CN106097763A - A kind of method for detecting parking stalls and device - Google Patents
A kind of method for detecting parking stalls and device Download PDFInfo
- Publication number
- CN106097763A CN106097763A CN201610661972.4A CN201610661972A CN106097763A CN 106097763 A CN106097763 A CN 106097763A CN 201610661972 A CN201610661972 A CN 201610661972A CN 106097763 A CN106097763 A CN 106097763A
- Authority
- CN
- China
- Prior art keywords
- parking stall
- parking
- video camera
- stall measure
- measure video
- 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
- G08G1/145—Traffic control systems for road vehicles indicating individual free spaces in parking areas where the indication depends on the parking areas
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/54—Surveillance or monitoring of activities, e.g. for recognising suspicious objects of traffic, e.g. cars on the road, trains or boats
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/63—Scene text, e.g. street names
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/625—License plates
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a kind of method for detecting parking stalls and device, the parking space state change in the method detection monitoring region, when detecting that parking space state changes, search the parking stall measure video camera being in idle condition, and therefrom choose and accurately detect the parking stall measure video camera with Car license recognition for performing parking space state, the parking stall changed by parking space state processes information and is sent to selected parking stall measure video camera, carries out parking space state and accurately detects and Car license recognition.The invention also discloses the device corresponding to said method, including pre-detection module and accurate detection module.Parking stall is processed information and is sent to multiple parking stall measure video camera and processes by the method and device of the present invention, shortens the detection time, saves network traffics, alleviates network load.
Description
Technical field
The invention belongs to field of video monitoring, particularly relate to a kind of method for detecting parking stalls and device.
Background technology
Along with the development in private car market, each big city more and more considers newly-built wisdom parking lot.Stop in wisdom
In the scheme of field, parking stall measure video camera is one of key technology and equipment.Parking stall measure video camera is mainly used in entering in parking lot
Row parking stall measure and Car license recognition.The most newly-built parking lot scale is increasing, and parking stall gets more and more, and every 1-3 car
Position is accomplished by being equipped with a parking stall measure video camera, and therefore a parking lot needs the parking stall measure video camera being equipped with the most,
Thus cause actual items very sensitive to the cost of parking stall measure video camera.
What general parking stall measure video camera was selected is the processor that arithmetic speed is the highest, and therefore parking stall measure video camera is carried out
The time of parking stall measure and Car license recognition is longer, the time that the time is about 5 seconds of general single parking stall measure, and carries out simultaneously
During multiple parking stall measure, required time can be multiplied.The time of Car license recognition is longer, the time meeting of general single Car license recognition
Reach more than 10 seconds.In actual use, parking stall measure and Car license recognition run simultaneously, due to task preemption, parking stall measure and
The time-consuming of Car license recognition all can increase accordingly, and owing to the priority of parking stall measure is high, can distribute more resource and carry out parking stall
Detection, therefore the single of the Car license recognition of low priority time-consumingly can increase to about 30 seconds.If it addition, increasing some extension inspections
Application exception parking situations such as () such as standdle carrier position warnings surveyed, can cause the process time of parking stall measure video camera to increase further
Add, the real-time of strong influence whole system and using effect, it is impossible to meet the use requirement of client.
In order to solve the problems referred to above, prior art is by using high performance process in parking stall measure video camera
Device, improves the processing speed of parking stall measure video camera, to meet the use requirement of user.Although the program solves parking stall measure
The processing speed of video camera is slow, the problem that parking stall measure and Car license recognition are time-consumingly grown, but the program also considerably increases parking stall inspection
Survey the cost of video camera.
Summary of the invention
It is an object of the invention to provide a kind of method for detecting parking stalls and device, can solve because of the place of parking stall measure video camera
Reason speed is slow, and parking stall measure and Car license recognition are the longest, and cause the parking stall measure time cannot meet user and use asking of requirement
Topic, will not increase again the cost of parking stall measure video camera.
To achieve these goals, technical solution of the present invention is as follows:
A kind of method for detecting parking stalls, is applied to parking stall measure video camera, and described method for detecting parking stalls includes parallel preliminary examination
Survey stage and accurate detection-phase, wherein said pre-detection stage includes:
Detect the parking space state change in the camera supervised region of this parking stall measure；
According to the testing result of parking space state change, when detecting that parking space state changes, search and accurately detect rank
Section duty is in the parking stall measure video camera of idle condition, is in the car of idle condition from accurate detection-phase duty
Position detection video camera chooses the parking stall measure video camera for performing accurate detection-phase, the car changed by parking space state
Position process information is sent to selected parking stall measure video camera；
Described accurate detection-phase includes:
Receive parking stall and process information, carry out parking space state and accurately detect and Car license recognition, and update the accurate detection of self
Stage duty is busy condition；
After parking space state accurately detects and terminates with Car license recognition, the accurate detection-phase duty updating self is sky
Not busy state.
Further, described being in the parking stall measure video camera of idle condition from accurate detection-phase duty is chosen
For performing the parking stall measure video camera of accurate detection-phase, including:
Inquire about whether this parking stall measure video camera accurate detection-phase duty is in idle condition；
If this parking stall measure video camera accurate detection-phase duty is in idle condition, then choose this parking stall measure
Video camera is as the parking stall measure video camera of the accurate detection-phase of execution, and judges whether the car that also parking space state changes
It is pending that position processes information, if had, continues whether inquiry other parking stall measure video cameras accurate detection-phase duty is located
In idle condition, choose other accurate detection-phase duties and be in the parking stall measure video camera of idle condition as performing essence
The really parking stall measure video camera of detection-phase；
If this parking stall measure video camera accurate detection-phase duty is in busy condition, then continue to inquire about other parking stalls
Whether detection video camera accurate detection-phase duty is in idle condition, chooses at other accurate detection-phase duties
In idle condition parking stall measure video camera as perform accurate detection-phase parking stall measure video camera.
Further, process information in described parking stall includes parking position monitoring image and parking stall configuration information, described parking position monitoring
Image is the encoded rear image generated of original image of parking stall measure video camera shooting, and parking position monitoring image comprises this parking stall
Detecting the image information of camera supervised all parking stalls, described parking stall configuration information is task list, this task list every
Article one, record corresponding parking stall numbering, this parking stall coordinate information in parking position monitoring image and this parking stall is carried out parking stall
State accurately detects No. ID of the parking stall measure video camera with Car license recognition.
Further, the described parking stall that changed by parking space state processes information and is sent to selected parking stall measure and takes the photograph
Multicast mode is used to be transmitted during camera.
Further, described in carry out parking space state and accurately detect and Car license recognition, including:
Accurately detect the parking stall with Car license recognition be divided into different detection regions by needing to carry out parking space state, and will be every
The parking space state in individual detection region accurately detects to be distributed to different parking stall measure video cameras with Car license recognition task and processes,
Then carry out collecting and obtain the parking space state of whole parking stall and accurately detect and license plate recognition result；
And/or, the License Plate Segmentation obtained by License Plate is become different regions, and by the character recognition of zones of different
Task is distributed to different parking stall measure video cameras and is carried out character recognition, then carries out collecting the character recognition obtaining whole car plate
Result.
The invention allows for a kind of stall testing apparatus, be applied to parking stall measure video camera, described stall testing apparatus
Including pre-detection module and accurate detection module, wherein said pre-detection module includes:
Detector unit, for detecting the parking space state change in the camera supervised region of this parking stall measure；
Query unit, for the testing result according to detector unit, when detecting that parking space state changes, searches essence
Really detection module duty is in the parking stall measure video camera of idle condition, is in the free time from accurate detection module duty
The parking stall measure video camera of state is chosen for performing the parking stall measure video camera accurately detected, parking space state is changed
Parking stall process information be sent to selected parking stall measure video camera；
Described accurate detection module includes:
Processing unit, is used for receiving parking stall and processes information, carry out parking space state and accurately detect and Car license recognition；
Updating block, for when carrying out parking space state and accurately detecting with Car license recognition, the duty updating self is
Busy condition, after parking space state accurately detects and terminates with Car license recognition, the duty updating self is idle condition.
Further,
Described query unit is being in the parking stall measure video camera of idle condition choosing from accurate detection module duty
Take in time performing the parking stall measure video camera of accurately detection, the following operation of execution:
Inquire about whether this parking stall measure video camera accurate detection module duty is in idle condition；
If this parking stall measure video camera accurate detection module duty is in idle condition, then choose this parking stall measure
Video camera is as performing the accurate parking stall measure video camera detected, and judges whether at the parking stall that also parking space state changes
Reason information is pending, if had, continues whether inquiry other parking stall measure video cameras accurate detection module duty is in sky
Not busy state, chooses other accurate detection module duties and is in the parking stall measure video camera of idle condition as performing accurately inspection
The parking stall measure video camera surveyed；
If this parking stall measure video camera accurate detection module duty is in busy condition, then continue to inquire about other parking stalls
Whether detection video camera accurate detection module duty is in idle condition, chooses at other accurate detection module duties
In the parking stall measure video camera that the parking stall measure video camera of idle condition accurately detects as execution.
Further, process information in described parking stall includes parking position monitoring image and parking stall configuration information, described parking position monitoring
Image is the encoded rear image generated of original image of parking stall measure video camera shooting, and parking position monitoring image comprises this parking stall
Detecting the image information of camera supervised all parking stalls, described parking stall configuration information is task list, this task list every
Article one, record corresponding parking stall numbering, this parking stall coordinate information in parking position monitoring image and this parking stall is carried out parking stall
State accurately detects No. ID of the parking stall measure video camera with Car license recognition.
Further, the parking stall that parking space state is changed by described query unit processes information and is sent to selected car
During the detection video camera of position, multicast mode is used to be transmitted.
Further, described processing unit carries out parking space state when accurately detecting with Car license recognition, performs following operation:
Accurately detect the parking stall with Car license recognition be divided into different detection regions by needing to carry out parking space state, and will be every
The parking space state in individual detection region accurately detects to be distributed to different parking stall measure video cameras with Car license recognition task and processes,
Then carry out collecting and obtain the parking space state of whole parking stall and accurately detect and license plate recognition result；
And/or, the License Plate Segmentation obtained by License Plate is become different regions, and by the character recognition of zones of different
Task is distributed to different parking stall measure video cameras and is carried out character recognition, then carries out collecting the character recognition obtaining whole car plate
Result.
The present invention proposes a kind of method for detecting parking stalls and device, by frequent for vehicles while passing region parking stall measure being imaged
The parking stall measure task of machine is distributed to other parking stall measure video camera being in idle condition and is processed, and makes vehicles while passing frequent
The parking stall measure time in region is greatly shortened, and in the case of not increasing parking stall measure video camera cost, solves because parking stall is examined
The processing speed surveying video camera is slow, and parking stall measure and Car license recognition are the longest, and cause the parking stall measure time cannot meet user
Use the problem required.Parking stall is processed information also by multicast packet and is sent to multiple parking stall measure video camera and enters by the present invention
Row processes, and saves network traffics, alleviates network load.
Accompanying drawing explanation
Fig. 1 is the flow chart of method for detecting parking stalls pre-detection stage of the present invention；
Fig. 2 is the flow chart of the accurate detection-phase of method for detecting parking stalls of the present invention；
Fig. 3 is the network diagram that the present embodiment multiple-camera fits into row parking stall measure；
Fig. 4 is the present embodiment stall testing apparatus structural representation.
Detailed description of the invention
Being described in further details technical solution of the present invention with embodiment below in conjunction with the accompanying drawings, following example are not constituted
Limitation of the invention.
The custom of stopping of people is all that the region from upper strata, parking lot, close to entrance starts to stop, and now parking lot is upper
Layer, have vehicle the most frequently close to the region of entrance and enter, and lower floor, the fragmentary vehicle in only local, region away from entrance
Come in and go out；When upper strata or after being fully parked with close to the region of entrance, may proceed to the region to lower floor, away from entrance and stop, lower floor, away from
The region of entrance has vehicle the most frequently and enters.I.e. being fully parked with the region of vehicle, the most only fragmentary vehicle in local is gone out, and car
More empty region, position has vehicle the most frequently and goes out.Therefore parking lot can often occur overall idle, the busiest feelings
Condition.
Can often occur overall idle for above-mentioned parking lot, the busiest situation, the present embodiment uses multiple stage parking stall
Detection video camera fits into row parking stall measure, the part parking stall measure task of the parking stall measure video camera of heavy duty zone is distributed
Detect to the parking stall measure video camera of clear area, to shorten the time of the parking stall measure of heavy duty zone.
Parking stall measure is divided into two stages, i.e. pre-detection stage and accurate detection-phase, same parking stall by the present embodiment
Detection video camera parallel processing pre-detection stage and accurate detection-phase, pre-detection stage periodically carries out pre-detection, simultaneously
The task of distributing accurate detection-phase according to testing result performs to local or other parking stall measure video cameras, accepts and processes essence
The really task result of detection-phase；Accurate detection-phase is responsible for performing parking space state and is accurately detected and Car license recognition.Easily
It is understood by, (as parking stall measure is taken the photograph in the equipment using central processing unit, microprocessor or programming device to process at present
Camera), different process threads can be used to process the task of different phase.Such as first thread processes pre-detection stage, the
Two threads process accurate detection-phase.The present embodiment first thread and the second thread can process parallel, and the second line
Task handled by journey can be given the parking stall measure video camera of this locality or other free time and be performed, when greatly reducing process
Between, improve the efficiency of process.
The present embodiment one method for detecting parking stalls, is applied to parking stall measure video camera, including parallel pre-detection stage and
Accurately detection-phase.Wherein pre-detection stage is as it is shown in figure 1, include:
Step S1, detect the camera supervised region of this parking stall measure parking space state change.
When the present embodiment carries out the installation of parking stall measure video camera in parking lot, first pass through reconnaissance at criminal scene, select the most single
Parking stall, double parking stall, the parking stall measure video camera model of three parking stalls；Then according to field condition, parking stall measure video camera is installed to
Suitably monitor position, make parking stall measure video camera cover all parking stalls specifying monitoring region；Exist finally by calibration technique
The monitored picture of parking stall measure video camera arranges the positional information of each parking stall.Parking stall measure video camera is according to the parking stall arranged
The all parking stalls specifying monitoring region are detected by positional information, it may be judged whether have car enter or leave parking stall, and to entrance
Or leave the vehicle in parking stall and carry out Car license recognition.Parking stall measure video camera coordinates display lamp, navigational panel, auxiliary terminal, management
Software, can realize forward parking guidance and reverse car seeking.
Parking space state pre-detection is completed by the parking stall measure video camera of region, parking stall, and parking space state accurately detects and car
Board identification can be completed by the parking stall measure video camera of region, parking stall, it is also possible to is taken the photograph by the parking stall measure of region, parking stall
Camera is distributed to other parking stall measure video camera and is completed.Carrying out parking space state accurately to detect and the parking stall inspection of Car license recognition
Survey video camera, be referred to as the parking stall measure video camera being in busy condition, do not carry out parking space state and accurately detect and Car license recognition
Parking stall measure video camera, is referred to as the parking stall measure video camera being in idle condition.The present embodiment is by managing server to all
The state of parking stall measure video camera is managed, when the state of every parking stall measure video camera changes, all by state
Situation of change is sent to manage server, record management server the state of real-time update every parking stall measure video camera.
The present embodiment parking space state pre-detection uses motion detection or other simple algorithm to realize, for parking space state being
No changing tentatively judges.For the parking stall measure scene of underground garage, as long as motion detection selects suitably detection
Region, detection threshold, it is possible to obtain a lowest pre-detection algorithm of loss.Due in the mainstream chip that video camera uses
The accelerating engine of capital integrated motion detection, the speed of the parking space state pre-detection therefore realized by motion detection algorithm is permissible
Reach the most real-time effect.
Step S2, the testing result changed according to parking space state, when detecting that parking space state changes, search accurately
Detection-phase duty is in the parking stall measure video camera of idle condition, is in idle shape from accurate detection-phase duty
The parking stall measure video camera of state is chosen the parking stall measure video camera for performing accurate detection-phase, parking space state is become
The parking stall changed processes information and is sent to selected parking stall measure video camera.
This parking stall measure video camera, according to the cycle time preset, is periodically supervised by the present embodiment parking stall measure video camera
All parking stalls in control region carry out parking space state pre-detection one by one, when detecting that parking space state changes and (i.e. has vehicle to enter
Enter or leave) time, then need to carry out the process of accurate detection-phase, do not carry out the process of accurate detection-phase.
In actual application, same parking lot can arrange multiple stage parking stall measure video camera, and management is the most corresponding respectively
Monitoring region.Therefore for the video monitoring system in a parking lot, it is provided with management server to manage all of parking stall
Detection video camera, the duty of record parking stall measure video camera, and provide inquiry service for parking stall measure video camera, receiving
After the inquiry of parking stall measure video camera, the parking stall measure video camera information of idle condition can be in its feedback.Certainly this pipe
It is same that reason server directly can also be substituted realization by a certain parking stall measure video camera in numerous parking stall measure video cameras
Function, repeat no more here.
The present embodiment, when detecting that parking space state changes, searches the parking stall measure video camera being in idle condition,
And therefrom choose and accurately detect the parking stall measure video camera with Car license recognition for performing parking space state.I.e. it is being in idle condition
Parking stall measure video camera in when choosing for performing parking stall measure video camera that parking space state accurately detects with Car license recognition, when
When this parking stall measure video camera is in idle condition, preferentially choose this parking stall measure video camera, other idle condition cars of reselection
Position detection video camera carries out parking space state and accurately detects and Car license recognition.Selection course is as follows:
Inquire about whether this parking stall measure video camera accurate detection-phase duty is in idle condition；
If this parking stall measure video camera accurate detection-phase duty is in idle condition, then choose this parking stall measure
Video camera is as the parking stall measure video camera of the accurate detection-phase of execution, and judges whether the car that also parking space state changes
It is pending that position processes information, if had, continues whether inquiry other parking stall measure video cameras accurate detection-phase duty is located
In idle condition, choose other accurate detection-phase duties and be in the parking stall measure video camera of idle condition as performing essence
The really parking stall measure video camera of detection-phase；
If this parking stall measure video camera accurate detection-phase duty is in busy condition, then continue to inquire about other parking stalls
Whether detection video camera accurate detection-phase duty is in idle condition, chooses at other accurate detection-phase duties
In idle condition parking stall measure video camera as perform accurate detection-phase parking stall measure video camera.
Specifically, first inquire about whether this parking stall measure video camera is in idle condition, if this parking stall measure video camera
It is in idle condition, and when the state only detecting a parking stall changes, then chooses this parking stall measure video camera and carry out
Parking space state accurately detects and Car license recognition.If this parking stall measure video camera is in busy condition, then according to state change occurs
Parking stall quantity, send inquiry request to management server, the letter of the idle parking stall measure video camera of acquisition request respective numbers
Breath；Management server is according to the state of the parking stall measure video camera recorded, from the parking stall measure video camera being in idle condition
In choose the parking stall measure video camera of respective numbers, and by the relevant information of selected parking stall measure video camera, including parking stall
The detection IP address of video camera and No. ID, feed back to the parking stall measure video camera sending inquiry request；Send the car of inquiry request
The parking stall measure video camera of the feedback that management server chosen by position detection video camera carries out parking space state and accurately detects and car plate knowledge
Not.If this parking stall measure video camera is in idle condition, and detects when the state of multiple parking stall changes, then by upper
Other stating that same method obtains that the parking stall quantity that quantity is the change of generation state subtracts one from management server is in idle shape
The information of the parking stall measure video camera of state, and choose this parking stall measure video camera and take the photograph from the parking stall measure managing server acquisition
Camera carries out parking space state and accurately detects and Car license recognition.
The accurate detection-phase of the present embodiment, as in figure 2 it is shown, include:
Step F1, reception parking stall process information, carry out parking space state and accurately detect and Car license recognition, and update the essence of self
Really detection-phase duty is busy condition；
Step F2, after parking space state accurately detects and Car license recognition terminates, update self accurate detection-phase work
State is idle condition.
The present embodiment accurately detects and after the choosing of parking stall measure video camera of Car license recognition completing to carry out parking space state,
The information that processed parking stall is sent to selected parking stall measure video camera and carries out parking space state and accurately detect and Car license recognition.Parking stall
Process information includes parking position monitoring image and parking stall configuration information.Parking position monitoring image is the original of parking stall measure video camera shooting
The image generated after image is encoded, such as jpeg image, it is camera supervised that parking position monitoring image comprises this parking stall measure
The image information of all parking stalls.Parking stall configuration information is a task list, corresponding one of each the record of this task list
Parking stall numbering, this parking stall coordinate information in parking position monitoring image and this parking stall is carried out parking space state accurately detect and car
No. ID of the parking stall measure video camera of board identification.
Such parking stall processes information and is convenient for multicast transmission, reduces data transmission quantity in a network.Easy to understand
, process information in parking stall can not also include task list, only this parking stall measure video camera with to carry out the car that processes
Carry out clean culture transmission between the detection video camera of position, but so can increase the transmission quantity in network, repeat no more here.
It is sent to multiple parking stall measure takes the photograph it should be noted that parking stall is processed information by the present embodiment parking stall measure video camera
When camera processes, in order to reduce data transmission quantity in a network, by multicast packet, parking stall can be processed information
It is sent to the parking stall measure video camera of multiple free time.Specifically, as it is shown on figure 3, parking stall measure video camera A receives management server
After the relevant information of idle parking stall measure video camera B and C sent, according to the IP ground of parking stall measure video camera B and C in this message
Location, is sent to parking stall measure video camera B and C by the multicast group address of this parking stall measure video camera A, notifies parking stall measure video camera
B and C adds this multicast group, and parking stall measure video camera A receives parking stall measure video camera B and C and adds multicast group and successfully feed back and disappear
After breath, the information that processed parking stall is sent to parking stall measure video camera B and C, parking stall measure video camera B and C by multicast packet
After receiving the parking stall process information that parking stall measure video camera A sends, exit the multicast group of parking stall measure video camera A.
The present embodiment carry out parking space state accurately detect the parking stall measure video camera with Car license recognition receive parking stall process
After information, according to parking stall therein configuration information and No. ID acquisition this parking stall measure video camera need of this parking stall measure video camera
Carry out the parking stall detected, and from parking position monitoring image, obtain the image information of this parking stall, this parking stall is carried out parking space state
Accurately detection and Car license recognition.The state of this equipment is set to busy condition by this parking stall measure video camera simultaneously, and reports pipe
Reason server.
Parking space state accurately detects and carries out car face detection exactly in the region, parking stall of parking position monitoring image, specifically detects
Cheng Wei: use algorithm for pattern recognition, the parking stall district respectively with different size of pattern match frame, in traversal parking position monitoring image
Territory, until finding car face or complete region, parking stall of traversal.If complete parking stall of traversal in parking space state accurately detects
Car face is not the most found in region, then judge that parking space state pre-detection occurs in that flase drop, stops next step Car license recognition task, will
The parking stall numbering of detected parking stall and testing result feed back to parking stall measure video camera or the management clothes sending parking stall configuration information
Business device, is then set to this parking stall measure video camera idle condition, and reports management server.If it is accurate at parking space state
Detection have found car face, then carry out next step Car license recognition task.
Car license recognition is divided into three steps: License Plate, License Plate Character Segmentation and character recognition.Initially with parking stall shape
State accurately detects similar method and carries out License Plate, i.e. uses algorithm for pattern recognition, uses different size of pattern match respectively
Frame, the region, parking stall of traversal parking position monitoring image, until finding car plate, then by the distribution characteristics of character in car plate, car plate is cut
It is slit into multiple different region, makes each region comprises a character, then the character in each region is identified, will
All characters identified combine in order and are the license plate number identified.After the present embodiment completes Car license recognition,
The parking stall measure that the license plate number numbered the parking stall of detected parking stall and identify feeds back to sending parking stall configuration information is taken the photograph
Camera or management server, be then set to this parking stall measure video camera idle condition, and report management server.
It should be noted that said method is to use a parking stall measure video camera accurately to detect, in order to enter one
Step shortening carries out parking space state and accurately detects the time with Car license recognition, and each needs can also be carried out parking stall shape by the present embodiment
State accurately detects the parking stall with Car license recognition and is divided into multiple detection region, and is accurately examined by the parking space state in each detection region
Survey and Car license recognition task is distributed to different parking stall measure video cameras and processed.In order to ensure that occurring without missing inspection surveys, multiple
Detection must have certain overlap between region.If there being car in picture, go-on-go surveys region method can be greatly shortened parking space state
Accurately detection and the time of Car license recognition, if not having car in picture, each detection region carries out what parking space state accurately detected
Time-consumingly can be more, but total time is also shorter than the method regardless of detection region, can contract further the most in this way
The time that short parking space state accurately detects.
It addition, character recognition can use a parking stall measure video camera to perform during Car license recognition, it is also possible to adopt
Process with distributed method, to shorten the time of character recognition.Specifically, License Plate Segmentation to be become different regions, and will
The character recognition task of zones of different is distributed to different parking stall measure video cameras and is carried out character recognition, then carries out collecting obtaining
The character identification result of whole car plate.
Owing to the parking stall measure of the present embodiment parking stall measure video camera is divided into pre-detection stage and accurate detection-phase, therefore
Can be processed by double business threads, double business threads include first thread and the second thread, and first thread is periodic
Perform pre-detection function, process line according to testing result distributed tasks to local or other parking stall measure video cameras tasks simultaneously
Journey, accepts and processes task result；Second thread is responsible for performing the task of accurate detection-phase, simultaneously that testing result is anti-
It is fed to this locality or the first thread of task initiating equipment.First thread processes with the second thread parallel, and first thread is periodically entered
Row pre-detection, when finding parking space state change, finds the idle parking stall measure video camera of the second thread and processes accurately detection
The task in stage.
The present embodiment passes through said method, when multiple cars that the parking stall measure video camera in the frequent region of vehicles while passing is monitored
When position has car turnover or this parking stall measure video camera to be in busy condition simultaneously, parking stall measure task is distributed to other and is in the free time
The parking stall measure video camera of state processes, and in the case of not increasing parking stall measure video camera cost, substantially reduces car
The time of position detection.
Illustrate below by citing, as shown in Figure 3:
By parking space state pre-detection, parking stall measure video camera A finds that surveillance area has the state of three parking stalls to become
Changing, it is idle condition that parking stall measure video camera A inquires about the state of this equipment, and sends inquiry request, request to management server
Obtain two parking stall measure video cameras being in idle condition, management server by the IP address of parking stall measure video camera B and C and
No. ID feeds back to parking stall measure video camera A, it is assumed that No. ID of parking stall measure video camera A, B, C is respectively 11,12,13.Parking stall is examined
Survey video camera A the original image of shooting in parking space state pre-detection is encoded, generate parking position monitoring image；Then to this car
Three parking stalls difference numbered 1,2,3 in the monitoring image of position, and generate the parking stall 1,2,3 seat in parking position monitoring image respectively
Mark information, and the parking stall measure task of numbered for parking stall 1,2,3 is respectively allocated to the parking stall measure that No. ID is 11,12,13 takes the photograph
Camera processes, and generates parking stall configuration information.Parking stall measure video camera A will comprise parking position monitoring image and parking stall configuration information
Parking stall process information be sent to this equipment and parking stall measure video camera B and C, parking stall measure video camera A is according to the car received
Position configuration information gets and needs to carry out parking space state and accurately detect and the parking stall number of Car license recognition is 1, and according to parking stall 1 at car
Coordinate information in the monitoring image of position obtains the image of parking stall 1 from parking position monitoring image, then according to the parking stall 1 got
Image carries out parking space state and accurately detects, and carrying out the time that parking space state accurately detects is 5 seconds, and accurate at parking space state
Detecting car face in detection, then proceed by Car license recognition, the time carrying out Car license recognition is 12 seconds, finally will identify
The license plate number gone out and parking stall numbering 1 feed back to parking stall measure video camera A.Parking stall measure video camera B is obtained by same method
The image of the parking stall 2 arrived, carries out parking space state and accurately detects, and carrying out the time that parking space state accurately detects is 7 seconds, and
Parking space state is not detected by car face in accurately detecting, the most do not carry out next step Car license recognition, and by parking space state preliminary examination
Testing result and the parking stall numbering 2 of measuring existing flase drop feed back to parking stall measure video camera A.Parking stall measure video camera C is by same
The image of parking stall 3 that gets of method, carry out parking space state and accurately detect, carrying out the time that parking space state accurately detects is 6
Second, and in parking space state accurately detects, detect car face, then carry out next step Car license recognition, carry out car plate knowledge
Other time is 10 seconds, and the license plate number that finally will identify that and parking stall numbering 3 feed back to parking stall measure video camera A.
By said method the parking space state of parking stall 1,2,3 accurately detected and Car license recognition task is respectively allocated to parking stall
Detection video camera A, B, C process, makes the parking space state of these three parking stall accurately detect and the time contracting of Car license recognition consuming
Short is 17 seconds, even and if use prior art not consider task preemption the most at least needs 40 seconds, therefore substantially reduce inspection
The survey time.
Corresponding to said method, the present embodiment one stall testing apparatus, it is applied to parking stall measure video camera.Such as Fig. 4 institute
Showing, this stall testing apparatus includes pre-detection module and accurate detection module, and wherein pre-detection module includes:
Detector unit, for detecting the parking space state change in the camera supervised region of this parking stall measure；
Query unit, for the testing result according to detector unit, when detecting that parking space state changes, searches essence
Really detection module duty is in the parking stall measure video camera of idle condition, is in the free time from accurate detection module duty
The parking stall measure video camera of state is chosen for performing the parking stall measure video camera accurately detected, parking space state is changed
Parking stall process information be sent to selected parking stall measure video camera；
Wherein accurately detection module includes:
Processing unit, is used for receiving parking stall and processes information, carry out parking space state and accurately detect and Car license recognition；
Updating block, for when carrying out parking space state and accurately detecting with Car license recognition, the duty updating self is
Busy condition, after parking space state accurately detects and terminates with Car license recognition, the duty updating self is idle condition.
Accordingly, query unit is being in the parking stall measure video camera of idle condition from accurate detection module duty
When choosing for performing the parking stall measure video camera accurately detected, perform to operate as follows:
Inquire about whether this parking stall measure video camera accurate detection module duty is in idle condition；
If this parking stall measure video camera accurate detection module duty is in idle condition, then choose this parking stall measure
Video camera is as performing the accurate parking stall measure video camera detected, and judges whether at the parking stall that also parking space state changes
Reason information is pending, if had, continues whether inquiry other parking stall measure video cameras accurate detection module duty is in sky
Not busy state, chooses other accurate detection module duties and is in the parking stall measure video camera of idle condition as performing accurately inspection
The parking stall measure video camera surveyed；
If this parking stall measure video camera accurate detection module duty is in busy condition, then continue to inquire about other parking stalls
Whether detection video camera accurate detection module duty is in idle condition, chooses at other accurate detection module duties
In the parking stall measure video camera that the parking stall measure video camera of idle condition accurately detects as execution.
Accordingly, the parking stall that parking space state is changed by query unit processes information and is sent to selected parking stall measure
During video camera, multicast mode is used to be transmitted.
Accordingly, processing unit carries out parking space state when accurately detecting with Car license recognition, performs following operation:
Accurately detect the parking stall with Car license recognition be divided into different detection regions by needing to carry out parking space state, and will be every
The parking space state in individual detection region accurately detects to be distributed to different parking stall measure video cameras with Car license recognition task and processes,
Then carry out collecting and obtain the parking space state of whole parking stall and accurately detect and license plate recognition result；
And/or, the License Plate Segmentation obtained by License Plate is become different regions, and by the character recognition of zones of different
Task is distributed to different parking stall measure video cameras and is carried out character recognition, then carries out collecting the character recognition obtaining whole car plate
Result.
The present embodiment pre-detection module and accurate detection module parallel processing, pre-detection module periodically carries out parking space state
Change pre-detection, when a change is detected, inquires about accurate detection module and is in the parking stall measure video camera of idle condition and carries out
Accurately detection.Thus substantially reduce the detection time, bring user preferably practical experience.
Above example is only limited in order to technical scheme to be described, without departing substantially from present invention essence
In the case of god and essence thereof, those of ordinary skill in the art are when making various corresponding change and change according to the present invention
Shape, but these change accordingly and deform the protection domain that all should belong to appended claims of the invention.
Claims (10)
1. a method for detecting parking stalls, is applied to parking stall measure video camera, it is characterised in that described method for detecting parking stalls includes also
The pre-detection stage of row and accurate detection-phase, wherein said pre-detection stage includes:
Detect the parking space state change in the camera supervised region of this parking stall measure；
According to the testing result of parking space state change, when detecting that parking space state changes, search accurate detection-phase work
It is in the parking stall measure video camera of idle condition as state, is in the parking stall inspection of idle condition from accurate detection-phase duty
Survey in video camera and choose the parking stall measure video camera for performing accurate detection-phase, at the parking stall changed by parking space state
Reason information is sent to selected parking stall measure video camera；
Described accurate detection-phase includes:
Receive parking stall and process information, carry out parking space state and accurately detect and Car license recognition, and update the accurate detection-phase of self
Duty is busy condition；
After parking space state accurately detects and terminates with Car license recognition, the accurate detection-phase duty updating self is idle shape
State.
Method for detecting parking stalls the most according to claim 1, it is characterised in that described at accurate detection-phase duty
The parking stall measure video camera for performing accurate detection-phase is chosen in the parking stall measure video camera of idle condition, including:
Inquire about whether this parking stall measure video camera accurate detection-phase duty is in idle condition；
If this parking stall measure video camera accurate detection-phase duty is in idle condition, then choose the shooting of this parking stall measure
Machine is as the parking stall measure video camera of the accurate detection-phase of execution, and judges whether at the parking stall that also parking space state changes
Reason information is pending, if had, continues whether inquiry other parking stall measure video cameras accurate detection-phase duty is in sky
Not busy state, chooses other accurate detection-phase duties and is in the parking stall measure video camera of idle condition as performing accurately inspection
The parking stall measure video camera in survey stage；
If this parking stall measure video camera accurate detection-phase duty is in busy condition, then continue to inquire about other parking stall measure
Whether video camera accurate detection-phase duty is in idle condition, chooses other accurate detection-phase duties and is in sky
The parking stall measure video camera of not busy state is as the parking stall measure video camera performing accurate detection-phase.
Method for detecting parking stalls the most according to claim 1, it is characterised in that described parking stall processes information and includes parking position monitoring
Image and parking stall configuration information, described parking position monitoring image is that the original image of parking stall measure video camera shooting is encoded rear raw
The image become, parking position monitoring image comprises the image information of the camera supervised all parking stalls of this parking stall measure, and described parking stall is joined
Confidence breath is task list, and each the record corresponding parking stall numbering of this task list, this parking stall are at parking position monitoring image
In coordinate information and this parking stall carried out parking space state accurately detect and No. ID of parking stall measure video camera of Car license recognition.
Method for detecting parking stalls the most according to claim 3, it is characterised in that the described parking stall that parking space state is changed
Multicast mode is used to be transmitted when process information is sent to selected parking stall measure video camera.
Method for detecting parking stalls the most according to claim 1, it is characterised in that described in carry out parking space state and accurately detect and car
Board identification, including:
Accurately detect the parking stall with Car license recognition be divided into different detection regions by needing to carry out parking space state, and by each inspection
Survey the parking space state in region accurately to detect and distribute to different parking stall measure video cameras with Car license recognition task and process, then
Carry out collecting and obtain the parking space state of whole parking stall and accurately detect and license plate recognition result；
And/or, the License Plate Segmentation obtained by License Plate is become different regions, and by the character recognition task of zones of different
Distribute to different parking stall measure video cameras and carry out character recognition, then carry out collecting the character recognition knot obtaining whole car plate
Really.
6. a stall testing apparatus, is applied to parking stall measure video camera, it is characterised in that described stall testing apparatus includes pre-
Detection module and accurate detection module, wherein said pre-detection module includes:
Detector unit, for detecting the parking space state change in the camera supervised region of this parking stall measure；
Query unit, for the testing result according to detector unit, when detecting that parking space state changes, searches accurately inspection
Survey module duty and be in the parking stall measure video camera of idle condition, be in idle condition from accurate detection module duty
Parking stall measure video camera in choose for performing the parking stall measure video camera accurately detected, the car changed by parking space state
Position process information is sent to selected parking stall measure video camera；
Described accurate detection module includes:
Processing unit, is used for receiving parking stall and processes information, carry out parking space state and accurately detect and Car license recognition；
Updating block, for when carrying out parking space state and accurately detecting with Car license recognition, the duty updating self is busy shape
State, after parking space state accurately detects and terminates with Car license recognition, the duty updating self is idle condition.
Stall testing apparatus the most according to claim 6, it is characterised in that described query unit is from accurate detection module
Duty is in when choosing for performing the parking stall measure video camera accurately detected in the parking stall measure video camera of idle condition,
Perform to operate as follows:
Inquire about whether this parking stall measure video camera accurate detection module duty is in idle condition；
If this parking stall measure video camera accurate detection module duty is in idle condition, then choose the shooting of this parking stall measure
Machine is as performing the accurate parking stall measure video camera detected, and judges whether that the parking stall that also parking space state changes processes letter
Ceasing pending, if had, continuing whether inquiry other parking stall measure video cameras accurate detection module duty is in idle shape
State, chooses other accurate detection module duties and is in the parking stall measure video camera of idle condition as performing accurately detection
Parking stall measure video camera；
If this parking stall measure video camera accurate detection module duty is in busy condition, then continue to inquire about other parking stall measure
Whether video camera accurate detection module duty is in idle condition, chooses other accurate detection module duties and is in sky
The parking stall measure video camera of not busy state is as the parking stall measure video camera performing accurately detection.
Stall testing apparatus the most according to claim 6, it is characterised in that described parking stall processes information and includes parking position monitoring
Image and parking stall configuration information, described parking position monitoring image is that the original image of parking stall measure video camera shooting is encoded rear raw
The image become, parking position monitoring image comprises the image information of the camera supervised all parking stalls of this parking stall measure, and described parking stall is joined
Confidence breath is task list, and each the record corresponding parking stall numbering of this task list, this parking stall are at parking position monitoring image
In coordinate information and this parking stall carried out parking space state accurately detect and No. ID of parking stall measure video camera of Car license recognition.
Stall testing apparatus the most according to claim 8, it is characterised in that parking space state is become by described query unit
When the parking stall process information changed is sent to selected parking stall measure video camera, multicast mode is used to be transmitted.
Stall testing apparatus the most according to claim 6, it is characterised in that described processing unit carries out parking space state essence
Really when detection and Car license recognition, the following operation of execution:
Accurately detect the parking stall with Car license recognition be divided into different detection regions by needing to carry out parking space state, and by each inspection
Survey the parking space state in region accurately to detect and distribute to different parking stall measure video cameras with Car license recognition task and process, then
Carry out collecting and obtain the parking space state of whole parking stall and accurately detect and license plate recognition result；
And/or, the License Plate Segmentation obtained by License Plate is become different regions, and by the character recognition task of zones of different
Distribute to different parking stall measure video cameras and carry out character recognition, then carry out collecting the character recognition knot obtaining whole car plate
Really.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610661972.4A CN106097763B (en) | 2016-08-12 | 2016-08-12 | A kind of method for detecting parking stalls and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610661972.4A CN106097763B (en) | 2016-08-12 | 2016-08-12 | A kind of method for detecting parking stalls and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106097763A true CN106097763A (en) | 2016-11-09 |
| CN106097763B CN106097763B (en) | 2019-03-08 |
Family
ID=57457000
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610661972.4A Active CN106097763B (en) | 2016-08-12 | 2016-08-12 | A kind of method for detecting parking stalls and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106097763B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN108491828A (en) * | 2018-04-20 | 2018-09-04 | 济南浪潮高新科技投资发展有限公司 | A kind of parking site detecting system and method for the pairwise similarity PVAnet based on level |
| CN108922235A (en) * | 2018-07-20 | 2018-11-30 | 西安艾润物联网技术服务有限责任公司 | Parking space state monitoring method, device and computer readable storage medium |
| CN114446077A (en) * | 2020-11-05 | 2022-05-06 | 上汽通用汽车有限公司 | Device and method for parking space detection, storage medium and vehicle |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101101648A (en) * | 2007-07-30 | 2008-01-09 | 深圳市融合视讯科技有限公司 | Parking lot booking system and its use method |
| CN101894482A (en) * | 2010-07-15 | 2010-11-24 | 东南大学 | Video technology-based roadside vacant parking position wireless network detection system and method |
| CN202677628U (en) * | 2012-04-16 | 2013-01-16 | 林汉溪 | Parking space searching, and management and control system |
| CN102915638A (en) * | 2012-10-07 | 2013-02-06 | 复旦大学 | Surveillance video-based intelligent parking lot management system |
| CN204178562U (en) * | 2014-11-10 | 2015-02-25 | 朱军 | The parking detector that a kind of removable video and wireless ultrasonic combine |
| CN104778855A (en) * | 2014-01-15 | 2015-07-15 | 北京同步科技有限公司 | Intelligent parking lot management system and management method |
| CN105225522A (en) * | 2015-10-09 | 2016-01-06 | 北京蓝卡科技股份有限公司 | Parking stall sense terminals, system and detection method |
-
2016
- 2016-08-12 CN CN201610661972.4A patent/CN106097763B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101101648A (en) * | 2007-07-30 | 2008-01-09 | 深圳市融合视讯科技有限公司 | Parking lot booking system and its use method |
| CN101894482A (en) * | 2010-07-15 | 2010-11-24 | 东南大学 | Video technology-based roadside vacant parking position wireless network detection system and method |
| CN202677628U (en) * | 2012-04-16 | 2013-01-16 | 林汉溪 | Parking space searching, and management and control system |
| CN102915638A (en) * | 2012-10-07 | 2013-02-06 | 复旦大学 | Surveillance video-based intelligent parking lot management system |
| CN104778855A (en) * | 2014-01-15 | 2015-07-15 | 北京同步科技有限公司 | Intelligent parking lot management system and management method |
| CN204178562U (en) * | 2014-11-10 | 2015-02-25 | 朱军 | The parking detector that a kind of removable video and wireless ultrasonic combine |
| CN105225522A (en) * | 2015-10-09 | 2016-01-06 | 北京蓝卡科技股份有限公司 | Parking stall sense terminals, system and detection method |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN108491828A (en) * | 2018-04-20 | 2018-09-04 | 济南浪潮高新科技投资发展有限公司 | A kind of parking site detecting system and method for the pairwise similarity PVAnet based on level |
| CN108922235A (en) * | 2018-07-20 | 2018-11-30 | 西安艾润物联网技术服务有限责任公司 | Parking space state monitoring method, device and computer readable storage medium |
| CN114446077A (en) * | 2020-11-05 | 2022-05-06 | 上汽通用汽车有限公司 | Device and method for parking space detection, storage medium and vehicle |
| CN114446077B (en) * | 2020-11-05 | 2023-10-03 | 上汽通用汽车有限公司 | Device and method for parking space detection, storage medium and vehicle |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106097763B (en) | 2019-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20220148303A1 (en) | Method and system for managing a parking lot based on intelligent imaging | |
| CN106097763A (en) | A kind of method for detecting parking stalls and device | |
| CN109410623A (en) | Parking navigation method and device | |
| US10401190B2 (en) | Vehicle control method and apparatus, and storage medium | |
| CN106355939B (en) | A kind of parking stall query-reservation method and system | |
| JP2020097393A (en) | Flight management system for flying object | |
| WO2016132587A1 (en) | Information processing device, road structure management system, and road structure management method | |
| CN104392628B (en) | A kind of parking lot car searching method and its system | |
| WO2017211136A1 (en) | Smart traffic management method and system, and computer storage medium | |
| CN105187790B (en) | Method, device and system for monitoring working state of vehicle-mounted terminal | |
| JPWO2015098442A1 (en) | Video search system and video search method | |
| US10587844B2 (en) | Image analysis system for analyzing dynamically allocated camera image, integrated control system including same, and operation method therefor | |
| US20100250309A1 (en) | Asset management system and method for an automotive vehicle | |
| CN103886755A (en) | Intersection abnormal parking rapid alarm system and method with red light running picture taking function | |
| CN110349406A (en) | A kind of congestion source origin cause of formation detection method and system | |
| CN107123301A (en) | Information of vehicles broadcasting system, method and apparatus | |
| CN111047231A (en) | Inventory method and system, computer system and computer readable storage medium | |
| WO2016132769A1 (en) | Imaging device, control method for imaging device, and control program | |
| CN112908014A (en) | Vehicle searching method and device for parking lot | |
| CN114023076B (en) | Specific vehicle tracking method based on multi-source heterogeneous data | |
| CN209168410U (en) | It is a kind of based on lamp draw control intelligence return seek vehicle guidance system | |
| WO2018205591A1 (en) | Target tracking method and target tracking apparatus | |
| CN110798618B (en) | Camera resource scheduling method and device in dynamic tracking | |
| CN106601009B (en) | A kind of parking information generation method and device | |
| CN110164180A (en) | A kind of parking lot management method and apparatus based on earth magnetism |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |