formed according to experiment condition, the experimental data is according to experiment condition and practical factor to be added and is formed
Data, the practical factor include at least light and angle.
4. the method for the parking space information intelligent recognition according to claim 1 based on cloud, which is characterized in that
Central processing unit in the bicycle recording apparatus carries out processing to driving video and forms just processing data, comprising:
The central processing unit in the bicycle recording apparatus carries out picture interception to driving video according to first time interval,
Interception picture is obtained, wherein the first time interval is set according to the processing speed of the central processing unit；
The central processing unit picks out the picture with parking stall number in the interception picture to form the just processing data.
5. the method for the parking space information intelligent recognition according to claim 1 based on cloud, which is characterized in that further include:
GPS module in the bicycle recording apparatus records road location；
The road location is sent to cell phone client by the bluetooth port in the bicycle recording apparatus；
The cell phone client receives the road location.
6. a kind of parking space information intelligent identifying system based on cloud characterized by comprising bicycle recording apparatus, mobile phone visitor
Family end and cloud server；
The bicycle recording apparatus includes central processing unit and bluetooth port；Wherein,
The central processing unit forms just processing data for carrying out processing to driving video, and the just processing data include band
The picture of parking stall number；
The bluetooth port, for the just processing data to be sent to cell phone client；
The cell phone client includes file module and cloud module；Wherein,
The file module, for receiving the just processing data；
The cloud module, for the just processing data to be uploaded to cloud server；
The cloud server, for receiving the just processing data；
The cloud server, be also used to by it is described just processing data and cloud database in initial data to operational model into
Row training study and verification, and the just processing data are identified to obtain recognition result by the operational model, by institute
It states that matching degree in recognition result is highest to be sent to the cell phone client, while the cloud is added in the just processing data
Database；
The operational model includes:
Module is read in, for reading in initial data in the cloud database in just the processing data and the cloud server；
Division module, for just processing data and the initial data to be integrally divided into learning data and verification data two
A part；
First study correction verification module, for creating binary class model using support vector machine algorithm, in the binary system
The first disaggregated model of multiclass is built from one group of binary classification model point on the basis of disaggregated model, learning data training institute is added
The first disaggregated model is stated, verification data first disaggregated model resulting to training is added and scores；
The study number is added for creating the second disaggregated model of multiclass using decision clump calculating method in second study correction verification module
According to training second disaggregated model, verification data second disaggregated model resulting to training is added and scores；
Evaluation module, for using standard index to assess first disaggregated model and second disaggregated model；
Identification module, for being identified to obtain recognition result to the just processing data；
The cloud module is also used to receive the highest recognition result of matching degree.
7. the parking space information intelligent identifying system according to claim 6 based on cloud, which is characterized in that
The cell phone client further includes authority module；
The authority module, for judging whether the cell phone client obtains permission, the permission includes at least bluetooth, file
It is positioned with mobile phone；Wherein,
When not obtaining the permission, the authority module reminds user to authorize；
When obtaining the permission, the cell phone client receives the just processing data, and the just processing data are uploaded
To cloud server.
8. the parking space information intelligent identifying system according to claim 6 based on cloud, which is characterized in that
In the initial state, the initial data includes ideal data and experimental data to the cloud database, wherein the reason
Think that data are the data formed according to experiment condition, the experimental data is according to experiment condition and practical factor to be added and is formed
Data, the practical factor include at least light and angle.
9. the parking space information intelligent identifying system according to claim 6 based on cloud, which is characterized in that
The central processing unit, for, to driving video progress picture interception, obtaining interception picture according to first time interval,
Described in first time interval according to the processing speed of the central processing unit set；
The central processing unit is also used to pick out in the picture with parking stall number in the interception picture to form the just processing
Data.
10. the parking space information intelligent identifying system according to claim 6 based on cloud, which is characterized in that the driving
Recording device further includes GPS module；
The GPS module, for being recorded to road location；
The bluetooth port is also used to the road location being sent to cell phone client；
The file module is also used to receive the road location.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910565499.3A CN110211421B (en) | 2019-06-27 | 2019-06-27 | Cloud-based parking space information intelligent identification method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910565499.3A CN110211421B (en) | 2019-06-27 | 2019-06-27 | Cloud-based parking space information intelligent identification method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110211421A true CN110211421A (en) | 2019-09-06 |
| CN110211421B CN110211421B (en) | 2021-06-04 |
Family
ID=67794958
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910565499.3A Active CN110211421B (en) | 2019-06-27 | 2019-06-27 | Cloud-based parking space information intelligent identification method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110211421B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101727755A (en) * | 2009-11-20 | 2010-06-09 | 杨沁沁 | Inquiring system for car location in garage |
| CN103473950A (en) * | 2012-06-06 | 2013-12-25 | 刘鉵 | Parking lot parking space monitoring method |
| 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 |
| CN104616525A (en) * | 2015-01-04 | 2015-05-13 | 深圳市安智车米汽车信息化有限公司 | Method and device for obtaining vehicle parking position information |
| CN106251288A (en) * | 2016-07-21 | 2016-12-21 | 北京理工大学 | The twin-line array image device subpixel image synthetic method analyzed based on Multi-resolution |
| JP2017004241A (en) * | 2015-06-10 | 2017-01-05 | 株式会社駐車場綜合研究所 | Parking position notification device, mobile terminal, parking position notification system, and program |
| US20170116858A1 (en) * | 2015-10-27 | 2017-04-27 | FUJllTSU LIMITED | Parking space detection apparatus and method, electronic apparatus |
| WO2018025066A1 (en) * | 2016-08-04 | 2018-02-08 | Fawaz Sultan | Parked car locator using qr codes |
| CN109035859A (en) * | 2018-09-07 | 2018-12-18 | 北京车联天下信息技术有限公司 | A kind of parking position localization method and device |
| CN109872360A (en) * | 2019-01-31 | 2019-06-11 | 斑马网络技术有限公司 | Localization method and device, storage medium, electric terminal |
-
2019
- 2019-06-27 CN CN201910565499.3A patent/CN110211421B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101727755A (en) * | 2009-11-20 | 2010-06-09 | 杨沁沁 | Inquiring system for car location in garage |
| CN103473950A (en) * | 2012-06-06 | 2013-12-25 | 刘鉵 | Parking lot parking space monitoring method |
| 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 |
| CN104616525A (en) * | 2015-01-04 | 2015-05-13 | 深圳市安智车米汽车信息化有限公司 | Method and device for obtaining vehicle parking position information |
| JP2017004241A (en) * | 2015-06-10 | 2017-01-05 | 株式会社駐車場綜合研究所 | Parking position notification device, mobile terminal, parking position notification system, and program |
| US20170116858A1 (en) * | 2015-10-27 | 2017-04-27 | FUJllTSU LIMITED | Parking space detection apparatus and method, electronic apparatus |
| CN106251288A (en) * | 2016-07-21 | 2016-12-21 | 北京理工大学 | The twin-line array image device subpixel image synthetic method analyzed based on Multi-resolution |
| WO2018025066A1 (en) * | 2016-08-04 | 2018-02-08 | Fawaz Sultan | Parked car locator using qr codes |
| CN109035859A (en) * | 2018-09-07 | 2018-12-18 | 北京车联天下信息技术有限公司 | A kind of parking position localization method and device |
| CN109872360A (en) * | 2019-01-31 | 2019-06-11 | 斑马网络技术有限公司 | Localization method and device, storage medium, electric terminal |
Non-Patent Citations (1)
| Title |
|---|
| 马爱琴等: "大数据背景下网络借贷信用风险评估", 《宜宾学院学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110211421B (en) | 2021-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fu et al. | Vehicular blockchain-based collective learning for connected and autonomous vehicles | |
| AU2017253916B2 (en) | Systems and methods for recommending an estimated time of arrival | |
| WO2021135653A1 (en) | Method and system for identifying abnormal stay of vehicle | |
| CN109204007B (en) | Unmanned suspension type magnetic suspension train and control method thereof | |
| CN107179769B (en) | AGV quantity configuration method based on real-time scheduling simulation and queuing theory | |
| US10960893B2 (en) | System and method for driver profiling corresponding to automobile trip | |
| US11615710B2 (en) | Providing indicators of availability in a shared vehicle environment | |
| AU2016265691A1 (en) | Parallel processing for solution space partitions | |
| CN110329266A (en) | Can course continuation mileage determine method, apparatus, terminal and storage medium | |
| CN110103953A (en) | For assisting method, equipment, medium and the system of the Driving control of vehicle | |
| CN110371132A (en) | Driver's adapter tube appraisal procedure and device | |
| CN204990432U (en) | Equipment inspection management system based on RFID | |
| CN116957174B (en) | Freight line integrated planning method and system based on data fusion | |
| CN110379165A (en) | A kind of road type prediction technique, device, equipment and storage medium | |
| CN110458604A (en) | A kind of net about driver's evaluation method, device, equipment and storage medium | |
| CN110060658A (en) | A kind of vehicle launch guide device | |
| WO2008126202A1 (en) | Load distribution program for storage system, load distribution method for storage system, and storage management device | |
| CN105812722A (en) | Grain transportation monitoring method and system | |
| CN111259119B (en) | Question recommending method and device | |
| Thies et al. | Project portfolio planning under CO2 fleet emission restrictions in the automotive industry | |
| CN110211421A (en) | Parking space information intelligent identification Method and system based on cloud | |
| US20220067848A1 (en) | Dynamic transport energy control | |
| Feng et al. | Embedded YOLO: A Real‐Time Object Detector for Small Intelligent Trajectory Cars | |
| US20230304806A1 (en) | Graph-based vehicle route optimization with vehicle capacity clustering | |
| Eglynas et al. | Evaluation of the energy consumption of container diesel trucks in a container terminal: A case study at Klaipeda port |
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 |