CN103116751A - Automatic license plate character recognition method - Google Patents
Automatic license plate character recognition method Download PDFInfo
- Publication number
- CN103116751A CN103116751A CN2013100281068A CN201310028106A CN103116751A CN 103116751 A CN103116751 A CN 103116751A CN 2013100281068 A CN2013100281068 A CN 2013100281068A CN 201310028106 A CN201310028106 A CN 201310028106A CN 103116751 A CN103116751 A CN 103116751A
- Authority
- CN
- China
- Prior art keywords
- license plate
- character
- plate
- image
- characters
- 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
Landscapes
- Character Input (AREA)
- Character Discrimination (AREA)
Abstract
The invention discloses an automatic license plate character recognition method. The method includes following steps: inputting color vehicle images; pretreating the images; choosing different algorithms for license plate positioning under different light conditions; subjecting a license plate to horizontal tilt correction and vertical tilt correction; adopting a method based on cluster connection and vertical projection to perform character segmentation to the license plate obtained in step d; using an improved template matching method to perform character recognition to characters obtained in step e; and outputting recognition results of step f. By the automatic license plate character recognition method, using performance of a whole automatic license plate character recognition system is evidently improved, in the step of license plate positioning, license plate positioning can be realized by choosing different algorithms under different light conditions and utilizing characteristics of license plate areas, and accurate positioning of license plate area positions can be guaranteed even under the circumstance that interferences occur.
Description
Technical field
The invention belongs to computer vision and area of pattern recognition, relate to the multi-door subjects such as pattern-recognition, Digital Image Processing, artificial intelligence, computer science, be specifically related to a kind of Method of Automatic Recognition for Character of Lcecse Plate.
Background technology
Along with the expansion of city size, the vehicles number of the inside, city increases sharply, and traditional labor management traffic can not adapt to this variation.Therefore, intelligent transportation system by large-scale application in urban traffic control and the scheduling in.Wherein car plate identification (Vehicle License Plate Recognition, VLPR) is the core in intelligent transportation system.
Car plate identification is the important component part in modern intelligent transportation system, uses very extensive.It is take technology such as pattern-recognition, computer vision, Digital Image Processing as the basis.Can realize the functions such as magnitude of traffic flow control index measurement, vehicle location, high way super speed robotization supervision, break in traffic rules and regulations candid photograph, toll station and parking lot fee collection management by a series of processing to the car plate data.The automatic identification of license plate prevents traffic jam for safeguarding traffic safety and urban public security, realizes that the traffic automation management has great significance.
Therefore, need a kind of Method of Automatic Recognition for Character of Lcecse Plate.
Summary of the invention
Goal of the invention: the present invention is directed to the defective that prior art exists aspect car plate identification, a kind of Method of Automatic Recognition for Character of Lcecse Plate is provided.
Technical scheme: for solving the problems of the technologies described above, a kind of Method of Automatic Recognition for Character of Lcecse Plate of the present invention adopts following technical scheme:
A kind of Method of Automatic Recognition for Character of Lcecse Plate comprises the following steps:
A, input color vehicle image;
B, the colored vehicle image that step a is obtained carry out pre-service;
C, by day in the reasonable situation of light, the colored vehicle image that adopts color-based point that the algorithm of locating license plate of vehicle of search and mathematical morphology is obtained step b carries out car plate and locates; By day in light situation at not good or night, adopt colored vehicle image that the algorithm of locating license plate of vehicle of intensity-based image obtains step b to carry out the car plate location; The present invention is take illuminance 8000lx as separation, and when illuminance during greater than 8000lx, the colored vehicle image that adopts color-based point that the algorithm of locating license plate of vehicle of search and mathematical morphology is obtained step b carries out car plate and locates; When illuminance during less than 8000lx, adopt colored vehicle image that the algorithm of locating license plate of vehicle of intensity-based image obtains step b to carry out the car plate location.
D, the car plate that respectively step c location is obtained carry out the horizontal tilt rectification and vertical bank is corrected;
E, employing are carried out Character segmentation based on the method for cluster connected sum vertical projection to the car plate that steps d obtains;
F, the improved template matching method of use obtain character to step e and carry out character recognition;
The recognition result of g, output step f.
Beneficial effect: Method of Automatic Recognition for Character of Lcecse Plate of the present invention, improve the usability of whole characters on license plate automatic recognition system significantly.In described car plate positioning step, the present invention can utilize the characteristics of license plate area, selects algorithms of different to carry out car plate under the different light rays condition and locates, and guarantees in the situation that accurately positioning licence plate regional location occurs disturbing also.
Further, pre-service described in step b comprises image smoothing method and/or filtering and noise reduction method.
Further, utilize car plate background color and the difference of characters on license plate color on gray scale to find out the car plate region based on the algorithm of locating license plate of vehicle of search color dot pair and mathematical morphology described in step c.
Further, the character color of supposing car plate is the A look, and the background color of car plate is the B look, describedly described in step c comprises the following steps based on the algorithm of locating license plate of vehicle of search color dot pair with mathematical morphology:
(1) coloured image that contains vehicle is carried out gray processing and binary conversion treatment, calculate the maximum variance threshold value, obtain corresponding bianry image;
(2) open up a region of memory in calculator memory, be used for the storage binary image data, the length and width of bianry image equate with the length and width of coloured image, and the value initialization of each pixel is 255;
(3) the A colour vegetarian refreshments in the scanning bianry image, if find B colour vegetarian refreshments in left side, coloured image relevant position, think the initial frontier point that this A colour vegetarian refreshments is characters on license plate and car plate background color, i.e. the initial pixel of characters on license plate, the position of this point of mark;
(4) continue the A colour vegetarian refreshments of scanning bianry image, if B colour vegetarian refreshments is found on the right side in the coloured image relevant position, think the termination frontier point that this A colour vegetarian refreshments is characters on license plate and car plate background color, i.e. the position of this point of the termination pixel of characters on license plate, and mark;
(5) zone is characters on license plate pixel zone if the termination pixel phase difference of the initial pixel of characters on license plate and characters on license plate within 50 pixels, is thought this piece, and in the image of step 3 with these points of black pixel point mark;
(6) complete scan entire image is until find out all A looks in image and B look color dot pair;
(7) using the two-value opening operation to process the color dot that step (6) is obtained processes;
(8) use the corrosion operation to process later color to processing to the two-value opening operation, further enlarge the right scope of color dot;
(9) use expansive working to color dot to processing, obtain a slice continuous pixel zone, this piece zone is exactly the car plate candidate regions.In the situation that light is more abundant, adopt color-based point to searching for and the algorithm of locating license plate of vehicle of mathematical morphology, this algorithm has a very high locating accuracy in the situation that light is better;
Further, described in step c, the algorithm of locating license plate of vehicle of intensity-based image comprises the following steps:
A, at first to the coloured image that contains vehicle is carried out gray processing and binary conversion treatment;
B, adopt the method every three line scanning delegation, the line number r that mark monochrome pixels point change frequency is maximum
m, and write down change frequency m;
C, with r
mBe the basis, statistics up and down 3 row, judge its monochrome pixels point change frequency whether with r
mApproach, the scope of the change frequency of the monochrome pixels point of license plate area is from 0.75m to 1.25m, and the number of times that namely in car plate, black and white changes is within a range of stability, but not the monochrome pixels change frequency of license plate area is unsettled;
D, repeating step b and step c obtain the up-and-down boundary of car plate, thereby determine the car plate candidate regions.
In the situation that light is not good, adopt improved algorithm of locating license plate of vehicle based on textural characteristics, this algorithm every 3 line scannings once can improve the real-time of system greatly.
Further, in steps d, the rectification of the horizontal tilt rectification of car plate and vertical bank is adopted respectively based on the horizontal tilt correction method of car plate marginal point ordinate variance minimum and carries out based on the whole vertical skew correction method of the car plate of single character vertical tilt angle degree.In license plate sloped aligning step, advanced driving board binaryzation, then carry out correction on different directions.
Further, first find out connected domain in image based on the method for cluster connected sum vertical projection described in step e, then by the priori of car plate, connected domain is screened.This algorithm is a kind of accuracy rate that can improve Character segmentation, and robustness is high algorithm again.
Further, described in step e, the method based on cluster connected sum vertical projection comprises the following steps:
(1) license plate image is carried out binary conversion treatment.
(2) image after processing is by line scanning, as two pel spacings from d＜=d
Const, think that these two pixels belong to a class, namely two pixels in character;
What (3) screening obtained is all kinds of, removes height less than the class of Height/2, is 7 characters and left and right side frame through obtaining class after screening;
(4) all kinds of reference positions according to from left to right name placement.License plate binary adopts modified local binarization method, experimental results show that the method in the situation that not good also can be good at of light goes characters on license plate and background to separate;
Further, d is defined as:
d(A,B)＝|x
1-x
2|+|y
1-y
2|，
The d of Chinese character
ConstValue is 2 or 3, the d of English character and numeral
ConstValue is 2.
Further, in step f, improved template matching method comprises the following steps:
(1) set up the matching template storehouse;
(2) characters on license plate is carried out the refinement computing, the stroke width that makes character is a pixel wide;
(3) the characters on license plate filtering discrete noise point that step (2) is obtained;
(4) will carry out normalized through refinement and the later characters on license plate of filtering discrete noise point, normalized is carried out later on the image inversion processing, and the car plate background is become white, and character becomes black;
(5) the characters on license plate image that step (4) is obtained and matching template storehouse compare, and obtain matching template;
(6) scanning car plate character picture, when running into a character pixels, seek nearest character pixels point in the pixel region scope of matching template 5 * 5, and calculate the size of nearest character pixels point and matching template minimum distance, calculate the minor increment sum of characters on license plate pixel and matching stencil, find out the minor increment sum of characters on license plate and matching template;
(7) scan matching masterplate image, when running into a character pixels, seek nearest character pixels point in the pixel region scope of characters on license plate 5 * 5, and calculate the size of nearest character pixels point and characters on license plate minimum distance, calculate the minor increment sum of matching stencil and characters on license plate, find out the minor increment sum of matching stencil and characters on license plate;
(8) compare the minor increment sum of (6) and (7), the corresponding templates that gets the small value is as the matching result of characters on license plate.
(9) according to above-mentioned steps, characters on license plate is mated one by one, finally obtain license board information.In described character recognition step, template matching method is improved, further the characters on license plate that obtains after cutting apart is carried out thinning processing, significantly improved and be matched to power.
Description of drawings
The process flow diagram of Fig. 1 Method of Automatic Recognition for Character of Lcecse Plate of the present invention.
Embodiment
Below in conjunction with the drawings and specific embodiments, further illustrate the present invention, should understand these embodiment only is used for explanation the present invention and is not used in and limits the scope of the invention, after having read the present invention, those skilled in the art all fall within the application's claims limited range to the modification of the various equivalent form of values of the present invention.
As shown in Figure 1, image or video interception that at first Method of Automatic Recognition for Character of Lcecse Plate of the present invention is taken by video camera, convert digitized image input computing machine (step 101) to, then input picture is carried out pre-service (step 102), then according to different light condition, use diverse ways to orient car plate position (step 103); Next by the slant correction (step 104) of the digital image processing methods such as binaryzation, horizontal tilt correction and vertical skew correction realization to license plate image; Again independent character is split (step 105) from license plate image; Technology by pattern-recognition realizes the automatic identification (step 106) to characters on license plate at last.
China's car plate mainly contains these four kinds of wrongly written or mispronounced characters of the blue end, yellow end surplus, black matrix wrongly written or mispronounced character, the dark red word of white background.The below is described in detail Method of Automatic Recognition for Character of Lcecse Plate of the present invention as an example of wrongly written or mispronounced character car plate of the blue end example.
Car plate location is the difficult point in Vehicle License Recognition System always, the accuracy of a large amount of interference and factor all can the affect car plate location such as the light irradiation is strong and weak, car plate is reflective in license plate image.The present invention proposes the algorithm of locating license plate of vehicle that two kinds of algorithm of locating license plate of vehicle combine: in the reasonable situation of light, adopt color-based point to the algorithm of locating license plate of vehicle of search and mathematical morphology by day; Adopt by day a kind of algorithm of locating license plate of vehicle of novel intensity-based image in light situation at not good or night.Algorithm of locating license plate of vehicle in the present invention needs the support of light sensor, and the intensity of the external environment condition light that receives by light sensor decides uses carry out the car plate location for which kind of algorithm.The present invention is take illuminance 8000lx as separation, when illuminance during greater than 8000lx, adopts color-based point to the algorithm of locating license plate of vehicle of search and mathematical morphology, the colored vehicle image that obtains to be carried out car plate and locates; When illuminance during less than 8000lx, the algorithm of locating license plate of vehicle that adopts the intensity-based image carries out the car plate location to the colored vehicle image that obtains.
Owing to may containing the color close with the car plate color on vehicle body, add the reasons such as oxidation causes that car plate fades, these disturb the precision that all may reduce the car plate location.The algorithm of locating license plate of vehicle based on search color dot pair and mathematical morphology that the present invention proposes utilizes car plate background color and the difference of characters on license plate color on gray scale to find out the car plate region, has good antijamming capability.Concrete steps are as follows:
(1) coloured image that contains vehicle is carried out gray processing and binary conversion treatment, calculate the maximum variance threshold value, obtain corresponding bianry image;
(2) open up a region of memory in calculator memory, be used for the view data of 8 of storages, the image length and width equate with coloured image, and the value initialization of each pixel is 255;
(3) the white pixel point in the scanning bianry image, if find the blue pixel point in left side, coloured image relevant position, think the initial frontier point that this white pixel point is characters on license plate and car plate background color, i.e. the initial pixel of characters on license plate, the position of this point of mark;
(4) continue the white pixel point of scanning bianry image, if the blue pixel point is found on the right side in the coloured image relevant position, think the termination frontier point that this white pixel point is characters on license plate and car plate background color, i.e. the position of this point of the termination pixel of characters on license plate, and mark;
(5) zone is characters on license plate pixel zone if the termination pixel phase difference of the initial pixel of characters on license plate and characters on license plate within 50 pixels, is thought this piece, and in the image of step 3 with these points of black pixel point mark;
(6) complete scan entire image is until all blue white colour points of finding out in image are right;
(7) using two-value to open the operation processing processes color dot.
(8) use the corrosion operation to split two-value and open the later color of operation to processing, carry out continuously ten times, further enlarge the right scope of color dot.
(9) use expansive working to continue color dot to obtain the pixel zone of a slice basic continous to processing, this piece zone is exactly the car plate candidate regions.
Notice that the blueness of using in the present invention is a vague definition, the blueness in actual license plate has mazarine and light blue, also will consider due to reasons such as oxidations, and the background color of car plate can shoal, so can only define a general qualifications for blueness.This condition is rule of thumb set, and can change according to different this conditions of light situation.As long as the pixel in coloured image satisfies this condition, can think that this pixel is the blue pixel point.Concrete, if this pixel satisfies following three conditions: (1) B〉1.5*R; (2) B〉1.5*G; (3) B〉100 just think that this pixel is the blue pixel point.Equally, white is also a fuzzy concept.If a pixel satisfies following four conditions: (1) R＜0.4*S; (2) G＜0.4*S; (3) B＜0.4*S; (4) S〉200 just think that this pixel is the white pixel point.S=R+G+B wherein.
The car plate candidate regions that obtain this moment may be the car plate region, may be also noise or interference region.Can utilize the priori of car plate to remove pseudo-license plate area, thereby obtain real license plate area.Usually the length of license plate area and wide all should be greater than certain pixel value, and the ratio that grows tall of car plate should be between 2: 1 to 4: 1.
In not good or night situation, the color characteristic of car plate is not obvious at light, if still adopt the color-based point may can not find color dot pair in image to searching algorithm, causes locating unsuccessfully.The present invention is in the situation that the car plate of a kind of intensity-based image of the not good employing of light is located new algorithm.At first image is carried out gray processing and binary conversion treatment.Because automotive license plate generally is suspended on bottom of car, so the present invention adopts searching method from the bottom up, not only can improve search speed like this, also can improve the success ratio of search.Employing is every the method for three line scanning delegation, the line number r that mark monochrome pixels point change frequency is maximum
m, and write down change frequency m.The car plate height about 50 pixels, in the horizontal scanning process of license plate area, the many row of monochrome pixels point change frequency will occur every 3 row in general, and 17 left and right appear in such guild continuously, write down these line numbers r
i(i=1,2 ...), with r
iBe the basis, statistics up and down 3 row, judge its monochrome pixels point change frequency whether with r
iSimilar, rule of thumb the scope of the change frequency of the monochrome pixels point of license plate area is from 0.75m to 1.25m as can be known, so the number of times that in car plate, black and white changes is within a range of stability, but not the monochrome pixels change frequency of license plate area is unsettled.The up-and-down boundary of orienting car plate by top algorithm is the height of characters on license plate substantially, but will be with location up-and-down boundary expanded range out in actual treatment, do like this is in order to leave certain processing space to follow-up license plate area slant correction, to avoid the car plate useful part is excised.
The difference that has fixedly caused shooting angle due to the position of camera, more or less there is the situation of inclination in the license plate image that obtains after the location.Tilting to affect follow-up License Plate Character Segmentation, causes segmentation errors, thus to carry out slant correction to license plate image before Character segmentation, for the character of back can accurately be cut apart ready.License plate slopedly be divided into three kinds of patterns: horizontal direction tilts, vertical direction tilts, level stack vertical direction tilts, therefore license plate sloped correction can be carried out from horizontal and vertical directions, usually first carrying out horizontal tilt proofreaies and correct, determine the up-and-down boundary of characters on license plate, carry out at last vertical skew correction.The car plate horizontal tilt correcting algorithm that is based on vertical edge spot projection variance minimum that the present invention uses and based on the car plate vertical skew correction algorithm of horizontal edge point variance minimum.
Carry out needing the gray scale license plate image is carried out binary conversion treatment before license plate sloped correction.The present invention adopts improved local binarization method, and it is in the situation that uneven illumination and car plate have stained can well characters on license plate and car plate background being separated.Suppose that (x, y) is the pixel of wanting binaryzation, its close region is the window of a w * w, and gray-scale value is f (x, y), calculates the threshold value of each pixel:
The α here is an empirical value, and span is between 0.40 and 0.65.
By statistics obtain the character pixels point in the shared ratio of whole license plate image in 30% left and right, if consider noise factor, this reaches 35% left and right than regular meeting.Therefore a global threshold K can be set.Suppose h (x, y) be license plate image at the statistical value of a certain gray-scale value, the definition K be:
If we only carry out binaryzation with global threshold K to car plate, be difficult to the negative effect that elimination uneven illumination or background acute variation are brought binaryzation.Therefore must carry out binary conversion treatment to license plate image in conjunction with threshold k and binaryzation method and just can reach desirable effect, concrete steps are as follows:
(1) calculate global threshold K.
(2) obtain the local threshold of all pixels by Bernsen binaryzation method.
(3) get between global threshold K and local threshold value the greater as the threshold value of pixel.
Therefore some discrete noises may appear in image in the binaryzation process, remove the noise process that is absolutely necessary.Filter method commonly used such as mean filter and medium filtering are not suitable for the processing character image, because may the filtering character pixels in the process of filtering.The present invention uses following algorithm:
(1) by the line scanning image, when finding a white pixel point, seek the quantity of white pixel point in its neighborhood of 3 * 3, consider 8 pixels adjacent with white pixel.
(2) set a threshold value, if around it, white pixel point number is greater than this threshold value, think that this point is not the discrete noise point, remove otherwise it is used as noise spot.In the present invention, the threshold value value is 5.
(3) scanning entire image is removed all discrete pixels.
Before carrying out the horizontal tilt correction, needs first carry out rim detection to the license plate image of binaryzation.Rim detection utilizes the zero crossing information of image first order derivative or second derivative that the basic foundation of judgement marginal point is provided.What the present invention used is the Sobel operator.
By obtaining all marginal points in image after rim detection, the central point that makes these marginal points is M (m
x, m
y), can be by asking for following formula:
X in formula
iWith y
iHorizontal ordinate and ordinate for the car plate marginal point.
Car plate edge image place coordinate origin is moved on to the M point, at this moment car plate marginal point ordinate average
After coordinate carried out conversion, variation had occured in the coordinate of car plate marginal point, wherein:
u
i＝x
i-m
x,v
i＝y
i-m
y
Suppose that B (u, v) is a bit in the two-dimensional direct angle coordinate system, this point is β with the angle of x axle, and its ordinate is v.The B point around clockwise direction rotation alpha angle, center to putting B'(u', v '), the some B' ordinate be v ', can obtain following derivation:
Pass through the later variance of coordinate transform and marginal point rotation as follows:
According to top derivation as can be known the rotation angle when marginal point ordinate variance obtains minimum value be the horizontal tilt angle, be exactly to σ here
2Differentiate, the α angle when derivative is zero is exactly the horizontal tilt angle.
Following formula can be got following formula through conversion:
So car plate inclination angle in the horizontal direction is:
Car plate is proofreaied and correct along the horizontal direction horizontal tilt that α completes car plate that turns clockwise.
In proofreading and correct later license plate image through horizontal tilt, car plate upper and lower side frame and left and right side frame, bumper etc. produce the characters on license plate vertical skew correction and disturb, and affect the accuracy of vertical skew correction.Therefore must accurately determine the up-and-down boundary of character, the up-and-down boundary of character both for vertical skew correction has reduced interference, has reduced interference for Character segmentation again after determining.
Use the Sobel vertical operator to ask for the vertical edge of bianry image, then vertical edge is carried out projection on horizontal direction, adding up the white pixel of every row counts out, if greater than the threshold value of setting, just think that characters on license plate is regional, if find first white pixel point quantity statistical value less than the row of threshold value, think that this row may be coboundary or the lower boundary of character zone.In actual application, often the up-and-down boundary of character is enlarged 2～3 pixels, doing like this is to proofread and correct for fear of horizontal tilt not exclusively to bring the character pixels loss.
Therefore the vertical bank angle of single character and the vertical bank angle of whole car plate are consistent, thereby can obtain by the vertical bank angle that calculates single characters on license plate whole car plate angle of inclination in vertical direction.For single character, when character did not tilt in the vertical direction, after binaryzation, character pixels put that projective distribution is the most intensive in vertical direction, and scope is minimum, homolographic projection point horizontal ordinate variance minimum; When character existed, character pixels put that drop shadow spread is wider in vertical direction, and homolographic projection point horizontal ordinate variance is larger.It is more severe that character tilts in the vertical direction, and the horizontal ordinate variance of subpoint is larger.So can carry out doing vertical projection after shear transformation to single character, when the horizontal ordinate variance obtains minimum value, corresponding shearing inclination be exactly the angle of inclination of single character.
If the character point number is M, the character pixels point coordinate is (x
i, y
i), i=1,2 ... M, through coordinate after shear transformation be transformed to (x '
i, y
i), the variance of subpoint horizontal ordinate is defined as:
Can obtain like this:
Wherein,
Make the horizontal ordinate variances sigma of character pixels point projection in vertical direction
2Obtain minimum value, need make σ
2Derivative to θ is 0, tries to achieve θ ' and is single character angle of inclination in vertical direction.The derivation of equation is as follows:
Because sec
2θ
0≠ 0, so have
Can obtain character and car plate inclined angle alpha in vertical direction is
License plate image has satisfied the requirement of License Plate Character Segmentation substantially through slant correction, next only needs the character in license plate area is cut out separately.The present invention proposes a kind of partitioning algorithm based on cluster connected sum vertical projection and complete this step operation.
First find out connected domain in image based on the Character Segmentation of License Plate of cluster analysis, then by the priori of car plate, connected domain screened, this algorithm can fine solution complex background condition under the character cutting problem.Specific algorithm is described below:
(1) license plate image is carried out binary conversion treatment.
(2) image after processing is by line scanning, as two pel spacings from d＜=d
Const, think that these two pixels belong to a class, namely two pixels in character.Wherein d is defined as:
d(A,B)＝|x
1-x
2|+|y
1-y
2|
The d of Chinese character
ConstThe d of value and English character, numeral
ConstValue is different.For (nWidth)/7 part d before image
ConstGet 3, do like this to solve the disconnected problem of Chinese character; To (nWidth*6)/7 part d after image
ConstGet 2.
What (3) screening obtained is all kinds of, removes height less than the class of Height/2, is 7 characters and left and right side frame through obtaining class after screening.
(4) if adhesion problems appears in character, the number of class less than 7 or certain class width much larger than the mean breadth of class.Divide processing with vertical projection method to these classes this moment, thereby find out the frontier point of two characters in middle the near zone searching vertical projection local minizing point of class.
(5) all kinds of reference positions according to from left to right name placement.
(6) number as fruit is 7, forwards (7) to.Otherwise think and mistake occurs when cutting apart.
(7) if the height of certain class is far longer than all the other 6 classes, may be the cause of car plate up and down rivet and characters on license plate adhesion, at this moment revise according to the mean value of all the other 6 class height; If the width of certain class is far longer than all the other 6 classes, may be the cause of the left and right side frame adhesion of the Far Left character of car plate or rightmost character and car plate, at this moment revise according to the mean value of all the other 6 class width.
Recognition of License Plate Characters is a kind of of character recognition, but and other character recognition (identifying as print character) difference, characters on license plate has himself characteristic: the character kind is limited; The characters on license plate quality is not as print character; Characters on license plate has certain rule, and namely first character is Chinese character, and second character is English character, and third and fourth character is English character or numerical character, and remaining character is numerical character.According to these characteristics, the present invention proposes the method in conjunction with thinning and minor increment template matches.Its basic thought is exactly that the characters on license plate after cutting apart is carried out refinement computing in binaryzation morphology, can solve like this because font style difference is brought the not high shortcoming of discrimination, and interference that can noise reduction, improve to a certain extent the discrimination of template matching method.Concrete steps are as follows:
(1) set up the matching template storehouse, template base comprises the abbreviation of each provinces and cities of China, 26 capitalization English letters, 10 arabic numeral, the width of the every pictures of template library is 20 pixels, be highly 40 pixels (need to prove that the car plate of identifying in the present invention is only the normal domestic car plate, does not comprise police and military license plate).
(2) characters on license plate is carried out the refinement computing, make the stroke width of character become a pixel wide.
(3) due to the noise spot that may occur isolating after refinement, these noise spots can exert an influence to final matching result, therefore in addition filtering.The thought of removing these discrete noise points in the present invention is the black pixel point in scan image, and the search black pixel point, if the black pixel point that finds is less than 2, think that this point is that filtering is put and given to discrete noise in its 3 * 3 close region.
(4) will carry out normalized through refinement and the later characters on license plate of noise remove, namely by image scaling, the characters on license plate image be become 20 pixels wide, the normal pictures that 40 pixels are high.The image inversion processing is carried out later in normalization, and the car plate background is become white, and character becomes black,
(5) characters on license plate image and template library are compared.Set first characters on license plate and only mate calculating with 32 provinces and cities' abbreviations, second characters on license plate only mates calculating with 26 capitalization English letters, the 3rd and the 4th characters on license plate and 10 arabic numeral also have the mothers of 26 capitalization English to mate calculating, latter two characters on license plate only and 10 arabic numeral mate calculating, do like this efficient that not only improves algorithm, can also improve the accuracy rate of identification.
(6) scanning car plate character picture when running into a character pixels, is sought nearest character pixels point, and is calculated the size of minimum distance in the setting range of one of the relevant position of template, scope in the present invention is set as 5 * 5 pixel region.Calculate the minor increment sum of characters on license plate pixel and institute's matching stencil, find out the minor increment sum of car plate and matching stencil.
(7) scan matching masterplate image when running into a character pixels, is sought nearest character pixels point, and is calculated the size of this minimum distance in the scope of one of the relevant position of characters on license plate, scope in the present invention is set as 5 * 5 pixel region.Calculate the minor increment sum of matching stencil and characters on license plate, find out the minor increment sum of matching stencil and characters on license plate.
(8) compare the minor increment sum of (6) and (7), the corresponding templates that gets the small value is as the matching result of characters on license plate.
(9) according to above-mentioned steps, 7 characters on license plate are mated one by one, finally obtain license board information.
Method of Automatic Recognition for Character of Lcecse Plate of the present invention improves the usability of whole characters on license plate automatic recognition system significantly.In described car plate positioning step, the present invention can utilize the characteristics of license plate area, selects algorithms of different to carry out car plate under the different light rays condition and locates, and guarantees in the situation that accurately positioning licence plate regional location occurs disturbing also.
Claims (10)
1. Method of Automatic Recognition for Character of Lcecse Plate comprises the following steps:
A, input color vehicle image;
B, the colored vehicle image that step a is obtained carry out pre-service;
C, by day in the reasonable situation of light, the colored vehicle image that adopts color-based point that the algorithm of locating license plate of vehicle of search and mathematical morphology is obtained step b carries out car plate and locates; By day in light situation at not good or night, adopt colored vehicle image that the algorithm of locating license plate of vehicle of intensity-based image obtains step b to carry out the car plate location;
D, the car plate that respectively step c location is obtained carry out the horizontal tilt rectification and vertical bank is corrected;
E, employing are carried out Character segmentation based on the method for cluster connected sum vertical projection to the car plate that steps d obtains;
F, the improved template matching method of use obtain character to step e and carry out character recognition;
The recognition result of g, output step f.
2. Method of Automatic Recognition for Character of Lcecse Plate as claimed in claim 1, is characterized in that, pre-service described in step b comprises image smoothing method and/or filtering and noise reduction method.
3. Method of Automatic Recognition for Character of Lcecse Plate as claimed in claim 1, it is characterized in that, utilize car plate background color and the difference of characters on license plate color on gray scale to find out the car plate region based on the algorithm of locating license plate of vehicle of search color dot pair and mathematical morphology described in step c.
4. Method of Automatic Recognition for Character of Lcecse Plate as claimed in claim 1, it is characterized in that, the character color of supposing car plate is the A look, and the background color of car plate is the B look, describedly described in step c comprises the following steps based on the algorithm of locating license plate of vehicle of search color dot pair with mathematical morphology:
(1) coloured image that contains vehicle is carried out gray processing and binary conversion treatment, calculate the maximum variance threshold value, obtain corresponding bianry image;
(2) open up a region of memory in calculator memory, be used for the storage binary image data, the length and width of bianry image equate with the length and width of coloured image, and the value initialization of each pixel is 255;
(3) the A colour vegetarian refreshments in the scanning bianry image, if find B colour vegetarian refreshments in left side, coloured image relevant position, think the initial frontier point that this A colour vegetarian refreshments is characters on license plate and car plate background color, i.e. the initial pixel of characters on license plate, the position of this point of mark;
(4) continue the A colour vegetarian refreshments of scanning bianry image, if B colour vegetarian refreshments is found on the right side in the coloured image relevant position, think the termination frontier point that this A colour vegetarian refreshments is characters on license plate and car plate background color, i.e. the position of this point of the termination pixel of characters on license plate, and mark;
(5) zone is characters on license plate pixel zone if the termination pixel phase difference of the initial pixel of characters on license plate and characters on license plate within 50 pixels, is thought this piece, and in the image of step 3 with these points of black pixel point mark;
(6) complete scan entire image is until find out all A looks in image and B look color dot pair;
(7) using the two-value opening operation to process the color dot that step (6) is obtained processes;
(8) use the corrosion operation to process later color to processing to the two-value opening operation, further enlarge the right scope of color dot;
(9) use expansive working to color dot to processing, obtain a slice continuous pixel zone, this piece zone is exactly the car plate candidate regions.
5. Method of Automatic Recognition for Character of Lcecse Plate as claimed in claim 1, is characterized in that, the algorithm of locating license plate of vehicle of the image of intensity-based described in step c comprises the following steps:
A, at first the coloured image that contains vehicle is carried out gray processing and binary conversion treatment;
B, adopt the method every three line scanning delegation, the line number r that mark monochrome pixels point change frequency is maximum
m, and write down change frequency m;
C, with r
mBe the basis, statistics up and down 3 row, judge its monochrome pixels point change frequency whether with r
mThe change frequency of row approaches, and the scope of the change frequency of the monochrome pixels point of license plate area is from 0.75m to 1.25m, and the number of times that namely in car plate, black and white changes is within a range of stability, but not the monochrome pixels change frequency of license plate area is unsettled;
D, repeating step b and step c obtain the up-and-down boundary of car plate, thereby determine the car plate candidate regions.
6. Method of Automatic Recognition for Character of Lcecse Plate as claimed in claim 1, it is characterized in that, in steps d, the horizontal tilt rectification of car plate and vertical bank are corrected and are adopted respectively based on the horizontal tilt correction method of car plate marginal point ordinate variance minimum and carry out based on the whole vertical skew correction method of the car plate of single character vertical tilt angle degree.
7. Method of Automatic Recognition for Character of Lcecse Plate as claimed in claim 1, is characterized in that, first finds out connected domain in image based on the method for cluster connected sum vertical projection described in step e, then by the priori of car plate, connected domain is screened.
8. Method of Automatic Recognition for Character of Lcecse Plate as claimed in claim 1, is characterized in that, the method based on cluster connected sum vertical projection described in step e comprises the following steps:
(1) license plate image is carried out binary conversion treatment.
(2) image after processing is by line scanning, as two pel spacings from d＜=d
Const, think that these two pixels belong to a class, namely two pixels in character;
What (3) screening obtained is all kinds of, removes height less than the class of Height/2, is 7 characters and left and right side frame through obtaining class after screening;
(4) all kinds of reference positions according to from left to right name placement.
9. Method of Automatic Recognition for Character of Lcecse Plate as claimed in claim 8, is characterized in that, d is defined as:
d(A,B)＝|x
1-x
2|+|y
1-y
2|，
The d of Chinese character
ConstValue is 2 or 3, the d of English character and numeral
ConstValue is 2.
10. Method of Automatic Recognition for Character of Lcecse Plate as claimed in claim 1, is characterized in that, in step f, improved template matching method comprises the following steps:
(1) set up the matching template storehouse;
(2) characters on license plate is carried out the refinement computing, the stroke width that makes character is a pixel wide;
(3) the characters on license plate filtering discrete noise point that step (2) is obtained;
(4) will carry out normalized through refinement and the later characters on license plate of filtering discrete noise point, normalized is carried out later on the image inversion processing, and the car plate background is become white, and character becomes black;
(5) the characters on license plate image that step (4) is obtained and matching template storehouse compare, and obtain matching template;
(6) scanning car plate character picture, when running into a character pixels, seek nearest character pixels point in the pixel region scope of matching template 5 * 5, and calculate the size of nearest character pixels point and matching template minimum distance, calculate the minor increment sum of characters on license plate pixel and matching stencil, find out the minor increment sum of characters on license plate and matching template;
(7) scan matching masterplate image, when running into a character pixels, seek nearest character pixels point in the pixel region scope of characters on license plate 5 * 5, and calculate the size of nearest character pixels point and characters on license plate minimum distance, calculate the minor increment sum of matching stencil and characters on license plate, find out the minor increment sum of matching stencil and characters on license plate;
(8) compare the minor increment sum of (6) and (7), the corresponding templates that gets the small value is as the matching result of characters on license plate.
(9) according to above-mentioned steps, characters on license plate is mated one by one, finally obtain license board information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310028106.8A CN103116751B (en) | 2013-01-24 | 2013-01-24 | A kind of Method of Automatic Recognition for Character of Lcecse Plate |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310028106.8A CN103116751B (en) | 2013-01-24 | 2013-01-24 | A kind of Method of Automatic Recognition for Character of Lcecse Plate |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103116751A true CN103116751A (en) | 2013-05-22 |
| CN103116751B CN103116751B (en) | 2016-07-06 |
Family
ID=48415123
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310028106.8A Expired - Fee Related CN103116751B (en) | 2013-01-24 | 2013-01-24 | A kind of Method of Automatic Recognition for Character of Lcecse Plate |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103116751B (en) |
Cited By (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103824078A (en) * | 2014-03-18 | 2014-05-28 | 厦门翼歌软件科技有限公司 | Complex scene multi-license plate positioning method |
| CN103886320A (en) * | 2014-01-10 | 2014-06-25 | 北京环境特性研究所 | Label automatic recognition method based on parallel support vector machine weighting |
| CN104102909A (en) * | 2014-07-23 | 2014-10-15 | 中科联合自动化科技无锡有限公司 | Vehicle characteristic positioning and matching method based on multiple-visual information |
| CN104399674A (en) * | 2014-09-11 | 2015-03-11 | 上海大学 | Smoke box brand visual inspection sorting system |
| CN104463134A (en) * | 2014-12-19 | 2015-03-25 | 深圳市捷顺科技实业股份有限公司 | License plate detection method and system |
| CN104463195A (en) * | 2014-11-08 | 2015-03-25 | 沈阳工业大学 | Printing style digital recognition method based on template matching |
| CN104636309A (en) * | 2015-02-06 | 2015-05-20 | 南京理工大学 | Matrix calculator based on machine vision and matrix identification method |
| CN104866849A (en) * | 2015-04-30 | 2015-08-26 | 天津大学 | Food nutrition label identification method based on mobile terminal |
| CN105279512A (en) * | 2015-10-22 | 2016-01-27 | 东方网力科技股份有限公司 | Tilt vehicle license plate recognition method and device |
| CN105447831A (en) * | 2015-12-14 | 2016-03-30 | 浙江宇视科技有限公司 | License plate image processing method and apparatus |
| CN105631445A (en) * | 2014-11-06 | 2016-06-01 | 通号通信信息集团有限公司 | Character recognition method and system for license plate with Chinese characters |
| CN105719488A (en) * | 2014-12-02 | 2016-06-29 | 杭州海康威视数字技术股份有限公司 | License plate recognition method and apparatus, and camera and system for license plate recognition |
| WO2016134666A1 (en) * | 2015-02-28 | 2016-09-01 | 华为技术有限公司 | Method and device for discovering friend through image matching |
| 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 |
| CN106485234A (en) * | 2016-10-21 | 2017-03-08 | 合肥哦走信息技术有限公司 | One kind is based on vehicle identification method in intelligent transportation system |
| CN106650732A (en) * | 2016-12-27 | 2017-05-10 | 深圳市捷顺科技实业股份有限公司 | License plate identification method and apparatus |
| CN106951896A (en) * | 2017-02-22 | 2017-07-14 | 武汉黄丫智能科技发展有限公司 | A kind of license plate image sloped correcting method |
| CN108596172A (en) * | 2018-04-11 | 2018-09-28 | 航天科技控股集团股份有限公司 | A kind of the vehicle binding device and method of car networking client |
| CN108830274A (en) * | 2018-05-03 | 2018-11-16 | 南京邮电大学 | Characters on license plate detects positioning and recognition methods automatically under a kind of complex background environment |
| CN108830314A (en) * | 2018-06-06 | 2018-11-16 | 西安纸贵互联网科技有限公司 | A kind of sxemiquantitative recognition methods of external diagnosis reagent and device |
| CN108985137A (en) * | 2017-06-02 | 2018-12-11 | 杭州海康威视数字技术股份有限公司 | A kind of licence plate recognition method, apparatus and system |
| CN108968991A (en) * | 2018-05-08 | 2018-12-11 | 平安科技（深圳）有限公司 | Hand bone X-ray bone age assessment method, apparatus, computer equipment and storage medium |
| CN109100370A (en) * | 2018-06-26 | 2018-12-28 | 武汉科技大学 | A kind of pcb board defect inspection method based on sciagraphy and connected domain analysis |
| CN109145915A (en) * | 2018-07-27 | 2019-01-04 | 武汉科技大学 | License plate rapid distortion antidote under a kind of complex scene |
| CN109241973A (en) * | 2018-08-21 | 2019-01-18 | 南京工程学院 | A kind of full-automatic soft dividing method of character under grain background |
| CN109460701A (en) * | 2018-09-10 | 2019-03-12 | 昆明理工大学 | A kind of character recognition method based on histogram in length and breadth |
| CN110020655A (en) * | 2019-04-19 | 2019-07-16 | 厦门商集网络科技有限责任公司 | A kind of character denoising method and terminal based on binaryzation |
| CN110443244A (en) * | 2019-08-12 | 2019-11-12 | 深圳市捷顺科技实业股份有限公司 | A kind of method and relevant apparatus of graphics process |
| CN110826569A (en) * | 2019-11-05 | 2020-02-21 | 泰康保险集团股份有限公司 | Bill image preprocessing method and device, medium and electronic equipment |
| CN111046872A (en) * | 2019-12-12 | 2020-04-21 | 深圳市杰恩世智能科技有限公司 | Optical character recognition method |
| CN111401364A (en) * | 2020-03-18 | 2020-07-10 | 深圳市市政设计研究院有限公司 | License plate positioning algorithm based on combination of color features and template matching |
| CN112016565A (en) * | 2020-10-27 | 2020-12-01 | 恒银金融科技股份有限公司 | Segmentation method for fuzzy numbers at account number of financial bill |
| CN111553317B (en) * | 2020-05-14 | 2023-08-08 | 北京惠朗时代科技有限公司 | Anti-fake code acquisition method and device, computer equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007011654A (en) * | 2005-06-30 | 2007-01-18 | Seiko Epson Corp | Method of reading embossed character |
| CN101398894A (en) * | 2008-06-17 | 2009-04-01 | 浙江师范大学 | Automobile license plate automatic recognition method and implementing device thereof |
| CN101833859A (en) * | 2010-05-14 | 2010-09-15 | 山东大学 | Self-triggering license plate identification method based on virtual coil |
| CN102254152A (en) * | 2011-06-17 | 2011-11-23 | 东南大学 | License plate location method based on color change points and color density |
| CN102385753A (en) * | 2011-11-17 | 2012-03-21 | 江苏大学 | Illumination-classification-based adaptive image segmentation method |
-
2013
- 2013-01-24 CN CN201310028106.8A patent/CN103116751B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007011654A (en) * | 2005-06-30 | 2007-01-18 | Seiko Epson Corp | Method of reading embossed character |
| CN101398894A (en) * | 2008-06-17 | 2009-04-01 | 浙江师范大学 | Automobile license plate automatic recognition method and implementing device thereof |
| CN101833859A (en) * | 2010-05-14 | 2010-09-15 | 山东大学 | Self-triggering license plate identification method based on virtual coil |
| CN102254152A (en) * | 2011-06-17 | 2011-11-23 | 东南大学 | License plate location method based on color change points and color density |
| CN102385753A (en) * | 2011-11-17 | 2012-03-21 | 江苏大学 | Illumination-classification-based adaptive image segmentation method |
Cited By (52)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103886320A (en) * | 2014-01-10 | 2014-06-25 | 北京环境特性研究所 | Label automatic recognition method based on parallel support vector machine weighting |
| CN103824078A (en) * | 2014-03-18 | 2014-05-28 | 厦门翼歌软件科技有限公司 | Complex scene multi-license plate positioning method |
| CN103824078B (en) * | 2014-03-18 | 2017-10-10 | 厦门翼歌软件科技有限公司 | The many license plate locating methods of complex scene |
| CN104102909A (en) * | 2014-07-23 | 2014-10-15 | 中科联合自动化科技无锡有限公司 | Vehicle characteristic positioning and matching method based on multiple-visual information |
| CN104102909B (en) * | 2014-07-23 | 2017-03-15 | 中科联合自动化科技无锡有限公司 | Vehicle characteristics positioning and matching process based on lenticular information |
| CN104399674A (en) * | 2014-09-11 | 2015-03-11 | 上海大学 | Smoke box brand visual inspection sorting system |
| CN105631445A (en) * | 2014-11-06 | 2016-06-01 | 通号通信信息集团有限公司 | Character recognition method and system for license plate with Chinese characters |
| CN104463195B (en) * | 2014-11-08 | 2018-08-07 | 沈阳工业大学 | Printing digit recognizing method based on template matches |
| CN104463195A (en) * | 2014-11-08 | 2015-03-25 | 沈阳工业大学 | Printing style digital recognition method based on template matching |
| CN105719488A (en) * | 2014-12-02 | 2016-06-29 | 杭州海康威视数字技术股份有限公司 | License plate recognition method and apparatus, and camera and system for license plate recognition |
| CN104463134B (en) * | 2014-12-19 | 2019-02-12 | 深圳市捷顺科技实业股份有限公司 | A kind of detection method of license plate and system |
| CN104463134A (en) * | 2014-12-19 | 2015-03-25 | 深圳市捷顺科技实业股份有限公司 | License plate detection method and system |
| CN104636309A (en) * | 2015-02-06 | 2015-05-20 | 南京理工大学 | Matrix calculator based on machine vision and matrix identification method |
| CN104636309B (en) * | 2015-02-06 | 2018-01-05 | 南京理工大学 | Matrix recognition methods based on machine vision |
| WO2016134666A1 (en) * | 2015-02-28 | 2016-09-01 | 华为技术有限公司 | Method and device for discovering friend through image matching |
| CN104866849A (en) * | 2015-04-30 | 2015-08-26 | 天津大学 | Food nutrition label identification method based on mobile terminal |
| CN105279512B (en) * | 2015-10-22 | 2018-07-10 | 东方网力科技股份有限公司 | A kind of inclination licence plate recognition method and device |
| CN105279512A (en) * | 2015-10-22 | 2016-01-27 | 东方网力科技股份有限公司 | Tilt vehicle license plate recognition method and device |
| CN105447831A (en) * | 2015-12-14 | 2016-03-30 | 浙江宇视科技有限公司 | License plate image processing method and apparatus |
| CN105447831B (en) * | 2015-12-14 | 2019-02-15 | 浙江宇视科技有限公司 | License plate image processing method and processing device |
| 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 |
| CN106485234A (en) * | 2016-10-21 | 2017-03-08 | 合肥哦走信息技术有限公司 | One kind is based on vehicle identification method in intelligent transportation system |
| CN106650732B (en) * | 2016-12-27 | 2020-04-28 | 深圳市捷顺科技实业股份有限公司 | License plate recognition method and device |
| CN106650732A (en) * | 2016-12-27 | 2017-05-10 | 深圳市捷顺科技实业股份有限公司 | License plate identification method and apparatus |
| CN106951896B (en) * | 2017-02-22 | 2020-01-03 | 武汉黄丫智能科技发展有限公司 | License plate image tilt correction method |
| CN106951896A (en) * | 2017-02-22 | 2017-07-14 | 武汉黄丫智能科技发展有限公司 | A kind of license plate image sloped correcting method |
| CN108985137A (en) * | 2017-06-02 | 2018-12-11 | 杭州海康威视数字技术股份有限公司 | A kind of licence plate recognition method, apparatus and system |
| CN108596172A (en) * | 2018-04-11 | 2018-09-28 | 航天科技控股集团股份有限公司 | A kind of the vehicle binding device and method of car networking client |
| CN108830274A (en) * | 2018-05-03 | 2018-11-16 | 南京邮电大学 | Characters on license plate detects positioning and recognition methods automatically under a kind of complex background environment |
| CN108968991A (en) * | 2018-05-08 | 2018-12-11 | 平安科技（深圳）有限公司 | Hand bone X-ray bone age assessment method, apparatus, computer equipment and storage medium |
| CN108968991B (en) * | 2018-05-08 | 2022-10-11 | 平安科技（深圳）有限公司 | Hand bone X-ray film bone age assessment method, device, computer equipment and storage medium |
| CN108830314A (en) * | 2018-06-06 | 2018-11-16 | 西安纸贵互联网科技有限公司 | A kind of sxemiquantitative recognition methods of external diagnosis reagent and device |
| CN109100370A (en) * | 2018-06-26 | 2018-12-28 | 武汉科技大学 | A kind of pcb board defect inspection method based on sciagraphy and connected domain analysis |
| CN109145915A (en) * | 2018-07-27 | 2019-01-04 | 武汉科技大学 | License plate rapid distortion antidote under a kind of complex scene |
| CN109145915B (en) * | 2018-07-27 | 2021-08-06 | 武汉科技大学 | Rapid distortion correction method for license plate under complex scene |
| CN109241973B (en) * | 2018-08-21 | 2022-02-08 | 南京工程学院 | Full-automatic soft segmentation method for characters under texture background |
| CN109241973A (en) * | 2018-08-21 | 2019-01-18 | 南京工程学院 | A kind of full-automatic soft dividing method of character under grain background |
| CN109460701B (en) * | 2018-09-10 | 2020-09-25 | 昆明理工大学 | Font identification method based on longitudinal and transverse histograms |
| CN109460701A (en) * | 2018-09-10 | 2019-03-12 | 昆明理工大学 | A kind of character recognition method based on histogram in length and breadth |
| CN110020655B (en) * | 2019-04-19 | 2021-08-20 | 厦门商集网络科技有限责任公司 | Character denoising method and terminal based on binarization |
| CN110020655A (en) * | 2019-04-19 | 2019-07-16 | 厦门商集网络科技有限责任公司 | A kind of character denoising method and terminal based on binaryzation |
| CN110443244B (en) * | 2019-08-12 | 2023-12-05 | 深圳市捷顺科技实业股份有限公司 | Graphics processing method and related device |
| CN110443244A (en) * | 2019-08-12 | 2019-11-12 | 深圳市捷顺科技实业股份有限公司 | A kind of method and relevant apparatus of graphics process |
| CN110826569A (en) * | 2019-11-05 | 2020-02-21 | 泰康保险集团股份有限公司 | Bill image preprocessing method and device, medium and electronic equipment |
| CN110826569B (en) * | 2019-11-05 | 2022-07-19 | 泰康保险集团股份有限公司 | Bill image preprocessing method, device, medium and electronic equipment |
| CN111046872A (en) * | 2019-12-12 | 2020-04-21 | 深圳市杰恩世智能科技有限公司 | Optical character recognition method |
| CN111046872B (en) * | 2019-12-12 | 2023-10-24 | 深圳市杰恩世智能科技有限公司 | Optical character recognition method |
| CN111401364B (en) * | 2020-03-18 | 2023-07-25 | 深圳市市政设计研究院有限公司 | License plate positioning algorithm based on combination of color features and template matching |
| CN111401364A (en) * | 2020-03-18 | 2020-07-10 | 深圳市市政设计研究院有限公司 | License plate positioning algorithm based on combination of color features and template matching |
| CN111553317B (en) * | 2020-05-14 | 2023-08-08 | 北京惠朗时代科技有限公司 | Anti-fake code acquisition method and device, computer equipment and storage medium |
| CN112016565A (en) * | 2020-10-27 | 2020-12-01 | 恒银金融科技股份有限公司 | Segmentation method for fuzzy numbers at account number of financial bill |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103116751B (en) | 2016-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103116751B (en) | A kind of Method of Automatic Recognition for Character of Lcecse Plate | |
| CN104751187B (en) | Meter reading automatic distinguishing method for image | |
| CN105373794B (en) | A kind of licence plate recognition method | |
| Patel et al. | Automatic number plate recognition system (anpr): A survey | |
| CN105160309B (en) | Three lanes detection method based on morphological image segmentation and region growing | |
| CN101334836B (en) | License plate positioning method incorporating color, size and texture characteristic | |
| CN102708356B (en) | Automatic license plate positioning and recognition method based on complex background | |
| CN102375982B (en) | Multi-character characteristic fused license plate positioning method | |
| CN100414561C (en) | Vehicle plate extracting method based on skiagraphy and mathematical morphology | |
| CN106650553A (en) | License plate recognition method and system | |
| CN105893949A (en) | Lane line detection method under complex road condition scene | |
| CN104392205A (en) | Abnormal vehicle license plate recognition method and system | |
| CN107122777A (en) | A kind of vehicle analysis system and analysis method based on video file | |
| CN104036262B (en) | A kind of method and system of LPR car plates screening identification | |
| CN104050450A (en) | Vehicle license plate recognition method based on video | |
| CN103065138A (en) | Recognition method of license plate number of motor vehicle | |
| CN104751142A (en) | Natural scene text detection algorithm based on stroke features | |
| CN103077384A (en) | Method and system for positioning and recognizing vehicle logo | |
| CN103870803A (en) | Vehicle license plate recognition method and system based on coarse positioning and fine positioning fusion | |
| CN103198315A (en) | License plate character segmentation algorithm based on character outline and template matching | |
| CN102364496A (en) | Method and system for identifying automobile license plates automatically based on image analysis | |
| CN103530608A (en) | Vehicle type judgment method and vehicle type judgment device | |
| CN104809433A (en) | Zebra stripe detection method based on maximum stable region and random sampling | |
| CN103034848A (en) | Identification method of form type | |
| Guo et al. | Mixed vertical-and-horizontal-text traffic sign detection and recognition for street-level scene |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160706 Termination date: 20190124 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |