CN103065138A - Recognition method of license plate number of motor vehicle - Google Patents
Recognition method of license plate number of motor vehicle Download PDFInfo
- Publication number
- CN103065138A CN103065138A CN2012105204044A CN201210520404A CN103065138A CN 103065138 A CN103065138 A CN 103065138A CN 2012105204044 A CN2012105204044 A CN 2012105204044A CN 201210520404 A CN201210520404 A CN 201210520404A CN 103065138 A CN103065138 A CN 103065138A
- Authority
- CN
- China
- Prior art keywords
- license plate
- pixel
- connected region
- plate area
- image
- 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
Images
Abstract
The invention relates to a recognition method of a license plate number of a motor vehicle, and belongs to the technical field of license plate number recognition. The recognition method comprises carrying out median filtering and noise reduction to an input image containing a license plate, locating multiple license plate areas by utilizing a binarization processing method and a morphological processing method, carrying out rotation correction and edge removing processing to the license plate areas one by one, carrying out character cutting by utilizing a vertical projection method, carrying out further edge removing processing to obtained character blocks in a cutting mode, carrying out character recognition by utilizing an enhanced edition template matching method, and therefore achieving fast and accurate license plate recognition. The recognition method improves the precision of license plate location, and fast and precise location of the license plate is achieved. The recognition method can be used in a real-time license place recognition system, and for processing high-speed data flow, meanwhile is high in recognition accuracy, and has certain robustness on illumination variation, influence of rain and mist, indistinct license plates, broken characters and other bad conditions.
Description
Technical field
The present invention relates to a kind of recognition methods of the motor vehicle trade mark, belong to the license plate recognition technology field.
Background technology
License plate recognition technology (VehicleLicense Plate Recognition, VLPR) refers to detect the vehicle on monitored road surface and the automatic lifting technology that a license plate information (containing Chinese character, English alphabet, arabic numeral and number plate color) processes of picking up the car.Car plate identification is one of important component part in the modern intelligent transportation system, uses very extensive.It is analyzed vehicle image or the video sequence of shot by camera take technology such as Digital Image Processing, pattern-recognition, computer visions as the basis, obtains the unique number-plate number of each automobile, thus the identifying of finishing.Can realize parking lot fee collection management by some subsequent treatment means, magnitude of traffic flow control index measurement, vehicle location, automobile burglar, high way super speed robotization supervision, electronic eye used for catching red light runner, toll station etc. function.For safeguarding traffic safety and urban public security, prevent traffic jam, realize that the traffic automation management has the meaning of reality.
Usually licence plate recognition method comprises following three steps: car plate location, License Plate Character Segmentation, Recognition of License Plate Characters.And the index of estimating licence plate recognition method mainly comprises processing time, recognition accuracy etc., more than in three steps each step all can have influence on the performance of whole licence plate recognition method.
Existing various licence plate recognition method mostly exists a lot of defectives.Thereby algorithm of locating license plate of vehicle is consuming time many and be subjected to easily the accuracy that car plate place complex background disturbs affects the location, because the diversity of the license plate image that gathers, and be subject to being permitted multifactorial impact when gathering image, such as the rainy day, dense fog, light etc., so that some license plate image quality appearance difference is in various degree arranged, in the ordinary course of things, the background of the image that gathers is very complicated, gathering image is the picture that gathers in high-speed motion, so the position of car plate is unfixing in the picture, the size of car plate is also different, above all disturbing factors, all brought difficulty to the license plate area location, existing method does not well address this problem.
Along with the development from the analogue camera to the high definition camera, the contradiction of image high-resolution and recognition speed is more and more noticeable.The advantage of high definition is self-evident, but when identifying, car plate also can cause some problems, the high definition picture is because the picture broad covered area, the identification of a plurality of car plates may appear in picture simultaneously, existing Vehicle License Plate Recognition System concentrates on mostly processes the image that only comprises a car, comprise at input picture in the situation of a lot of motor vehicles, existing Vehicle License Plate Recognition System can not show very high performance.In addition, the HD video code stream is very large, and the processing speed of Vehicle License Plate Recognition System has been proposed very high requirement, if processing speed is crossed the phenomenon that may cause leaking slowly car, thereby is difficult to realize lifting to vehicle snapshot rate and car plate recognition accuracy.
In addition, at present in the Recognition Algorithm of License Plate part, owing in the car plate of China English alphabet and arabic numeral are arranged not only, also comprise the numerous and diverse Chinese character of stroke, cause its identification difficulty external only much bigger to the situation of letter, numeral identification, the Some Domestic road conditions is poor in addition, the car plate damage ratio is more serious, and the number plate of vehicle of actual motion can be subject to the impacts such as mud, oil, lacquer, characters on license plate often has the phenomenon of fracture, the number plate manufacture craft character that also can occur lack of standardization shoals, fogs simultaneously, makes the difficulty of Recognition of License Plate Characters become large.Template matching method is used widely in the car plate automatic identification field because of its fireballing characteristic, but under above all abominable conditions, the precision of template matching method is limited by very large, in addition, template matching method is not good for the rotation resistance, has limited to a certain extent its range of application yet.Use other methods such as neural network to identify, can bring no small lifting processing complexity again, in general, existing licence plate recognition method, possess at the same time also exist on higher processing speed and the larger recognition accuracy a lot of not enough.
Summary of the invention
The objective of the invention is to propose a kind of recognition methods of the motor vehicle trade mark, existing binary Images Processing is combined with the morphology processing, carry out many car plates location, then by slightly to smart car plate being rotated rectification, the enhanced edition template matches that combines with the rarefaction representation theory is to realize the high-performance car plate identification of low computation complexity, high recognition accuracy.
The recognition methods of the motor vehicle trade mark that the present invention proposes may further comprise the steps:
(1) read in coloured image or a gray level image that contains car plate, if this image is coloured image, then be translated into gray level image, the note gray level image is initial pictures G;
(2) utilize median filter method that initial pictures G is carried out pre-service, weaken salt-pepper noise, obtain pretreated image
(3) utilize binary conversion treatment and morphology disposal route, from pretreated image
In orient license plate area, obtain pretreated image
In the set { P of all license plate areas
i(i=1,2 .., N), wherein N is total number of license plate area, P
iBe i license plate area, concrete steps are as follows:
(3-1) at pretreated image
On carry out rim detection, obtain pretreated image
Edge image E, edge image E is bianry image, if the pixel value of the pixel (x, y) among the edge image E is 1, then this pixel (x, y) is marginal point, if the pixel value of pixel (x, y) is 0, then this pixel (x, y) is non-marginal point;
(3-2) with among the edge image E, be set to 0 with the pixel value of marginal point on the edge that the borderline phase of edge image E is connected, obtain pretreated edge image
(3-3) definition one wide be W1 pixel, the height be the structural element of H1 pixel, to above-mentioned pretreated edge image
Carry out closing operation of mathematical morphology, obtain the edge image E after the closed operation
1
(3-4) one of definition is wide be that W2 pixel, height are the structural element of H2 pixel, and edge image E is exported in closed operation
1Carry out the morphology opening operation, obtain the edge image E behind the opening operation
2
The set that the pixel that (3-5) is communicated with in the definition image forms is connected region, according to the edge image E behind the opening operation
2In all pixel values be the connectedness of 1 pixel, to the edge image E behind the opening operation
2In all pixel values be that 1 pixel is divided, the connected region that the pixel that obtains being communicated with forms makes edge image E
2In each pixel value be that 1 pixel is corresponding with a connected region;
(3-6) each connected region is judged that obtain the set of the corresponding connected region of all motor vehicle trades mark to be identified, determining step is as follows:
(3-6-1) from the connected region of step (3-5), get a connected region A who did not judge and judge, if all connected regions in the connected region of step (3-5) all were judged, then carry out step (3-6-5);
The ratio of width to height of the connected region that (3-6-2) the setting car plate is corresponding is 3～5.5, the ratio of width to height of the bounding box of connected region A is compared with the ratio of width to height of setting license plate area, if the ratio of width to height of the bounding box of connected region A is greater than the upper limit of the ratio of width to height of setting license plate area, or the ratio of width to height of the bounding box of connected region A is less than the lower limit of the ratio of width to height of setting license plate area, judge that then connected region A is not the connected region corresponding with car plate, and carry out step (3-6-1);
If the ratio of width to height of the bounding box of connected region A is less than or equal to the upper limit of the ratio of width to height of setting license plate area, or the ratio of width to height of the bounding box of connected region A is more than or equal to the lower limit of the ratio of width to height of setting license plate area, then judge connected region A as drafting connected region corresponding to car plate, and carry out step (3-6-3);
(3-6-3) from pretreated image
In, take out a gray level image that the pixel in the bounding box with connected region A is corresponding
Adopt adaptive threshold, with gray level image
Binaryzation obtains the bianry image B corresponding with the pixel in bounding box connected region A after the binaryzation
A, set the in the horizontal direction frequency threshold value that replaces of black and white of a connected region corresponding with car plate, if bianry image B
AThe number of times that replaces of black and white is judged that then connected region A is not connected region corresponding to car plate, and is carried out step (3-6-1), if bianry image B less than frequency threshold value in the horizontal direction
AThe number of times that replaces of black and white is judged that then connected region A drafts connected region corresponding to car plate, and is carried out step (3-6-4) more than or equal to frequency threshold value in the horizontal direction;
(3-6-4) calculate bianry image B
AInterior pixel value is that 1 pixel number accounts for bianry image B
AInterior all pixel number purpose ratio α, if α is not between 0.3～0.4, judge that then connected region A is not connected region corresponding to car plate, if α is between 0.3～0.4, judge that then connected region A is connected region corresponding to car plate, get back to step (3-6-1), carry out the judgement of next connected region;
The set of (3-6-5) establishing connected region corresponding to all car plates is { A
i(i=1,2 .., N), wherein A
iBe i the connected region that car plate is corresponding, in order from the set { A of connected region
iThe middle taking-up connected region A corresponding with each car plate
iBounding box, to the pixel in the bounding box at pretreated image
The pixel of interior correspondence carries out binary conversion treatment, obtains a pixel set P
i, note pixel set P
iBe license plate area, thereby the set that obtains all license plate areas is { P
i(i=1,2 .., N), wherein P
iBe i license plate area;
(4) to above-mentioned license plate area P
iBe rotated and correct and the processing of trimming edge the license plate area after obtaining processing
Concrete steps are as follows:
(4-1) utilize edge detection operator, from license plate area P
iMiddle extraction and license plate area P
iCorresponding binary edge map E
Pi, calculate binary edge map E
PiDraw the winter (Radon) conversion, from the matrix that conversion obtains, obtain the matrix peak value, the angle θ corresponding with the matrix peak value is license plate area P
iAnglec of rotation θ, according to the anglec of rotation θ of car plate to license plate area P
iIn all pixels carry out reverse rotation θ angle, rotated roughly the license plate area P after the rectification
I1
(4-2) remove license plate area P
I1In the constant band of pixel value of boundary pixel up and down, the license plate area P that is tentatively tightened
I2
The license plate area P that (4-3) will tentatively tighten
I2In the pixel value of every delegation pixel sue for peace, the pixel value sum of establishing the capable pixel of m is S
m, by sequence number m inferior ordered pair S from small to large
mSort, obtain S
mFirst maximum value S
M0, establish and S
M0Corresponding line number is m
0, to the license plate area P of above-mentioned preliminary deflation
I2In the 1st～m
0Pixel value is that 1 pixel carries out fitting of a polynomial one time in the row, the license plate area P that is tentatively tightened according to polynomial slope k once
I2The pitch angle
According to the pitch angle
License plate area P to preliminary deflation
I2Interior all pixels carry out reverse rotation, obtain the license plate area P after meticulous rotation is corrected
I3
(4-4) remove license plate area P
I3In the constant band of up and down boundary pixel pixel value, obtain the license plate area P that secondary tightens
I4
(4-5) utilize interpolation method, to the license plate area P of secondary deflation
I4Carry out proportional zoom, so that license plate area is high for setting number pixel H, obtain a license plate area after the processing behind the proportional zoom
(5) adopt vertical projection method, to the license plate area after the above-mentioned processing
Carry out Character segmentation and process the license plate area after the computing
Two-value perspective view in vertical direction, and searching trough point, from the trough point whole car plate is carried out Character segmentation, obtain 7 character blocks, remove in each character block the up and down constant band of the pixel value of boundary pixel, character block after obtaining tightening, and the character block after will tightening utilizes interpolation method to zoom to
Namely wide
Individual pixel, a skyer pixel obtains character block to be identified;
(6) template matching method of employing enhanced edition carries out character recognition to above-mentioned character block to be identified, and the implementation step is as follows:
(6-1) from above-mentioned character block to be identified, get a character block C to be identified
Ij, the character block C that this is to be identified
IjBe divided into four zones, the pixel in each piece zone is lined up a column vector, and four column vectors corresponding with four zones are got up in succession, consist of this character block C to be identified
IjProper vector
(6-2) gather the car plate samples pictures, to each character that may occur in the car plate, from samples pictures, manually mark the respective symbols piece, and with the character block size scaling
Carry out proper vector according to the method in the step (6-1) and extract, the proper vector of all characters is saved as a dictionary D, each proper vector is a base among the dictionary D;
(6-3) adopt iteration threshold to select projection algorithm (Iterative Threshold-Selective Projection algorithm), find the solution above-mentioned C
IjProper vector
The most sparse solution vector under dictionary D
Solution vector
Each component ω
Ij-kExpression C
IjProper vector
K base in dictionary D
Weight;
(6-4) calculate under the two norm meanings each base among the dictionary D
By with
Corresponding weights omega
Ij-kCarry out vector and C behind the proportional zoom
IjProper vector
Between apart from d
Ij-k, establish apart from d
Ij-kIn minor increment be d
Ij-k0, with this minor increment d
Ij-k0Base among the corresponding dictionary D is
Corresponding character is character block C to be identified
IjRepresented character:
(6-5) repeating step (6-1)-(6-4) is until the license plate area after processing
In all character blocks to be identified all identify complete;
(7) repeating step (4)-(6) are until the set { P of license plate area
i(i=1,2 ..., N) in all license plate areas all oneself is disposed, the output recognition result.
The licence plate recognition method that the present invention proposes, its advantage is:
1, the present invention's use combines Morphology Algorithm and carries out license plate locating method with binary Images Processing, take full advantage of the fireballing advantage of binary Images Processing, introduce simultaneously Morphology Algorithm, can effectively remove unwanted marginal information for the interference of car plate location, remedy when utilizing binary Images Processing to carry out the car plate location for the comparatively responsive deficiency of marginal information.The present invention has promoted the precision that car plate is located by introducing priori to a certain degree greatly under on the less prerequisite of generalization impact, realized the location not only fast but also accurate to car plate.
2, the present invention can disposablely detect a plurality of license plate areas in the single image simultaneously, and fast processing identification, has realized preferably many car plates recognition function.
3, the present invention has adopted a kind of by thick rotation antidote to essence, can effectively correct the tilt phenomenon that license plate area occurs, for follow-up cut apart, recognizer provides good input, thereby greatly promote the accuracy of identification of a whole set of licence plate recognition method, and the anti-revolving property of introducing has thus enlarged the scope of application of this licence plate recognition method.
4, the present invention improves the conventional template matching algorithm, and is theoretical by using rarefaction representation, on the basis of conventional template matching algorithm speed advantage, further promoted computing velocity, greatly improved the identification accuracy simultaneously.The present invention has carried out meticulousr Division identification for the character of easily obscuring after template matches, thereby has very low identification error rate.
5, the present invention can be used for the Real-time Vehicle License Plate recognition system, processes high-speed data-flow, has simultaneously very high recognition accuracy, and the mal-conditions such as, character fracture fuzzy for illumination variation, misty rain impact, car plate have certain robustness.
Description of drawings
Fig. 1 is the FB(flow block) of the licence plate recognition method that proposes of the present invention.
Fig. 2 is the FB(flow block) of carrying out the license plate area location in the inventive method.
Embodiment
The licence plate recognition method that the present invention proposes, its FB(flow block) may further comprise the steps as shown in Figure 1:
(1) read in coloured image or a gray level image that contains car plate, if this image is coloured image, then be translated into gray level image, the note gray level image is initial pictures G;
(2) utilize median filter method that initial pictures G is carried out pre-service, weaken salt-pepper noise, obtain pretreated image
(3) utilize binary conversion treatment and morphology disposal route, from pretreated image
In orient license plate area, obtain pretreated image
In the set { P of all license plate areas
i(i=1,2 .., N), wherein N is total number of license plate area, P
iBe i license plate area, its FB(flow block) as shown in Figure 2, concrete steps are as follows:
(3-1) at pretreated image
On carry out rim detection, can carry out rim detection according to the different edge detection operator of practical application scene choice for use, in experiment, used the Sobel operator to carry out rim detection, obtain pretreated image
Edge image E, edge image E is bianry image, if the pixel value of the pixel (x, y) among the edge image E is 1, then this pixel (x, y) is marginal point, represents that it is positioned at pretreated image
In the edge on, if the pixel value of pixel (x, y) is 0, then this pixel (x, y) is non-marginal point, represents that this point is not positioned at pretreated image
In the edge on;
(3-2) with among the edge image E, be set to 0 with the pixel value of marginal point on the edge that the borderline phase of edge image E is connected, be about to this part point and be classified as non-marginal point, obtain pretreated edge image
(3-3) definition one wide be W1 pixel, the height be the structural element of H1 pixel, to above-mentioned pretreated edge image
Carry out closing operation of mathematical morphology, obtain the edge image E after the closed operation
1, having chosen wide in one embodiment of the present of invention is that 30 pixels, height are that the structural element of 1 pixel carries out closed operation;
Expand first in the morphology operations, after the process of corroding, be called closing operation of mathematical morphology, expansion algorithm is as follows:
With a certain specific structural element, each pixel of scan image, bianry image with structural element and its covering is done AND-operation, if all AND-operation results are 0, then this pixel pixel value of computing output image is 0, otherwise this pixel pixel value of computing output image is 1;
Erosion algorithm is as follows:
With a certain specific structural element, each pixel of scan image, bianry image with structural element and its covering is done AND-operation, if all AND-operation results are 1, then this pixel pixel value of computing output image is 1, otherwise this pixel pixel value of computing output image is 0;
Edge image E expands first, finishes the closed operation of edge image after corroding again, obtains closed operation output edge image E
1
(3-4) one of definition is wide be that W2 pixel, height are the structural element of H2 pixel, and edge image E is exported in closed operation
1Carry out the morphology opening operation, obtain the edge image E behind the opening operation
2, chosen the structural element of wide 30 pixels, high 1 pixel in one embodiment of the present of invention, to closed operation output edge image E
1Carry out the morphology opening operation, corrode first in the morphology operations, after the process that expands, be called the morphology opening operation;
The set that the pixel that (3-5) is communicated with in the definition image forms is connected region, according to the edge image E behind the opening operation
2In all pixel values be the connectedness of 1 pixel, to the edge image E behind the opening operation
2In all pixel values be that 1 pixel is divided, the connected region that the pixel that obtains being communicated with forms makes edge image E
2In each pixel value be that 1 pixel is corresponding with a connected region;
(3-6) each connected region is judged that obtain the set of the corresponding connected region of all car plates to be identified, determining step is as follows:
(3-6-1) from the connected region of step (3-5), get a connected region A who did not judge and judge, if all connected regions in the connected region of step (3-5) all were judged, then carry out step (3-6-5);
The ratio of width to height of the connected region that (3-6-2) the setting car plate is corresponding is 3～5.5, the ratio of width to height of the bounding box of connected region A is compared with the ratio of width to height of setting license plate area, if the ratio of width to height of the bounding box of connected region A is greater than the upper limit of the ratio of width to height of setting license plate area, or the ratio of width to height of the bounding box of connected region A is less than the lower limit of the ratio of width to height of setting license plate area, judge that then connected region A is not the connected region corresponding with car plate, and carry out step (3-6-1);
If the ratio of width to height of the bounding box of connected region A is less than or equal to the upper limit of the ratio of width to height of setting license plate area, or the ratio of width to height of the bounding box of connected region A is more than or equal to the lower limit of the ratio of width to height of setting license plate area, then judge connected region A as drafting connected region corresponding to car plate, and carry out step (3-6-3);
(3-6-3) from pretreated image
In, take out a gray level image that the pixel in the bounding box with connected region A is corresponding
Adopt adaptive threshold, with gray level image
Binaryzation obtains the bianry image B corresponding with the pixel in bounding box connected region A after the binaryzation
ASet the in the horizontal direction frequency threshold value that replaces of black and white of a connected region corresponding with car plate, because the domestic number-plate number is 7, the black and white alternate frequency should be above 14 times in the horizontal direction for interior bianry image corresponding to pixel of the connected region bounding box scope that car plate is corresponding, so this threshold value is set to 14 in this experiment, if bianry image B
AThe number of times that replaces of black and white is judged that then connected region A is not connected region corresponding to car plate, and is carried out step (3-6-1), if bianry image B less than frequency threshold value in the horizontal direction
AThe number of times that replaces of black and white is judged that then connected region A drafts connected region corresponding to car plate, and is carried out step (3-6-4) more than or equal to frequency threshold value in the horizontal direction;
(3-6-4) calculate bianry image B
AInterior pixel value is that 1 pixel number accounts for bianry image B
AInterior all pixel number purpose ratio α, if α is not between 0.3～0.4, judge that then connected region A is not connected region corresponding to car plate, if α is between 0.3～0.4, judge that then connected region A is connected region corresponding to car plate, get back to step (3-6-1), carry out the judgement of next connected region;
The set of (3-6-5) establishing connected region corresponding to all car plates is { A
i(i=1,2 .., N), wherein A
iBe i the connected region that car plate is corresponding, in order from the set { A of connected region
iThe middle taking-up connected region A corresponding with each car plate
iBounding box, to the pixel in the bounding box at pretreated image
The pixel of interior correspondence carries out binary conversion treatment, obtains a pixel set P
i, note pixel set P
iBe license plate area, thereby the set that obtains all license plate areas is { P
i(i=1,2 .., N), wherein P
iBe i license plate area;
(4) to above-mentioned license plate area P
iBe rotated and correct and the processing of trimming edge the license plate area after obtaining processing
Prepare for follow-up separating character and identification, concrete steps are as follows:
(4-1) utilize edge detection operator, from license plate area P
iMiddle extraction and license plate area P
iCorresponding binary edge map E
Pi, calculate binary edge map E
PiDraw the winter (Radon) conversion, from the matrix that conversion obtains, obtain the matrix peak value, the angle θ corresponding with the matrix peak value is license plate area P
iAnglec of rotation θ, according to the anglec of rotation θ of car plate to license plate area P
iIn all pixels carry out reverse rotation θ angle, rotated roughly the license plate area P after the rectification
I1
(4-2) remove license plate area P
I1In the constant band of pixel value of boundary pixel up and down, remove whereby the car plate edge, license plate area is tightened the license plate area P that is tentatively tightened
I2
The license plate area P that (4-3) will tentatively tighten
I2In the pixel value of every delegation pixel sue for peace, the pixel value sum of establishing the capable pixel of m is S
m, by sequence number m inferior ordered pair S from small to large
mSort, obtain S
mFirst maximum value S
M0, establish and S
M0Corresponding line number is m
0, to the license plate area P of above-mentioned preliminary deflation
I2In the 1st～m
0Pixel value is that 1 pixel carries out fitting of a polynomial one time in the row, the license plate area P that is tentatively tightened according to polynomial slope k once
I2The pitch angle
According to the pitch angle
License plate area P to preliminary deflation
I2Interior all pixels carry out reverse rotation, obtain the license plate area P after meticulous rotation is corrected
I3
(4-4) remove license plate area P
I3In the constant band of up and down boundary pixel pixel value, further remove whereby the car plate edge, license plate area is further tightened, obtain the license plate area P that secondary tightens
I4
(4-5) utilize interpolation method, to the license plate area P of secondary deflation
I4Carry out proportional zoom, so that license plate area is high for setting for example 30 pixels of number pixel H(), obtain a license plate area after the processing behind the proportional zoom
(5) adopt vertical projection method, to the license plate area after the above-mentioned processing
Carry out Character segmentation and process the license plate area after the computing
Two-value perspective view in vertical direction, and searching trough point, from the trough point whole car plate is carried out Character segmentation, obtain 7 character blocks, remove in each character block the up and down constant band of the pixel value of boundary pixel, character block after obtaining tightening, and the character block after will tightening utilizes interpolation method to zoom to
Namely wide
Individual pixel, a high H pixel obtains character block to be identified, the H value is 30 in one embodiment of the present of invention, because external factor is a bit smudgy clear the time, the better effects if of this method is so this dividing method can effectively be resisted certain mal-condition when car plate;
(6) template matching method of employing enhanced edition carries out character recognition to above-mentioned character block to be identified, and the implementation step is as follows:
(6-1) from above-mentioned character block to be identified, get a character block C to be identified
Ij, the character block C that this is to be identified
IjBe divided into four zones, the pixel in each piece zone is lined up a column vector, and four column vectors corresponding with four zones are got up in succession, consist of this character block C to be identified
IjProper vector
(6-2) gather the car plate samples pictures, to each character that may occur in the car plate, from samples pictures, manually mark the respective symbols piece, and with the character block size scaling
Carry out proper vector according to the method in the step (6-1) and extract, the proper vector of all characters is saved as a dictionary D, each proper vector is a base among the dictionary D;
(6-3) adopt iteration threshold to select projection algorithm (Iterative Threshold-Selective Projection algorithm), find the solution above-mentioned C
IjProper vector
The most sparse solution vector under dictionary D
Solution vector
Each component ω
Ij-kExpression C
IjProper vector
K base in dictionary D
Weight;
(6-4) calculate under the two norm meanings each base among the dictionary D
By with
Corresponding weights omega
Ij-kCarry out vector and C behind the proportional zoom
IjProper vector
Between apart from d
Ij-k, establish apart from d
Ij-kIn minor increment be d
Ij-k0, with this minor increment d
Ij-k0Base among the corresponding dictionary D is
Corresponding character is character block C to be identified
IjRepresented character relatively is difficult to the character range distinguished when recognition result falls into some, during such as O, D, G, needs to adopt the geometric properties of kinds of characters further to distinguish, because C in the method
IjProper vector
The most sparse solution vector under dictionary D
Has sparse property, so do not need to calculate C
IjProper vector
The distance of all bases in the dictionary D, thereby effective boosting algorithm speed, and to each base
Press ω
Ij-kCarry out proportional zoom, in the situation of boosting algorithm complexity not, also effectively raise recognition accuracy;
(6-5) repeating step (6-1)-(6-4) is until the license plate area after processing
In all character blocks to be identified all identify complete;
(7) repeating step (4)-(6) are until the set { P of license plate area
iAll license plate areas all are disposed in (i=1,2 .., N), the output recognition result.
Claims (1)
1. the recognition methods of a motor vehicle trade mark is characterized in that the method may further comprise the steps:
(1) read in coloured image or a gray level image that contains car plate, if this image is coloured image, then be translated into gray level image, the note gray level image is initial pictures G;
(2) utilize median filter method that initial pictures G is carried out pre-service, weaken salt-pepper noise, obtain pretreated image
(3) utilize binary conversion treatment and morphology disposal route, from pretreated image
In orient license plate area, obtain pretreated image
In the set { P of all license plate areas
i(i=1,2 .., N), wherein N is total number of license plate area, P
iBe i license plate area, concrete steps are as follows:
(3-1) at pretreated image
On carry out rim detection, obtain pretreated image
Edge image E, edge image E is bianry image, if the pixel value of the pixel (x, y) among the edge image E is 1, then this pixel (x, y) is marginal point, if the pixel value of pixel (x, y) is 0, then this pixel (x, y) is non-marginal point;
(3-2) with among the edge image E, be set to 0 with the pixel value of marginal point on the edge that the borderline phase of edge image E is connected, obtain pretreated edge image
(3-3) definition one wide be W1 pixel, the height be the structural element of H1 pixel, to above-mentioned pretreated edge image
Carry out closing operation of mathematical morphology, obtain the edge image E after the closed operation
1
(3-4) one of definition is wide be that W2 pixel, height are the structural element of H2 pixel, and edge image E is exported in closed operation
1Carry out the morphology opening operation, obtain the edge image E behind the opening operation
2
The set that the pixel that (3-5) is communicated with in the definition image forms is connected region, according to the edge image E behind the opening operation
2In all pixel values be the connectedness of 1 pixel, to the edge image E behind the opening operation
2In all pixel values be that 1 pixel is divided, the connected region that the pixel that obtains being communicated with forms makes edge image E
2In each pixel value be that 1 pixel is corresponding with a connected region;
(3-6) each connected region is judged that obtain the set of the corresponding connected region of all motor vehicle trades mark to be identified, determining step is as follows:
(3-6-1) from the connected region of step (3-5), get a connected region A who did not judge and judge, if all connected regions in the connected region of step (3-5) all were judged, then carry out step (3-6-5);
The ratio of width to height of the connected region that (3-6-2) the setting car plate is corresponding is 3～5.5, the ratio of width to height of the bounding box of connected region A is compared with the ratio of width to height of setting license plate area, if the ratio of width to height of the bounding box of connected region A is greater than the upper limit of the ratio of width to height of setting license plate area, or the ratio of width to height of the bounding box of connected region A is less than the lower limit of the ratio of width to height of setting license plate area, judge that then connected region A is not the connected region corresponding with car plate, and carry out step (3-6-1);
If the ratio of width to height of the bounding box of connected region A is less than or equal to the upper limit of the ratio of width to height of setting license plate area, or the ratio of width to height of the bounding box of connected region A is more than or equal to the lower limit of the ratio of width to height of setting license plate area, then judge connected region A as drafting connected region corresponding to car plate, and carry out step (3-6-3);
(3-6-3) from pretreated image
In, take out a gray level image that the pixel in the bounding box with connected region A is corresponding
Adopt adaptive threshold, with gray level image
Binaryzation obtains the bianry image B corresponding with the pixel in bounding box connected region A after the binaryzation
A, set the in the horizontal direction frequency threshold value that replaces of black and white of a connected region corresponding with car plate, if bianry image B
AThe number of times that replaces of black and white is judged that then connected region A is not connected region corresponding to car plate, and is carried out step (3-6-1), if bianry image B less than frequency threshold value in the horizontal direction
AThe number of times that replaces of black and white is judged that then connected region A drafts connected region corresponding to car plate, and is carried out step (3-6-4) more than or equal to frequency threshold value in the horizontal direction;
(3-6-4) calculate bianry image B
AInterior pixel value is that 1 pixel number accounts for bianry image B
AInterior all pixel number purpose ratio α, if α is not between 0.3～0.4, judge that then connected region A is not connected region corresponding to car plate, if α is between 0.3～0.4, judge that then connected region A is connected region corresponding to car plate, get back to step (3-6-1), carry out the judgement of next connected region;
The set of (3-6-5) establishing connected region corresponding to all car plates is { A
i(i=1,2 .., N), wherein A
iBe i the connected region that car plate is corresponding, in order from the set { A of connected region
iThe middle taking-up connected region A corresponding with each car plate
iBounding box, to the pixel in the bounding box at pretreated image
The pixel of interior correspondence carries out binary conversion treatment, obtains a pixel set P
i, note pixel set P
iBe license plate area, thereby the set that obtains all license plate areas is { P
i(i=1,2 .., N), wherein P
iBe i license plate area;
(4) to above-mentioned license plate area P
iBe rotated and correct and the processing of trimming edge the license plate area after obtaining processing
Concrete steps are as follows:
(4-1) utilize edge detection operator, from license plate area P
iMiddle extraction and license plate area P
iCorresponding binary edge map E
Pi, calculate binary edge map E
PiDraw the winter (Radon) conversion, from the matrix that conversion obtains, obtain the matrix peak value, the angle θ corresponding with the matrix peak value is license plate area P
iAnglec of rotation θ, according to the anglec of rotation θ of car plate to license plate area P
iIn all pixels carry out reverse rotation θ angle, rotated roughly the license plate area P after the rectification
I1
(4-2) remove license plate area P
I1In the constant band of pixel value of boundary pixel up and down, the license plate area P that is tentatively tightened
I2
The license plate area P that (4-3) will tentatively tighten
I2In the pixel value of every delegation pixel sue for peace, the pixel value sum of establishing the capable pixel of m is S
m, by sequence number m inferior ordered pair S from small to large
mSort, obtain S
mFirst maximum value S
M0, establish and S
M0Corresponding line number is m
0, to the license plate area P of above-mentioned preliminary deflation
I2In the 1st～m
0Pixel value is that 1 pixel carries out fitting of a polynomial one time in the row, the license plate area P that is tentatively tightened according to polynomial slope k once
I2The pitch angle
According to the pitch angle
License plate area P to preliminary deflation
I2Interior all pixels carry out reverse rotation, obtain the license plate area P after meticulous rotation is corrected
I3
(4-4) remove license plate area P
I3In the constant band of up and down boundary pixel pixel value, obtain the license plate area P that secondary tightens
I4
(4-5) utilize interpolation method, to the license plate area P of secondary deflation
I4Carry out proportional zoom, so that license plate area is high for setting number pixel H, obtain a license plate area after the processing behind the proportional zoom
(5) adopt vertical projection method, to the license plate area after the above-mentioned processing
Carry out Character segmentation and process the license plate area after the computing
Two-value perspective view in vertical direction, and searching trough point, from the trough point whole car plate is carried out Character segmentation, obtain 7 character blocks, remove in each character block the up and down constant band of the pixel value of boundary pixel, character block after obtaining tightening, and the character block after will tightening utilizes interpolation method to zoom to
Namely wide
Individual pixel, a high H pixel obtains character block to be identified;
(6) template matching method of employing enhanced edition carries out character recognition to above-mentioned character block to be identified, and the implementation step is as follows:
(6-1) from above-mentioned character block to be identified, get a character block C to be identified
Ij, the character block C that this is to be identified
IjBe divided into four zones, the pixel in each piece zone is lined up a column vector, and four column vectors corresponding with four zones are got up in succession, consist of this character block C to be identified
IjProper vector
(6-2) gather the car plate samples pictures, to each character that may occur in the car plate, from samples pictures, manually mark the respective symbols piece, and with the character block size scaling
Carry out proper vector according to the method in the step (6-1) and extract, the proper vector of all characters is saved as a dictionary D, each proper vector is a base among the dictionary D;
(6-3) adopt iteration threshold to select projection algorithm (Iterative Threshold-Selective Projection algorithm), find the solution above-mentioned C
IjProper vector
The most sparse solution vector under dictionary D
Solution vector
Each component ω
Ij-kExpression C
IjProper vector
K base in dictionary D
Weight;
(6-4) calculate under the two norm meanings each base among the dictionary D
By with
Corresponding weights omega
Ij-kCarry out vector and C behind the proportional zoom
IjProper vector
Between apart from d
Ij-k, establish apart from d
Ij-kIn minor increment be d
Ij-k0, with this minor increment d
Ij-k0Base among the corresponding dictionary D is
Corresponding character is character block C to be identified
IjRepresented character;
(6-5) repeating step (6-1)-(6-4) is until the license plate area after processing
In all character blocks to be identified all identify complete;
(7) repeating step (4)-(6) are until the set { P of license plate area
iAll license plate areas all are disposed in (i=1,2 .., N), the output recognition result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210520404.4A CN103065138B (en) | 2012-12-06 | 2012-12-06 | Recognition method of license plate number of motor vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210520404.4A CN103065138B (en) | 2012-12-06 | 2012-12-06 | Recognition method of license plate number of motor vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103065138A true CN103065138A (en) | 2013-04-24 |
| CN103065138B CN103065138B (en) | 2015-07-15 |
Family
ID=48107761
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210520404.4A Active CN103065138B (en) | 2012-12-06 | 2012-12-06 | Recognition method of license plate number of motor vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103065138B (en) |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103984949A (en) * | 2014-06-11 | 2014-08-13 | 四川九洲电器集团有限责任公司 | License plate positioning method and system based on high and low cap transformation and connected domain |
| CN104036262A (en) * | 2014-06-30 | 2014-09-10 | 南京富士通南大软件技术有限公司 | Method and system for screening and recognizing LPR license plate |
| CN104156731A (en) * | 2014-07-31 | 2014-11-19 | 成都易默生汽车技术有限公司 | License plate recognition system based on artificial neural network and method |
| CN104156704A (en) * | 2014-08-04 | 2014-11-19 | 胡艳艳 | Novel license plate identification method and system |
| CN104239878A (en) * | 2014-08-30 | 2014-12-24 | 电子科技大学 | License plate character recognition method based on probability extremum search |
| CN104239864A (en) * | 2014-09-16 | 2014-12-24 | 哈尔滨恒誉名翔科技有限公司 | Freight car number identification system based on image processing |
| CN104408438A (en) * | 2014-12-10 | 2015-03-11 | 黑龙江真美广播通讯器材有限公司 | Truck number identification system |
| CN105139659A (en) * | 2015-10-09 | 2015-12-09 | 东方网力科技股份有限公司 | Vehicle license plate recognition method and device |
| CN105184227A (en) * | 2015-08-13 | 2015-12-23 | 安徽超远信息技术有限公司 | Traffic signal control system with automobile high beam light detection function |
| CN105447487A (en) * | 2015-08-27 | 2016-03-30 | 中山大学 | Evaluation method and system for vehicle license plate identification system |
| CN106127206A (en) * | 2016-06-28 | 2016-11-16 | 北京智芯原动科技有限公司 | The vertical angles detection method of a kind of car plate and device |
| CN106127707A (en) * | 2016-06-23 | 2016-11-16 | 南京大学金陵学院 | A kind of processing method of supersonic blood image |
| CN106295528A (en) * | 2016-08-01 | 2017-01-04 | 长安大学 | A kind of vehicle checking method based on multi-part spatial relation GMM modeling |
| CN106469267A (en) * | 2015-08-20 | 2017-03-01 | 深圳市腾讯计算机系统有限公司 | A kind of identifying code sample collection method and system |
| CN106874904A (en) * | 2017-01-09 | 2017-06-20 | 北京大学深圳研究生院 | A kind of car plate picture antidote and device |
| CN109190621A (en) * | 2018-09-06 | 2019-01-11 | 东北大学 | Greasy weather automatic license plate recognition method |
| CN109191859A (en) * | 2018-09-28 | 2019-01-11 | 徐常柏 | A kind of parking lot Vehicle License Plate Recognition System |
| CN109829460A (en) * | 2019-04-02 | 2019-05-31 | 安徽省安瑞机电科技有限公司 | A kind of License Plate Character Segmentation recognition methods and its system applied to parking management system |
| CN110414507A (en) * | 2019-07-11 | 2019-11-05 | 和昌未来科技（深圳）有限公司 | Licence plate recognition method, device, computer equipment and storage medium |
| CN110458161A (en) * | 2019-07-15 | 2019-11-15 | 天津大学 | A kind of mobile robot doorplate location method of combination deep learning |
| CN110659638A (en) * | 2019-09-24 | 2020-01-07 | 北京精英路通科技有限公司 | License plate recognition method and device, computer equipment and storage medium |
| US10706330B2 (en) | 2015-10-01 | 2020-07-07 | Intellivision Technologies Corp | Methods and systems for accurately recognizing vehicle license plates |
| CN112116723A (en) * | 2020-08-10 | 2020-12-22 | 中标慧安信息技术股份有限公司 | License plate recognition-based non-inductive charging method and device |
| CN113554033A (en) * | 2021-09-18 | 2021-10-26 | 深圳市一号互联科技有限公司 | Text recognition method, device and system of intelligent text robot |
| CN114141025A (en) * | 2021-12-03 | 2022-03-04 | 江苏越捷智能停车系统有限公司 | Unattended intelligent parking lot system |
| CN114821078A (en) * | 2022-05-05 | 2022-07-29 | 北方工业大学 | License plate recognition method and device, electronic equipment and storage medium |
| US11587327B2 (en) | 2015-10-01 | 2023-02-21 | Intellivision Technologies Corp | Methods and systems for accurately recognizing vehicle license plates |
| CN116311215A (en) * | 2023-05-22 | 2023-06-23 | 成都运荔枝科技有限公司 | License plate recognition method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6473517B1 (en) * | 1999-09-15 | 2002-10-29 | Siemens Corporate Research, Inc. | Character segmentation method for vehicle license plate recognition |
| CN101408933A (en) * | 2008-05-21 | 2009-04-15 | 浙江师范大学 | Method for recognizing license plate character based on wide gridding characteristic extraction and BP neural network |
| CN101789080A (en) * | 2010-01-21 | 2010-07-28 | 上海交通大学 | Detection method for vehicle license plate real-time positioning character segmentation |
| CN102708356A (en) * | 2012-03-09 | 2012-10-03 | 沈阳工业大学 | Automatic license plate positioning and recognition method based on complex background |
-
2012
- 2012-12-06 CN CN201210520404.4A patent/CN103065138B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6473517B1 (en) * | 1999-09-15 | 2002-10-29 | Siemens Corporate Research, Inc. | Character segmentation method for vehicle license plate recognition |
| CN101408933A (en) * | 2008-05-21 | 2009-04-15 | 浙江师范大学 | Method for recognizing license plate character based on wide gridding characteristic extraction and BP neural network |
| CN101789080A (en) * | 2010-01-21 | 2010-07-28 | 上海交通大学 | Detection method for vehicle license plate real-time positioning character segmentation |
| CN102708356A (en) * | 2012-03-09 | 2012-10-03 | 沈阳工业大学 | Automatic license plate positioning and recognition method based on complex background |
Non-Patent Citations (1)
| Title |
|---|
| 李雄 等: "几何特征形态学车牌识别系统研究", 《计算机仿真》 * |
Cited By (40)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103984949A (en) * | 2014-06-11 | 2014-08-13 | 四川九洲电器集团有限责任公司 | License plate positioning method and system based on high and low cap transformation and connected domain |
| CN103984949B (en) * | 2014-06-11 | 2017-07-21 | 四川九洲电器集团有限责任公司 | License plate locating method and its system based on high bot-hat transformation and connected domain |
| CN104036262A (en) * | 2014-06-30 | 2014-09-10 | 南京富士通南大软件技术有限公司 | Method and system for screening and recognizing LPR license plate |
| CN104036262B (en) * | 2014-06-30 | 2017-09-19 | 南京富士通南大软件技术有限公司 | A kind of method and system of LPR car plates screening identification |
| CN104156731A (en) * | 2014-07-31 | 2014-11-19 | 成都易默生汽车技术有限公司 | License plate recognition system based on artificial neural network and method |
| CN104156731B (en) * | 2014-07-31 | 2017-08-25 | 成都易默生汽车技术有限公司 | Vehicle License Plate Recognition System and method based on artificial neural network |
| CN104156704A (en) * | 2014-08-04 | 2014-11-19 | 胡艳艳 | Novel license plate identification method and system |
| CN104239878B (en) * | 2014-08-30 | 2017-05-10 | 电子科技大学 | License plate character recognition method based on probability extremum search |
| CN104239878A (en) * | 2014-08-30 | 2014-12-24 | 电子科技大学 | License plate character recognition method based on probability extremum search |
| CN104239864A (en) * | 2014-09-16 | 2014-12-24 | 哈尔滨恒誉名翔科技有限公司 | Freight car number identification system based on image processing |
| CN104408438A (en) * | 2014-12-10 | 2015-03-11 | 黑龙江真美广播通讯器材有限公司 | Truck number identification system |
| CN105184227A (en) * | 2015-08-13 | 2015-12-23 | 安徽超远信息技术有限公司 | Traffic signal control system with automobile high beam light detection function |
| CN106469267B (en) * | 2015-08-20 | 2019-12-17 | 深圳市腾讯计算机系统有限公司 | Verification code sample collection method and system |
| CN106469267A (en) * | 2015-08-20 | 2017-03-01 | 深圳市腾讯计算机系统有限公司 | A kind of identifying code sample collection method and system |
| CN105447487A (en) * | 2015-08-27 | 2016-03-30 | 中山大学 | Evaluation method and system for vehicle license plate identification system |
| US11587327B2 (en) | 2015-10-01 | 2023-02-21 | Intellivision Technologies Corp | Methods and systems for accurately recognizing vehicle license plates |
| US10706330B2 (en) | 2015-10-01 | 2020-07-07 | Intellivision Technologies Corp | Methods and systems for accurately recognizing vehicle license plates |
| CN105139659A (en) * | 2015-10-09 | 2015-12-09 | 东方网力科技股份有限公司 | Vehicle license plate recognition method and device |
| CN106127707A (en) * | 2016-06-23 | 2016-11-16 | 南京大学金陵学院 | A kind of processing method of supersonic blood image |
| CN106127707B (en) * | 2016-06-23 | 2019-04-16 | 南京大学金陵学院 | A kind of processing method of supersonic blood image |
| CN106127206A (en) * | 2016-06-28 | 2016-11-16 | 北京智芯原动科技有限公司 | The vertical angles detection method of a kind of car plate and device |
| CN106127206B (en) * | 2016-06-28 | 2019-04-05 | 北京智芯原动科技有限公司 | A kind of vertical angles detection method and device of license plate |
| CN106295528A (en) * | 2016-08-01 | 2017-01-04 | 长安大学 | A kind of vehicle checking method based on multi-part spatial relation GMM modeling |
| CN106874904A (en) * | 2017-01-09 | 2017-06-20 | 北京大学深圳研究生院 | A kind of car plate picture antidote and device |
| CN109190621A (en) * | 2018-09-06 | 2019-01-11 | 东北大学 | Greasy weather automatic license plate recognition method |
| CN109191859A (en) * | 2018-09-28 | 2019-01-11 | 徐常柏 | A kind of parking lot Vehicle License Plate Recognition System |
| CN109829460A (en) * | 2019-04-02 | 2019-05-31 | 安徽省安瑞机电科技有限公司 | A kind of License Plate Character Segmentation recognition methods and its system applied to parking management system |
| CN110414507A (en) * | 2019-07-11 | 2019-11-05 | 和昌未来科技（深圳）有限公司 | Licence plate recognition method, device, computer equipment and storage medium |
| CN110458161A (en) * | 2019-07-15 | 2019-11-15 | 天津大学 | A kind of mobile robot doorplate location method of combination deep learning |
| CN110458161B (en) * | 2019-07-15 | 2023-04-18 | 天津大学 | Mobile robot doorplate positioning method combined with deep learning |
| CN110659638B (en) * | 2019-09-24 | 2023-04-07 | 北京精英路通科技有限公司 | License plate recognition method and device, computer equipment and storage medium |
| CN110659638A (en) * | 2019-09-24 | 2020-01-07 | 北京精英路通科技有限公司 | License plate recognition method and device, computer equipment and storage medium |
| CN112116723A (en) * | 2020-08-10 | 2020-12-22 | 中标慧安信息技术股份有限公司 | License plate recognition-based non-inductive charging method and device |
| CN112116723B (en) * | 2020-08-10 | 2021-06-22 | 中标慧安信息技术股份有限公司 | License plate recognition-based non-inductive charging method and device |
| CN113554033A (en) * | 2021-09-18 | 2021-10-26 | 深圳市一号互联科技有限公司 | Text recognition method, device and system of intelligent text robot |
| CN114141025A (en) * | 2021-12-03 | 2022-03-04 | 江苏越捷智能停车系统有限公司 | Unattended intelligent parking lot system |
| CN114821078B (en) * | 2022-05-05 | 2023-03-14 | 北方工业大学 | License plate recognition method and device, electronic equipment and storage medium |
| CN114821078A (en) * | 2022-05-05 | 2022-07-29 | 北方工业大学 | License plate recognition method and device, electronic equipment and storage medium |
| CN116311215A (en) * | 2023-05-22 | 2023-06-23 | 成都运荔枝科技有限公司 | License plate recognition method |
| CN116311215B (en) * | 2023-05-22 | 2023-11-17 | 成都运荔枝科技有限公司 | License plate recognition method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103065138B (en) | 2015-07-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103065138B (en) | Recognition method of license plate number of motor vehicle | |
| CN110178167B (en) | Intersection violation video identification method based on cooperative relay of cameras | |
| CN111178291B (en) | Parking payment system and parking payment method | |
| CN102708356B (en) | Automatic license plate positioning and recognition method based on complex background | |
| CN103324930B (en) | A kind of registration number character dividing method based on grey level histogram binaryzation | |
| CN110969160B (en) | License plate image correction and recognition method and system based on deep learning | |
| CN101334836B (en) | License plate positioning method incorporating color, size and texture characteristic | |
| CN102509098B (en) | Fisheye image vehicle identification method | |
| Jusoh et al. | Application of freeman chain codes: An alternative recognition technique for Malaysian car plates | |
| Roy et al. | License plate detection and character recognition system for commercial vehicles based on morphological approach and template matching | |
| CN103116751A (en) | Automatic license plate character recognition method | |
| CN106650553A (en) | License plate recognition method and system | |
| CN105005771A (en) | Method for detecting full line of lane based on optical flow point locus statistics | |
| CN103903018A (en) | Method and system for positioning license plate in complex scene | |
| CN106815583B (en) | Method for positioning license plate of vehicle at night based on combination of MSER and SWT | |
| CN101520841A (en) | Real-time and anti-interference method for positioning license plate in high-definition TV video | |
| CN103824452A (en) | Lightweight peccancy parking detection device based on full view vision | |
| CN103268489A (en) | Motor vehicle plate identification method based on sliding window searching | |
| CN102419820A (en) | Method for rapidly detecting car logo in videos and images | |
| CN111695565A (en) | Automobile mark accurate positioning method based on road barrier fuzzy image | |
| CN114387591A (en) | License plate recognition method, system, equipment and storage medium | |
| CN103927512B (en) | vehicle identification method | |
| Ismail | License plate Recognition for moving vehicles case: At night and under rain condition | |
| CN103295003A (en) | Vehicle detection method based on multi-feature fusion | |
| CN106203273A (en) | The lane detection system of multiple features fusion, method and senior drive assist system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C53 | Correction of patent of invention or patent application | ||
| CB02 | Change of applicant information |
Address after: 830000 the Yellow River Road, the Xinjiang Uygur Autonomous Region, No. 191, No. Applicant after: CHINACCS INFORMATION INDUSTRY CO., LTD. Address before: 830000 the Yellow River Road, the Xinjiang Uygur Autonomous Region, No. 191, No. Applicant before: Xinjiang Public Information Industry Co., Ltd. |
|
| CB03 | Change of inventor or designer information |
Inventor after: Li Weixin Inventor after: Shu Hongxin Inventor after: Wang Xiuying Inventor after: Ge Li Inventor after: Wen Jiangtao Inventor after: He Guangming Inventor before: Li Weixin Inventor before: Ge Li Inventor before: Wen Jiangtao |
|
| COR | Change of bibliographic data |
Free format text: CORRECT: INVENTOR; FROM: LI WEIXIN YI LI WEN JIANGTAO TO: LI WEIXIN SHU HONGXIN WANG XIUYING YI LI WEN JIANGTAO HE GUANGMING Free format text: CORRECT: APPLICANT; FROM: GONGZHONG INFORMATION INDUSTRY CO., LTD., XINJIANG TO: CHINA COMSERVICE PUBLIC INFORMATION INDUSTRY CO., LTD. |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |