CN107329977A - A kind of false-trademark car postsearch screening method based on probability distribution - Google Patents
A kind of false-trademark car postsearch screening method based on probability distribution Download PDFInfo
- Publication number
- CN107329977A CN107329977A CN201710391814.6A CN201710391814A CN107329977A CN 107329977 A CN107329977 A CN 107329977A CN 201710391814 A CN201710391814 A CN 201710391814A CN 107329977 A CN107329977 A CN 107329977A
- Authority
- CN
- China
- Prior art keywords
- car
- bayonet socket
- false
- record
- probability
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2462—Approximate or statistical queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
Abstract
A kind of false-trademark car postsearch screening method based on probability distribution, comprises the following steps:S1. obtain bayonet socket and cross car record data, and carry out data cleansing and obtain bayonet socket crossing car record data;S2. the sequence of car record data is crossed to bayonet socket, vehicle is extracted and travels bayonet socket to vector;S3. the spatial probability distribution that vehicle is flowed between calculating bayonet socket;S4. bayonet socket record and database are contrasted, obtains the false car plate set of preliminary screening;S5. the car plate for meeting spatial probability distribution is obtained from the false car plate set of preliminary screening based on the spatial probability distribution that vehicle in S3 is flowed to;S6. the spatial probability distribution flowed to according to vehicle, determines each character-recognition errors probability;S7. the probability and character-recognition errors probability, synthetic determination number plate false-trademark probability of spatial distribution are met according to car plate.The present invention can overcome bayonet socket accuracy of identification not enough to a certain extent, effectively reduce false-trademark investigation scope, improve accuracy at target.
Description
Technical field
The invention belongs to intelligent transportation field, more particularly to a kind of false-trademark car postsearch screening method based on probability distribution.
Background technology
In recent years, continuing to develop with Chinese national economy, vehicle guaranteeding organic quantity constantly increases, and various traffic offences are disobeyed
Zhang Xianxiang also increasingly increases, wherein " false-trademark ", " deck " are with the illegal activities seriously endangered.Vehicle " false-trademark " phenomenon, refers to
Be vehicle forge, adulterium automotive number plate, it is illegal to use the non-existent car in motor vehicle management registers vehicle information
The phenomenon of the trade mark.
" false-trademark " can cause serious harm.Often exceeded the speed limit wantonly using the vehicle of false car plate, do not press traffic lights row
Sail, very disruptive traffic order.Once generation traffic accident, these drivers are under the driving of idea of leaving things to chance, and often selection is escaped
Ease, makes policeman in charge of the case be difficult to determine vehicle.Meanwhile, " false-trademark " car also tend to be offender tool used in crime, increase is broken
Case difficulty.Investigate and prosecute " false-trademark " vehicle, it has also become the vital task of various regions public security department and vehicle supervision department.
At present, " false-trademark " car is excavated the information and date storehouse mainly gathered by bayonet socket and contrasted, and is not present in database
Definition be " false-trademark " car, because bayonet socket number plate accuracy of identification is limited, the false-trademark car of preliminary screening often up to hundreds of thousands is needed
Carry out postsearch screening.From the point of view of the false-trademark car screening recognition methods that existing document and disclosed patent are proposed, vacation is related at present
Method for distinguishing main method is known in board screening can be divided into two classes:
(1) it is based on auxiliary equipment.Such as sides of the number of patent application CN201210187968.0 using reserved safety monitoring password
Formula.Vehicle safety detection code is reserved in traffic police's Internal Management System platform, vehicle is believed by handheld terminal in law enforcement traffic police scene
Breath and safety monitoring password are contrasted with reserved information, determine whether false-trademark car;Number of patent application CN201320577360.9 is adopted
With a kind of false license plate recognition device based on RFID technique, pass through the electronic tag for constituting radio frequency chip and microelectronic chip
Installed in vehicle body, judged using REID vehicle whether false-trademark deck.
(2) detection recognition method contrasted based on information of vehicles, such as number of patent application 201510744990.4 uses picture
Similarity identification.The SIFT feature of vehicle region in picture is extracted first, after clustering algorithm discretization, is converted into neighborhood special
Levy, based on vehicle Expressive Features, then using random forest method carry out similarity study, obtain similarity prediction mould
Type, for judging whether two vehicles belong to similar vehicle in picture.
There are some drawbacks in practical application in the above method:The first detection recognition method based on auxiliary equipment, is needed
To be installed to motor vehicle and be difficult to promote in extras, reality;Second of method based on the comparison of vehicle appearance information, light
Larger according to, ambient influnence, accuracy rate is not high.The drawbacks of in order to solve the above method, realize and fast and effeciently analyze extensive hand over
Logical data, from doubtful " false-trademark " vehicle of a large amount of primary dcreening operations, real " false-trademark " car of accurate lock is, it is necessary to a kind of new technical side
Case meets the demand of traffic control department.
The content of the invention
The present invention proposes one kind and effectively identification mistake and real " false-trademark " car can be made a distinction, and is substantially reduced
The investigation scope of " false-trademark " car, without extras, deployment is convenient, and applicability is wide, and recognition accuracy is higher, is greatly enhanced
The false-trademark car postsearch screening method based on probability distribution of follow up check and efficiency of deploying to ensure effective monitoring and control of illegal activities.
The technical solution adopted by the present invention is:
A kind of false-trademark car postsearch screening method based on probability distribution, comprises the following steps:
S1. obtain bayonet socket and cross car record data, and carry out data cleansing and obtain bayonet socket crossing car record data;
S2. car record data of being made a slip of the tongue to original cards sorts, and extracts vehicle traveling bayonet socket to vectorial (Ki, Kj), Ki and Kj tables
Show that bayonet socket is numbered, be put into together with HPHM in set K, HPHM represents number plate of vehicle;
S3. the Spatial Probability Pij that vehicle is flowed between calculating bayonet socket, and all probability (Ki, Kj, Pij) are stored in set P
In;
S4. car record data is crossed based on bayonet socket in S1 and obtains car plate set H, and pipe database is driven with car and compare preliminary screening
False car plate, obtains the false car plate set F1 of preliminary screening;
S5. the normal number of hops Jnor of each car in the spatial probability distribution set of computations F1 flowed to based on vehicle in S3
With abnormal number of hops Jp, and the car plate for meeting spatial probability distribution is put into set H1, does not meet spatial probability distribution
Car plate is put into set H2;
S6. Recognition of License Plate Characters error probability Lx is calculated based on character accounting in set H1 and set H2;
S7. the normal number of hops Jnor of each car and exception number of hops Jp and characters on license plate in set F1 is based on to know
Other error probability Lx is to car plate postsearch screening, synthetic determination car plate false-trademark probability.The spatial character that the present invention is travelled using vehicle,
The concept of probability distribution is proposed, by calculating the probability that vehicle is redirected each time, the continuity of vehicle spatially is judged.Such as
The continuity of fruit track of vehicle spatially is higher, illustrates that the track has larger possibility to be a car;If track of vehicle
Continuity spatially is relatively low, illustrates that the track has larger possibility to be many cars, that is to say, that the identification of the number plate
Accuracy is relatively low, and the car plate for not meeting spatial probability distribution is excluded by calculating.Simultaneously as tollgate devices are to different characters
Accuracy of identification is different, will spatially be more conform with the car plate of distribution probability and does not meet the car plate of probability distribution, is divided into two
Set, counts character accounting in two set, if obvious errors occurs in character accounting, illustrates the character recognition accuracy respectively
May be relatively low, can be by character recognition probability, the higher car plate of exclusive segment identification error rate again.
Further, to cross car record data acquisition methods as follows for step S1 bayonet socket:Original cards in a cycle are obtained to make a slip of the tongue
Car record data, and according to the data cleansing rule of setting, the data not being inconsistent normally are deleted, and retain the dimension of needs, wrap
Include bayonet socket numbering, brand number, spend the car time.
Further, the step of step S2 obtains set K is as follows:
(1) it is grouped, according to car time-sequencing is crossed in each group, is then grasped below each group of progress according to brand number
Make:
Step 1: taking out first record, record 1 is denoted as;
Step 2: taking out next record, record 2 is denoted as;
Step 3: calculating the time difference Δ T of record 1 and record 2;If time difference Δ T is less than threshold value T, step 4 is gone to;
If time difference Δ T is more than threshold value T, record 2 is assigned to record 1, step 2 is gone to;
Step 4: the bayonet socket numbering composition bayonet socket vector that number plate and two are recorded is put into set K to (HPHM, Ki, Kj)
In;Record 2 is assigned to record 1, step 2 is gone to;
(2) all groups are traveled through, set K is obtained.
Further, in step S3 calculate vehicle flow to Spatial Probability Pij the step of include:Each in statistics set K
The quantity of vectorial (Ki, Kj), is designated as cout (Ki, Kj), then bayonet socket Ki flows out vehicle summation and isCar
Flow to probability from bayonet socket Ki to bayonet socket Kj
Further, the car plate set H in step S4 crosses unduplicated car plate in car record data for bayonet socket in S1.
Further, preliminary screening is to there will be no car and drive car plate set in pipe database to form preliminary screening in step S4
False car plate set F1.
Further, the normal number of hops Jnor of each car and abnormal number of hops Jp are walked in set of computations F1 in step S5
Suddenly include:
(i) car plate in set F1, obtains each car plate corresponding all records in set K;
(ii) if the car plate does not have corresponding record in set K, number of times Jnor that the number plate is normally redirected and
Abnormal number of hops Jp is designated as 0;
(iii) if the car plate has corresponding record in set K, then obtained according to (Ki, the Kj) of each record
It is corresponding in set P to flow to probability P ij, if Pij is more than or equal to threshold value Pi, then it is normal to think that vehicle is this time redirected
, if Pij is less than threshold value Pi, then it is abnormal to think that vehicle is this time redirected;
(iv) the number of times Jnor that each car plate is normally redirected, i.e. Pij are counted>=Pi number of times and abnormal time redirected
Number Jp, i.e. Pij<Pi number of times.
Further, the step of Recognition of License Plate Characters error probability Lx is calculated in step S6 includes:Respectively statistics set H1 and
The accounting of each character is designated as Lx1 and Lx2 in set H2, and wherein x is represented may character, each character in calculating H2 set
Compared to the error Lx=ABS ((Lx2-Lx1)/Lx1) of accounting in H1.
Further, car plate postsearch screening formula is as follows in step S7:
FB numerical value is bigger, and the possibility for representing false-trademark is higher, otherwise
Recognize that the possibility of mistake is higher;ε is empirically worth, and typically takes cycle number of days.
The present invention be in order to overcome in a practical situation, because light, angle, number plate such as are stained at the factor, bayonet socket for number
The discrimination of board is unable to reach in the limitation of 100% (typically in 96%-98% or so), actual conditions, the very possible handle of bayonet socket
Some character recognition, normal Car license recognition into car plate not in car drives pipe database, cause primary dcreening operation into other characters
False-trademark car list is excessive, and artificial verification workload is big.
The present invention's is contemplated that:Next bayonet socket that vehicle passes through, it should meet spatially exponential probability distribution, if certain
The individual trade mark does not relatively meet spatial probability distribution, it is more likely that be simultaneously by two different Car license recognitions into same car
Board, that is, recognize mistake.Meanwhile, car plate is made up of different characters, and each character recognition probability is different, for by knowing
The car plate of the higher character composition of other probability is preferentially investigated, and the influence of identification mistake can be reduced as far as possible, so as to great
Artificial investigation scope is reduced, and improves false-trademark hit rate.
Beneficial effects of the present invention are mainly manifested in:Can preferably it overcome due to false-trademark car caused by bayonet socket identification mistake
Primary dcreening operation list is excessive, greatly reduces investigation scope, improves false-trademark hit rate, practicality good;Road network structure need not be relied on, is fitted
It is stronger with property.
Brief description of the drawings
Fig. 1 is flow chart of the invention.
The spatial probability distribution figure that Fig. 2 flows to for the vehicle of the present invention.
Embodiment
The present invention is further described with reference to specific embodiment, but does not limit the invention to these tools
Body embodiment.One skilled in the art would recognize that present invention encompasses potentially included in Claims scope
All alternatives, improvement project and equivalents.
A kind of reference picture 1, false-trademark car postsearch screening method based on probability distribution, comprises the following steps:
S1. obtain bayonet socket and cross car record data, and carry out data cleansing and obtain bayonet socket crossing car record data;
Bayonet socket is referred to using skills such as advanced photoelectricity, computer, image procossing, pattern-recognition, remote data access
Art, car lane, bicycle lane to monitoring section carry out round-the-clock real-time monitoring and record dependent image data, and automatically
Obtain the data such as passage time, place, travel direction, brand number, number plate color, the body color of vehicle.The car excessively of vehicle
Record can be stored in database with format data.
Obtain a cycle inner bayonet and cross car record data.In order to reduce the too small contingency brought of sample, the cycle can be with
Selecting must grow a bit, generally 1-6 months, and prioritizing selection is 3 months.
There is some dirty datas, including no license board information, car plate None- identified in original bayonet socket data, partial character can not
Identification etc., washes these dirty datas, and retain the dimension of needs, including bayonet socket numbering, excessively brand number, car time.
S2. the sequence of car record data is crossed to bayonet socket, vehicle traveling bayonet socket is extracted to vectorial (Ki, Kj), Ki and Kj represent card
Mouth numbering, is put into set K, HPHM represents number plate of vehicle together with HPHM;
Vehicle can be captured constantly in normal driving process by bayonet socket, and vehicle has higher probability and compared in theory
Neighbouring bayonet socket capture, the probability captured by more remote bayonet socket is lower.If a vehicle is often caught by the relatively low bayonet socket of probability
Obtain, illustrate that the vehicle less meets spatial probability distribution.100% is unable to reach in view of the accuracy of identification of bayonet socket, it is possible to led
The different vehicle on the way travelled is caused, same number plate is identified as, so as to cause vehicle not meet spatial probability distribution, anti-mistake
For, meet the car plate of spatial probability distribution, recognize that correct possibility is higher.
In reality, due to bayonet socket failure, network failure, the capture rate of bayonet socket is unable to reach the factor such as 100%, and vehicle exists
When by part bayonet socket, it is possible to will not be recorded.It is generally acknowledged that vehicle is from 1 bayonet socket, regular hour
Do not captured by any bayonet socket inside, it may be possible to there occurs shortage of data (it could also be possible that stationary vehicle), shortage of data has can
The bayonet socket of next capture vehicle can be caused less to meet spatial probability distribution.This time is referred to as threshold value T, if two, vehicle
Interval time between bayonet socket has exceeded threshold value T, and this group of bayonet socket is to being not involved in calculating.
Extract vehicle traveling bayonet socket as follows to the process of vector:
(1) data after S1 is cleaned, are grouped according to brand number, according to car time-sequencing is crossed in each group, so
Operated afterwards below each group of progress:
Step 1: taking out first record, record 1 is denoted as;
Step 2: taking out next record, record 2 is denoted as;
Step 3: calculating the time difference Δ T of record 1 and record 2;If time difference Δ T is less than threshold value T, step 4 is gone to;
If time difference Δ T is more than threshold value T, record 2 is assigned to record 1, step 2 is gone to;
Step 4: the bayonet socket numbering composition bayonet socket vector that number plate and two are recorded is put into set K to (HPHM, Ki, Kj)
In;Record 2 is assigned to record 1, step 2 is gone to;
(2) all groups are traveled through, set K is obtained.
S3. the Spatial Probability Pij that vehicle is flowed between calculating bayonet socket, and all probability (Ki, Kj, Pij) are stored in set P
In;
Vehicle is calculated from a bayonet socket according to set K, the probability of other each bayonet sockets is reached, this probability is claimed
To flow to probability between bayonet socket.Flow to probability and reflect the next bayonet socket of vehicle spatially probability distribution.Bayonet socket flows to probability
(Ki, Kj)=(vehicle number that bayonet socket Kj is reached from bayonet socket Ki)/from bayonet socket Ki vehicle fleet.It is every in statistics set K
The quantity of one vector (Ki, Kj), is designated as cout (Ki, Kj), then bayonet socket Ki flows out vehicle summation, is
Bayonet socket Ki flows to probability to bayonet socket Kj'sCalculate between all bayonet sockets pair
Flow to probability, if the current record number between two bayonet sockets is zero, then current probability is designated as 0%.
S4. car record data is crossed based on bayonet socket in S1 and obtains car plate set H, and pipe database is driven with car and compare preliminary screening
False car plate, obtains the false car plate set F1 of preliminary screening;
Specifically, according to car record data is crossed in S1, obtaining unduplicated car plate, obtaining the collection of all car plates in the cycle
Close H.Car plate in set H is driven into being compared in pipe database with car, if car plate is not in database, set F1 is put into
In, F1 is the false-trademark set of preliminary screening.
S5. the normal number of hops Jnor of each car in the spatial probability distribution set of computations F1 flowed to based on vehicle in S3
With abnormal number of hops Jp, and the car plate for meeting spatial probability distribution is put into set H1, does not meet spatial probability distribution
Car plate is put into set H2;Specific steps include:
(i) car plate in set F1, obtains each car plate corresponding all records in set K;
(ii) if the car plate does not have corresponding record in set K, number of times Jnor that the number plate is normally redirected and
Abnormal number of hops Jp is designated as 0;
(iii) if the car plate has corresponding record in set K, then obtained according to (Ki, the Kj) of each record
It is corresponding in set P to flow to probability P ij, if Pij is more than or equal to threshold value Pi, then it is normal to think that vehicle is this time redirected
, if Pij is less than threshold value Pi, then it is abnormal to think that vehicle is this time redirected;Threshold value Pi values are 0.2%.
(iv) the number of times Jnor that each car plate is normally redirected, i.e. Pij are counted>=Pi number of times and abnormal time redirected
Number Jp, i.e. Pij<Pi number of times.
If vehicle, which is redirected, does not meet spatial probability distribution, illustrate that the car plate has larger possibility to be identification mistake, anti-mistake
Come, meet probability distribution, illustrate that the Car license recognition correctness is higher.
S6. Recognition of License Plate Characters error probability Lx is calculated based on character accounting in set H1 and set H2;
When sample is sufficiently large, the frequency that each character of car plate occurs should tend to a stationary value, if some character goes out
Existing frequency is higher, illustrate other characters misidentify into the character possibility it is higher, in turn, if some character occur
Flat rate than relatively low, illustrate that the character has larger possibility to be identified as other characters.
By the element in set F1, it is divided into two set H1 and H2 according to probability is redirected, wherein set H1 is to redirect probability
Element more than or equal to 0.2%, set H1 is to redirect the general element less than 0.2%.Due to the car plate in set H1, it is more conform with
Spatial probability distribution, therefore, the character recognition accuracy in set H1 are higher, conversely, character recognition probability is relatively low in H2.Respectively
The accounting of each character is designated as Lx1 and Lx2 in statistics set H1 and set H2, and wherein x is represented may character, calculating H2 set
In each character compared to accounting in H1 error Lx=ABS ((Lx2-Lx1)/Lx1).Lx can be approximately be used for estimate every
A kind of probability of character-recognition errors.
S7. the normal number of hops Jnor of each car and exception number of hops Jp and characters on license plate in set F1 is based on to know
Other error probability Lx is to car plate postsearch screening, synthetic determination car plate false-trademark probability.Whether Spatial Probability is met according to vehicle flow direction
Distribution, can judge that two different car plates, either with or without same car plate is identified as, are not met by removing to a certain extent
The car plate of spatial probability distribution, can remove the car plate of this part identification mistake.In remaining car plate, different car plates are by difference
Character composition, the successful probability of each character recognition is different, for the car plate being made up of the higher character of identification probability, such as
Fruit is not in car drives pipe data, and the possibility of false-trademark is very high, can preferentially be investigated.
Finally can be according to formulaFB numerical value is bigger, represents false-trademark
Possibility is higher, otherwise recognizes that the possibility of mistake is higher.ε is empirically worth, and typically takes cycle number of days.
The spatial character that the present invention is travelled using vehicle, it is proposed that the concept of probability distribution, by calculating vehicle each time
The probability redirected, judges the continuity of vehicle spatially.If the continuity of track of vehicle spatially is higher, illustrate the rail
Mark has larger possibility to be a car;If the continuity of track of vehicle spatially is relatively low, illustrate that the track has larger
Possibility is many cars, that is to say, that the recognition correct rate of the number plate is relatively low, and Spatial Probability point is not met by calculating exclusion
The car plate of cloth.Simultaneously as tollgate devices are different to different character recognition precision, distribution probability will be spatially more conform with
Car plate and do not meet the car plate of probability distribution, be divided into two set, count respectively two gather in character accounting, if character
There are obvious errors in accounting, illustrates that the character recognition accuracy may be relatively low, can be by character recognition probability, exclusion portion again
The car plate for dividing identification error rate higher.
A kind of concrete application embodiment is as follows:
S1. bayonet socket crosses the extraction of car data:
Obtain a cycle inner bayonet and cross car record data, retain the dimension needed, including bayonet socket numbering, brand number mistake
The car time.
The present embodiment has extracted Hangzhou 1 day-January 30 January in 2016, adds up data on the 30th, altogether comprising 489 cards
Mouthful, 129534497 record altogether, bayonet socket data format such as table 1 below:
Table 1
Field | Data type | Implication |
KKID | VARchar(20) | Bayonet socket ID |
HPHM | VARchar(10) | Brand number |
HPLX | VARchar(2) | Number plate species |
JGSJ | VARchar(20) | Spend the car time |
One road section of one of KKID correspondences, HPHM+HPZL uniquely determines an automobile.JGSJ is accurate to the second,
(in following steps, brand number contains number plate species, repeats no more)
The cleaning of bayonet socket data:
Due to brand number be bayonet system according to picture recognition, number plate discrimination is unable to reach 100%, original bayonet socket number
It is sky according to there are some dirty datas, including car plate, it is impossible to recognize, partial character None- identified etc..Clean the partial data, portion
Division is for example shown in table 2 below:
Table 2
Sequence number | Brand number | Spend the car time |
1 | 2016-01-15 14:52:51 | |
2 | NULL | 2016-01-20 19:32:30 |
3 | Peaceful B711T | 2016-01-25 11:31:34 |
4 | Zhejiang A00NT | 2016-01-25 20:54:04 |
5 | Zhejiang A025X | 2016-01-21 14:18:13 |
6 | None- identified | 2016-01-10 22:49:28 |
S2. car record ordering is crossed, and extracts bayonet socket vector
Cross car record ordering:According to brand number, the car time is spent, data are ranked up.Partial data is as shown in table 3 below
(ellipsis part is non-display portion).
Table 3
Sequence number | Brand number | Bayonet socket ID | Spend the car time |
1 | Zhejiang A2M1** | 31000300007402 | 2016-01-04 07:51:09 |
2 | Zhejiang A2M1** | 31000300010702 | 2016-01-04 08:48:26 |
3 | Zhejiang A2M1** | 31000300010904 | 2016-01-04 08:50:13 |
4 | Zhejiang A2M1** | 31000300004504 | 2016-01-04 08:50:38 |
5 | Zhejiang A2M1** | 31000300004502 | 2016-01-04 08:50:58 |
6 | Zhejiang A2M1** | 31000300019902 | 2016-01-04 08:53:36 |
7 | Zhejiang A2M1** | 31000300005402 | 2016-01-04 08:59:18 |
··· | ······ | ······ | ······ |
To sorted record, satisfactory bayonet socket is taken out to vector.In the present embodiment, threshold value T is set to 15 points
Clock.By taking table 3 as an example, the process for taking out bayonet socket pair is as follows:
1st, record 1, record 2 are taken out;
2nd, record 1 and 2 time differences of record are calculated, is 57mins17s>15mins, gives up record 1;
3rd, record 3 is taken out, the time difference for calculating record 2 and record 3 is 1mins47s<15mins, will (Zhejiang A2M1**,
31000300010702,31000300010904) it is put into set K.
4th, a record is removed, repeats to operate above.
The above 7 crosses car record, can take out 5 bayonet sockets pair.
S3. probability is flowed between calculating bayonet socket
All (K in statistics set Ki,Kj), it can obtain from bayonet socket KiOutflow, flows to bayonet socket KjVehicle number.Statistics
count(Ki,Kj), it can obtain from KiThe vehicle fleet of outflow, obtain it is as shown in table 4 below (ellipsis part be non-display part
Point).
Table 4
Bayonet socket Ki | Bayonet socket Kj | count(Ki,Kj) | count(Ki) | Probability |
31000300000102 | 31000300001804 | 35433 | 156351 | 22.7% |
31000300000102 | 31000300012619 | 35384 | 156351 | 22.6% |
31000300000102 | 31000300001802 | 26530 | 156351 | 17.0% |
31000300000102 | 31000300027001 | 18117 | 156351 | 11.6% |
31000300000102 | 31000300009719 | 10139 | 156351 | 6.5% |
31000300000102 | 31000300000502 | 5298 | 156351 | 3.4% |
31000300000102 | 31000300000904 | 4236 | 156351 | 2.7% |
31000300000102 | 31000300000503 | 3885 | 156351 | 2.5% |
31000300000102 | 31000300002504 | 2150 | 156351 | 1.4% |
31000300000102 | 31000300000504 | 1190 | 156351 | 0.8% |
31000300000102 | 31000300000902 | 1180 | 156351 | 0.8% |
31000300000102 | 31000300025819 | 962 | 156351 | 0.6% |
31000300000102 | 31000300005002 | 820 | 156351 | 0.5% |
31000300000102 | 31000300012120 | 810 | 156351 | 0.5% |
······ | ······ | ······ | ······ | ··· |
······ | ······ | ······ | ······ | ··· |
Bayonet socket flows to probability and bayonet socket distribution and road network structure is embodied in another dimension.
Bayonet socket 31000300004304 is calculated to the probability that flows to of other bayonet sockets, and probability flashback is arranged, folding is drawn
Line chart, probability is in obvious exponential distribution.Stream is equally calculated to bayonet socket 31000300003801 and bayonet socket 31000300006604
To probability and curve map is drawn, probability is also into obvious exponential distribution.Three bayonet sockets flow to the scatter chart of probability, such as Fig. 2
It is shown.Wherein Y-axis represents probability, and X-axis represents other bayonet sockets (according to probability inverted order).
S4. bayonet socket record drives the contrast of pipe database with car, primarily determines that false-trademark car scope:
In the present embodiment, car drive pipe data only comprising " Zhejiang A " beginning related data, non-Zhejiang A number plates can not judge be
No is false-trademark, therefore the delineation of false-trademark scope is " Zhejiang A " number plate.Record unduplicated using car is crossed in MapReduce acquisitions S1
Car plate, only retains so that " these number plates are driven pipe data with car and contrasted, driven if being not included in car by the number plate of Zhejiang A " beginnings
In pipe database, it is put into set F1, F1 is the false-trademark car list of primary dcreening operation.
In the present embodiment, 235642 number plates are had for the doubtful false-trademark of primary dcreening operation.
S5. each normal number of hops of car and abnormal number of hops in set of computations F1.
According to the car plate in set F1, each car plate corresponding all records in set K are obtained.According to each note
It is corresponding in (Ki, Kj) acquisition set P of record to flow to probability P ij.Partial results are as follows:
Table 5
Count the number of times Jnor that each number plate is normally redirected, i.e. Pij>=Pi number of times and the abnormal number of times redirected
Jp, i.e. Pij<Pi number of times.If the number plate does not have corresponding record in set K, the number of times that the number plate is normally redirected
Jnor and exception number of hops Jp are designated as 0.Partial results such as table 6 below:
Table 6
Sequence number | Brand number | Normal number of hops | Abnormal number of hops |
1 | Zhejiang AA59** | 293 | 7 |
2 | Zhejiang A925** | 187 | 0 |
3 | Zhejiang A2EM** | 371 | 6 |
4 | Zhejiang A2KA** | 270 | 2 |
5 | Zhejiang A255** | 167 | 4 |
6 | Zhejiang AK5X** | 66 | 0 |
7 | Zhejiang AC29** | 164 | 4 |
8 | Zhejiang A9EN** | 259 | 4 |
9 | Zhejiang A295** | 458 | 0 |
10 | Zhejiang AH52** | 258 | 3 |
······ | ······ | ······ | ······ |
······ | ······ | ······ | ······ |
S6. Recognition of License Plate Characters error probability is calculated.
We are by the element in set F1, according to probability is redirected, and are divided into two set H1 and H2, and wherein H1 is included
66460616 elements, H2 includes 23970273 elements.Car plate is made up of 7 characters, and wherein front two represents place, rear five
Position represents car plate.In the present embodiment, front two with " based on the A " of Zhejiang, thus we mainly consider after 5 characters on license plate.
5 character accountings after car plate in H1 and H2 set are counted respectively, following form is obtained:
Table 7
We have seen that, in set H1 and set H2,3,5, Q, U these character accountings relatively, recognize error probability compared with
Small, these character accounting difference of T, X, N are larger, and identification error probability is higher.
S7. postsearch screening, the sequence of false-trademark possibility:
False-trademark possibility FB can be calculated by below equation.
ε values are 15 in the present embodiment.
Partial results are as follows.
Table 8
In the present embodiment, from more than 20 ten thousand doubtful deck, the higher number plate (FB of 1895 deck possibilities is filtered out
>0), the range shorter of screening more than 100 times.By actual verification, sorted if only according to " doubtful false-trademark " occurrence number,
In preceding 50 doubtful false-trademarks, only 4 are defined as false-trademark, and remaining is identification mistake, is sorted according to this method, preceding 50 doubtful vacations
In board, there are 24 to be defined as false-trademark, accuracy rate improves 6 times.
Claims (9)
1. a kind of false-trademark car postsearch screening method based on probability distribution, comprises the following steps:
S1. original cards are obtained to make a slip of the tongue car record data, and carry out data cleansing and obtain bayonet socket crossing car record data;
S2. the sequence of car record data is crossed to bayonet socket, vehicle traveling bayonet socket is extracted and bayonet socket, which is compiled, to be represented to vectorial (Ki, Kj), Ki and Kj
Number, it is put into together with HPHM in set K, HPHM represents number plate of vehicle;
S3. the Spatial Probability Pij that vehicle is flowed between calculating bayonet socket, and all probability (Ki, Kj, Pij) are stored in set P;
S4. car record data is crossed based on bayonet socket in S1 and obtains car plate set H, and pipe database is driven with car and compare the false car of preliminary screening
Board, obtains the false car plate set F1 of preliminary screening;
S5. the normal number of hops Jnor of each car and different in the spatial probability distribution set of computations F1 flowed to based on vehicle in S3
Normal number of hops Jp, and the car plate for meeting spatial probability distribution is put into set H1, do not meet the car plate of spatial probability distribution
It is put into set H2;
S6. Recognition of License Plate Characters error probability Lx is calculated based on character accounting in set H1 and set H2;
S7. the normal number of hops Jnor of each car and exception number of hops Jp and Recognition of License Plate Characters in set F1 are based on wrong
Probability Lx is to car plate postsearch screening, synthetic determination car plate false-trademark probability by mistake.
2. a kind of false-trademark car postsearch screening method based on probability distribution according to claim 1, it is characterised in that:Step
It is as follows that S1 bayonet socket crosses car record data acquisition methods:Original cards in a cycle are obtained to make a slip of the tongue car record data, and according to setting
Fixed data cleansing rule, deletes the data not being inconsistent normally, and retains the dimension of needs, including bayonet socket numbering, brand number,
Spend the car time.
3. a kind of false-trademark car postsearch screening method based on probability distribution according to claim 1, it is characterised in that:Step
The step of S2 obtains set K is as follows:
(1) it is grouped, according to car time-sequencing is crossed in each group, is then operated below each group of progress according to brand number:
Step 1: taking out first record, record 1 is denoted as;
Step 2: taking out next record, record 2 is denoted as;
Step 3: calculating the time difference Δ T of record 1 and record 2;If time difference Δ T is less than threshold value T, step 4 is gone to;If
Time difference Δ T is more than threshold value T, and record 2 is assigned into record 1, step 2 is gone to;
Step 4: the bayonet socket numbering composition bayonet socket vector that number plate and two are recorded is put into set K to (HPHM, Ki, Kj);
Record 2 is assigned to record 1, step 2 is gone to;
(2) all groups are traveled through, set K is obtained.
4. a kind of false-trademark car postsearch screening method based on probability distribution according to claim 1, it is characterised in that:Step
In S3 calculate vehicle flow to Spatial Probability Pij the step of include:The quantity of each in statistics set K vectorial (Ki, Kj), note
For cout (Ki, Kj), then bayonet socket Ki flows out vehicle summation and isVehicle is from bayonet socket Ki to bayonet socket Kj's
Flow to probability
5. a kind of false-trademark car postsearch screening method based on probability distribution according to claim 1, it is characterised in that:Step
Car plate set H in S4 crosses unduplicated car plate in car record data for bayonet socket in S1.
6. a kind of false-trademark car postsearch screening method based on probability distribution according to claim 5, it is characterised in that:Step
Preliminary screening is to there will be no car and drive car plate set in pipe database to form the false car plate set F1 of preliminary screening in S4.
7. a kind of false-trademark car postsearch screening method based on probability distribution according to one of claim 1~6, its feature exists
In:The normal number of hops Jnor of each car and abnormal number of hops Jp steps include in set of computations F1 in step S5:
(i) car plate in set F1, obtains each car plate corresponding all records in set K;
(ii) if the car plate does not have corresponding record in set K, the number of times Jnor and exception that the number plate is normally redirected
Number of hops Jp is designated as 0;
(iii) if the car plate has corresponding record in set K, then obtain set P according to (Ki, the Kj) of each record
In it is corresponding flow to probability P ij, if Pij be more than or equal to threshold value Pi, then it is normal to think that vehicle is this time redirected, if
Pij is less than threshold value Pi, then it is abnormal to think that vehicle is this time redirected;
(iv) the number of times Jnor that each car plate is normally redirected, i.e. Pij are counted>=Pi number of times and the abnormal number of times redirected
Jp, i.e. Pij<Pi number of times.
8. a kind of false-trademark car postsearch screening method based on probability distribution according to claim 7, it is characterised in that:Step
The step of Recognition of License Plate Characters error probability Lx is calculated in S6 includes:Each character in difference statistics set H1 and set H2
Accounting is designated as Lx1 and Lx2, and wherein x is represented may character, error of each character compared to accounting in H1 in calculating H2 set
Lx=ABS ((Lx2-Lx1)/Lx1).
9. a kind of false-trademark car postsearch screening method based on probability distribution according to claim 8, it is characterised in that:Step
Car plate postsearch screening formula is as follows in S7:
FB numerical value is bigger, and the possibility for representing false-trademark is higher, otherwise recognizes
The possibility of mistake is higher;ε is empirically worth, and typically takes cycle number of days.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710391814.6A CN107329977B (en) | 2017-05-27 | 2017-05-27 | A kind of false-trademark vehicle postsearch screening method based on probability distribution |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710391814.6A CN107329977B (en) | 2017-05-27 | 2017-05-27 | A kind of false-trademark vehicle postsearch screening method based on probability distribution |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107329977A true CN107329977A (en) | 2017-11-07 |
CN107329977B CN107329977B (en) | 2019-08-16 |
Family
ID=60193242
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710391814.6A Active CN107329977B (en) | 2017-05-27 | 2017-05-27 | A kind of false-trademark vehicle postsearch screening method based on probability distribution |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107329977B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109087508A (en) * | 2018-08-30 | 2018-12-25 | 广州市市政工程设计研究总院有限公司 | Contiguous zone traffic analysis method and system based on high definition bayonet data |
CN109448363A (en) * | 2018-09-30 | 2019-03-08 | 佳都新太科技股份有限公司 | Suspected vehicles based on trajectory predictions intelligently sealed method, system and processing terminal |
CN110164138A (en) * | 2019-05-17 | 2019-08-23 | 湖南科创信息技术股份有限公司 | Based on bayonet to the recognition methods for the fake license plate vehicle for flowing to probability and system, medium |
CN112614347A (en) * | 2020-12-22 | 2021-04-06 | 杭州海康威视系统技术有限公司 | Fake plate detection method and device, computer equipment and storage medium |
CN113011895A (en) * | 2021-03-31 | 2021-06-22 | 腾讯科技(深圳)有限公司 | Associated account sample screening method, device and equipment and computer storage medium |
CN116547702A (en) * | 2020-10-29 | 2023-08-04 | 日本制铁株式会社 | Identification device, identification method, and program |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2013214143A (en) * | 2012-03-30 | 2013-10-17 | Fujitsu Ltd | Vehicle abnormality management device, vehicle abnormality management system, vehicle abnormality management method, and program |
CN105702047A (en) * | 2016-03-04 | 2016-06-22 | 浙江宇视科技有限公司 | License plate identification error filtering method and apparatus in fake-license plate analysis |
CN105719489A (en) * | 2016-03-24 | 2016-06-29 | 银江股份有限公司 | Fake-licensed vehicle detection method based on bayonet vehicle flow direction probability |
CN106022296A (en) * | 2016-06-01 | 2016-10-12 | 银江股份有限公司 | Fake plate vehicle detection method based on vehicle hot spot area probability aggregation |
-
2017
- 2017-05-27 CN CN201710391814.6A patent/CN107329977B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2013214143A (en) * | 2012-03-30 | 2013-10-17 | Fujitsu Ltd | Vehicle abnormality management device, vehicle abnormality management system, vehicle abnormality management method, and program |
CN105702047A (en) * | 2016-03-04 | 2016-06-22 | 浙江宇视科技有限公司 | License plate identification error filtering method and apparatus in fake-license plate analysis |
CN105719489A (en) * | 2016-03-24 | 2016-06-29 | 银江股份有限公司 | Fake-licensed vehicle detection method based on bayonet vehicle flow direction probability |
CN106022296A (en) * | 2016-06-01 | 2016-10-12 | 银江股份有限公司 | Fake plate vehicle detection method based on vehicle hot spot area probability aggregation |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109087508A (en) * | 2018-08-30 | 2018-12-25 | 广州市市政工程设计研究总院有限公司 | Contiguous zone traffic analysis method and system based on high definition bayonet data |
CN109087508B (en) * | 2018-08-30 | 2021-09-21 | 广州市市政工程设计研究总院有限公司 | High-definition bayonet data-based adjacent area traffic volume analysis method and system |
CN109448363A (en) * | 2018-09-30 | 2019-03-08 | 佳都新太科技股份有限公司 | Suspected vehicles based on trajectory predictions intelligently sealed method, system and processing terminal |
CN110164138A (en) * | 2019-05-17 | 2019-08-23 | 湖南科创信息技术股份有限公司 | Based on bayonet to the recognition methods for the fake license plate vehicle for flowing to probability and system, medium |
CN116547702A (en) * | 2020-10-29 | 2023-08-04 | 日本制铁株式会社 | Identification device, identification method, and program |
CN112614347A (en) * | 2020-12-22 | 2021-04-06 | 杭州海康威视系统技术有限公司 | Fake plate detection method and device, computer equipment and storage medium |
CN113011895A (en) * | 2021-03-31 | 2021-06-22 | 腾讯科技(深圳)有限公司 | Associated account sample screening method, device and equipment and computer storage medium |
CN113011895B (en) * | 2021-03-31 | 2023-07-18 | 腾讯科技(深圳)有限公司 | Associated account sample screening method, device and equipment and computer storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN107329977B (en) | 2019-08-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107329977B (en) | A kind of false-trademark vehicle postsearch screening method based on probability distribution | |
CN104200669B (en) | Fake-licensed car recognition method and system based on Hadoop | |
CN106022296B (en) | A kind of fake-licensed car detection method of the probability polymerization based on vehicle hot spot region | |
CN106251635A (en) | The recognition methods of a kind of deck suspicion license plate number and system | |
CN101587643B (en) | Identification method of fake-licensed cars | |
CN105513368B (en) | A kind of false-trademark car screening technique based on uncertain information | |
CN105719489B (en) | A kind of fake-licensed car detection method that probability is flowed to based on bayonet vehicle | |
CN105575118B (en) | One kind is lost and drives personnel's screening technique | |
CN103136528B (en) | A kind of licence plate recognition method based on dual edge detection | |
CN105118305B (en) | Motor pool outlet vehicle management platform | |
CN104766479A (en) | Automobile identity recognition method and device based on ultrahigh frequency radio frequency and video image dual-recognition matching | |
CN106297304A (en) | A kind of based on MapReduce towards the fake-licensed car recognition methods of extensive bayonet socket data | |
CN106710225B (en) | Vehicle number plate violation identification method and monitoring platform | |
CN104021677A (en) | Method and device for detecting turnover of number plate | |
CN103514745A (en) | Fake license plate vehicle recognition method based on intelligent transportation | |
CN104750800A (en) | Motor vehicle clustering method based on travel time characteristic | |
CN112734219B (en) | Vehicle transportation running behavior analysis method and system | |
CN102122348A (en) | Practical method for recovering fuzzy license plate image | |
CN112925820B (en) | Method, device and system for identifying vehicle evasion toll | |
CN112509325B (en) | Video deep learning-based off-site illegal automatic discrimination method | |
CN103310199A (en) | Vehicle model identification method based on high-resolution remote sensing data | |
CN108305461A (en) | A kind of determination method and apparatus for evading expense suspected vehicles | |
CN103700262A (en) | Automatic area acquisition method for fake-licensed vehicles | |
CN113470369B (en) | Method and system for judging true number plate of fake-licensed vehicle based on multi-dimensional information | |
CN112967410A (en) | Method for identifying toll-evasion vehicles based on longest public subsequence |
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 | ||
CP01 | Change in the name or title of a patent holder | ||
CP01 | Change in the name or title of a patent holder |
Address after: 310012 1st floor, building 1, 223 Yile Road, Hangzhou City, Zhejiang Province Patentee after: Yinjiang Technology Co.,Ltd. Address before: 310012 1st floor, building 1, 223 Yile Road, Hangzhou City, Zhejiang Province Patentee before: ENJOYOR Co.,Ltd. |