CN109389122A - A kind of license plate locating method and device - Google Patents

A kind of license plate locating method and device Download PDF

Info

Publication number
CN109389122A
CN109389122A CN201710656000.0A CN201710656000A CN109389122A CN 109389122 A CN109389122 A CN 109389122A CN 201710656000 A CN201710656000 A CN 201710656000A CN 109389122 A CN109389122 A CN 109389122A
Authority
CN
China
Prior art keywords
license plate
plate area
initial
character block
bit 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
Application number
CN201710656000.0A
Other languages
Chinese (zh)
Other versions
CN109389122B (en
Inventor
林翠翠
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Hikvision Digital Technology Co Ltd
Original Assignee
Hangzhou Hikvision Digital Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hangzhou Hikvision Digital Technology Co Ltd filed Critical Hangzhou Hikvision Digital Technology Co Ltd
Priority to CN201710656000.0A priority Critical patent/CN109389122B/en
Publication of CN109389122A publication Critical patent/CN109389122A/en
Application granted granted Critical
Publication of CN109389122B publication Critical patent/CN109389122B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/14—Image acquisition
    • G06V30/148—Segmentation of character regions
    • G06V30/153—Segmentation of character regions using recognition of characters or words
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/14—Image acquisition
    • G06V30/148—Segmentation of character regions
    • G06V30/158—Segmentation of character regions using character size, text spacings or pitch estimation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
    • G06V20/625—License plates
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Character Input (AREA)
  • Vehicle Waterproofing, Decoration, And Sanitation Devices (AREA)
  • Traffic Control Systems (AREA)

Abstract

The embodiment of the present application provides a kind of license plate locating method and device.The described method includes: obtaining the bit image undetermined comprising license plate, and the bit image undetermined is detected, obtains initial license plate area；Determine each character block in the initial license plate area；According to determining each character block, judge whether the initial license plate area is multilayer license plate area；If it is, determining the license plate area of the bit image undetermined according at least one layer of license plate area.Using scheme provided by the embodiments of the present application, the efficiency of License Plate process can be improved.

Description

A kind of license plate locating method and device
Technical field
This application involves field of intelligent transportation technology, more particularly to a kind of license plate locating method and device.
Background technique
License plate is vehicle " identity card ", is an important information for being different from other motor vehicles.License plate recognition technology It has been widely used in the scenes such as bayonet, parking lot and electronic police, to obtain the number plate information of vehicle in scene.
It generally may include License Plate, Character segmentation, character recognition three when carrying out Car license recognition to license plate image Link.According to the distributed architecture of characters on license plate, license plate can be divided into single layer license plate and the double-deck license plate.Character is distributed in single layer license plate In one layer；Character is distributed in upper layer and lower layer in the double-deck license plate, that is, includes upper layer character and lower layer's character.For the double-deck license plate Region, in License Plate link, usually using that more tomographic image region of character in the double-deck license plate area as license plate area.
In the related technology, to comprising license plate bit image undetermined carry out License Plate when, can to image to be positioned into Row Primary Location obtains initial license plate area；Initial license plate area is matched one by one with the preset double-deck plate template, is sentenced Whether the license plate to break in the initial license plate area belongs to the double-deck license plate, if belonged to, according to the double-deck license plate mould of successful match Plate determines the license plate area of bit image undetermined from initial license plate area.
When constructing the preset double-deck plate template, generally according to the dimension information of the double-deck license plate, upper and lower level character ratio, The features such as colouring information are constructed.Different types of license plate, features described above are different.For example, each bilayer license plate in Fig. 1 It has different sizes, the features such as different upper and lower level character ratios, different colors.And it is fixed carrying out license plate to license plate image Before position, the type of license plate in image can not be often learnt.To guarantee to obtain preferably when carrying out License Plate to license plate image Positioning result needs for the different types of double-deck license plate building plate template, and then causes to need in above-mentioned License Plate mode Want the matched double-deck license plate template number larger, License Plate process it is inefficient.
Summary of the invention
The embodiment of the present application has been designed to provide a kind of license plate locating method and device, to improve License Plate process Efficiency.Specific technical solution is as follows.
In order to achieve the above object, the embodiment of the present application provides a kind of license plate locating method, which comprises
The bit image undetermined comprising license plate is obtained, and the bit image undetermined is detected, obtains initial license plate area；
Determine each character block in the initial license plate area；
According to determining each character block, judge whether the initial license plate area is multilayer license plate area；
If it is, determining the license plate area of the bit image undetermined according at least one layer of license plate area.
Optionally, the described the step of bit image undetermined is detected, obtains initial license plate area, comprising:
According to the jump feature of license plate pixel value, the bit image undetermined is detected, initial license plate area is obtained；Or Person,
According to the pixel value tag at license plate edge, the bit image undetermined is detected, initial license plate area is obtained.
Optionally, the step of each character block in the determination initial license plate area, comprising:
Binary conversion treatment is carried out to the pixel value of the initial license plate area using preset different threshold values, obtains each place Initial license plate image after reason；
The constant each connected region of pixel value is determined from each treated initial license plate image；
Image-region corresponding to external frame by each connected region is determined as each in the initial license plate area A character block.
Optionally, the step of each character block in the determination initial license plate area, comprising:
Determine each doubtful character block in the initial license plate area；
According to the distance between the size characteristic of preset characters on license plate or preset characters on license plate feature, from each doubtful Character block is filtered out in character block.
Optionally, after determining each character block in the initial license plate area, the method also includes:
Judge whether the total quantity of each character block is greater than preset quantity threshold value；
If it is, execution is described according to determining each character block, judge whether the initial license plate area is multilayer The step of license plate area.
Optionally, described according at least one layer of license plate area, the step of determining the license plate area of the bit image undetermined, packet It includes:
By the coboundary for each character block for including at least one layer of license plate area and lower boundary, be identified as it is described to Position coboundary and the lower boundary of the license plate area of image；
Judge at least one layer of license plate area the left margin of leftmost side character block and the initial license plate area left margin it Between distance whether be greater than pre-determined distance threshold value；If it is, by the left side of leftmost side character block at least one layer of license plate area Boundary is determined as the left margin of the license plate area of the bit image undetermined；If it is not, then by the left margin of the initial license plate area It is determined as the left margin of the license plate area of the bit image undetermined；
Judge at least one layer of license plate area the right margin of rightmost side character block and the initial license plate area right margin it Between distance whether be greater than pre-determined distance threshold value；If it is, by the right of rightmost side character block at least one layer of license plate area Boundary is determined as the right margin of the license plate area of the bit image undetermined；If it is not, then by the right margin of the initial license plate area It is determined as the right margin of the license plate area of the bit image undetermined.
Optionally, described according at least one layer of license plate area, the step of determining the license plate area of the bit image undetermined, packet It includes:
Each layer license plate area for including by the initial license plate area, is identified as the license plate area of the bit image undetermined Domain；Alternatively,
By comprising the most one layer of license plate area of character number of blocks be determined as the license plate area of the bit image undetermined.
Optionally, when judging the initial license plate area is not multilayer license plate area, the method also includes:
The initial license plate area is determined as to the license plate area of the bit image undetermined.
Optionally, after the license plate area for determining the bit image undetermined, the method also includes:
The apex coordinate of boundary rectangle comprising same layer character block is exported as positioning result.
In order to achieve the above object, the embodiment of the present application also provides a kind of license plate positioning device, and described device includes:
Initial alignment module is detected for obtaining the bit image undetermined comprising license plate, and to the bit image undetermined, Obtain initial license plate area；
Character block determining module, for determining each character block in the initial license plate area；
License plate judgment module, for judging whether the initial license plate area is multilayer according to determining each character block License plate area；
First locating module, for determining the license plate area of the bit image undetermined according at least one layer of license plate area.
Optionally, the initial alignment module, is specifically used for:
According to the jump feature of license plate pixel value, the bit image undetermined is detected, initial license plate area is obtained；Or Person,
The initial alignment module, is specifically used for:
According to the pixel value tag at license plate edge, the bit image undetermined is detected, initial license plate area is obtained.
Optionally, the character block determining module, comprising:
Binaryzation submodule, for carrying out two-value to the pixel value of the initial license plate area using preset different threshold values Change processing, obtains each treated initial license plate image；
Region determines submodule, for determining the constant each company of pixel value from each treated initial license plate image Logical region；
Character block determines submodule, for image-region corresponding to the external frame by each connected region, is determined as Each character block in the initial license plate area.
Optionally, the character block determining module, is specifically used for:
Determine each doubtful character block in the initial license plate area, according to the size characteristic of preset characters on license plate or The distance between preset characters on license plate feature filters out character block from each doubtful character block.
Optionally, the device further include:
Quantity judgment module, for judging each word after determining each character block in the initial license plate area Whether the total quantity of symbol block is greater than preset quantity threshold value, if it is, executing the license plate judgment module.
Optionally, first locating module, comprising:
Up-and-down boundary determines submodule, the coboundary of each character block for that will include at least one layer of license plate area and Lower boundary is identified as coboundary and the lower boundary of the license plate area of the bit image undetermined；
Left margin determines submodule, for judge at least one layer of license plate area the left margin of leftmost side character block with it is described Whether the distance between initial license plate area left margin is greater than pre-determined distance threshold value；If it is, by least one layer of license plate area The left margin of middle leftmost side character block is determined as the left margin of the license plate area of the bit image undetermined；If it is not, then will be described The left margin of initial license plate area is determined as the left margin of the license plate area of the bit image undetermined；
Right margin determines submodule, for judge at least one layer of license plate area the right margin of rightmost side character block with it is described Whether the distance between initial license plate area right margin is greater than pre-determined distance threshold value；If it is, by least one layer of license plate area The right margin of middle rightmost side character block is determined as the right margin of the license plate area of the bit image undetermined；If it is not, then will be described The right margin of initial license plate area is determined as the right margin of the license plate area of the bit image undetermined.
Optionally, first locating module, is specifically used for:
Each layer license plate area for including by the initial license plate area, is identified as the license plate area of the bit image undetermined Domain；Alternatively,
First locating module, is specifically used for:
By comprising the most one layer of license plate area of character number of blocks be determined as the license plate area of the bit image undetermined.
Optionally, described device further include:
Second locating module, for when judging the initial license plate area is not multilayer license plate area, will it is described just Beginning license plate area is determined as the license plate area of the bit image undetermined.
Optionally, described device further include:
As a result output module, for that after the license plate area for determining the bit image undetermined, will include same layer character The apex coordinate of the boundary rectangle of block is exported as positioning result.
The embodiment of the present application provides a kind of electronic equipment, which includes processor, memory；
The memory, for storing executable code；
The processor, by reading the executable code stored in memory, to be mentioned for executing the embodiment of the present application The license plate locating method of confession.This method comprises:
The bit image undetermined comprising license plate is obtained, and the bit image undetermined is detected, obtains initial license plate area；
Determine each character block in the initial license plate area；
According to determining each character block, judge whether the initial license plate area is multilayer license plate area；
If it is, determining the license plate area of the bit image undetermined according at least one layer of license plate area.
License plate locating method and device provided by the embodiments of the present application can detect image to be positioned, obtain just Beginning license plate area determines each character block in the initial license plate area, according to determining each character block, judges initial vehicle Whether board region is multilayer license plate area；If it is, determining the license plate area of bit image undetermined according at least one layer of license plate area Domain.
That is, the embodiment of the present application can determine each character block from initial license plate area, according to each character The position distribution of block judges whether initial license plate area is multilayer license plate area, without a large amount of double-deck license plate moulds of matching one by one Plate, therefore can be improved the efficiency of License Plate process.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below There is attached drawing needed in technical description to be briefly described.It should be evident that the accompanying drawings in the following description is only this Some embodiments of application for those of ordinary skill in the art without creative efforts, can be with It obtains other drawings based on these drawings.
Fig. 1 is partial double license plate instance graph；
Fig. 2 is a kind of flow diagram of license plate locating method provided by the embodiments of the present application；
Fig. 2 a is a kind of exemplary diagram of connected region and corresponding character block in the embodiment of the present application；
Fig. 2 b is a kind of exemplary diagram for the character block that step S202 is determined in Fig. 2；
Fig. 3 a is a kind of flow diagram of step S202 in Fig. 2；
Fig. 3 b is a kind of flow diagram of step S204 in Fig. 2；
Fig. 4 is a kind of structural schematic diagram of license plate positioning device provided by the embodiments of the present application；
Fig. 5 is a kind of structural schematic diagram of electronic equipment provided by the embodiments of the present application.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present application, technical solutions in the embodiments of the present application carries out clear, complete Whole description.Obviously, described embodiment is only a part of the embodiment of the application, instead of all the embodiments.Base Embodiment in the application, those of ordinary skill in the art are obtained all without making creative work Other embodiments shall fall in the protection scope of this application.
The embodiment of the present application provides a kind of license plate locating method and device, is applied to electronic equipment, which can With but be not limited to computer, tablet computer, smart phone, automobile data recorder, the monitoring camera for being equipped with embedded system etc.. Using technical solution provided by the embodiments of the present application, the efficiency of License Plate process can be improved.Below by specific embodiment, The application is described in detail.
Fig. 2 is a kind of flow diagram of license plate locating method provided by the embodiments of the present application, is applied to electronic equipment.It should Method includes the following steps S201~S204:
Step S201: the bit image undetermined comprising license plate is obtained, and image to be positioned is detected, obtains initial license plate Region.
Wherein, it may include image capture device as the electronic equipment internal of executing subject, image can also not included Acquire equipment.
When electronic equipment internal includes image capture device, electronic equipment is obtaining the bit image undetermined comprising license plate When, it can specifically include: receiving the bit image undetermined comprising license plate of image capture device acquisition.
When electronic equipment internal does not include image capture device, which can be with external image capture device It is connected, electronic equipment can specifically include when obtaining the bit image undetermined comprising license plate: obtains image capture device acquisition Bit image undetermined comprising license plate.
The bit image undetermined comprising license plate obtained can be what image capture device acquired in real time, may not be in real time Acquisition, but what image capture device stored after being collected in advance.
Above-mentioned bit image undetermined is it is to be understood that the image of License Plate will be carried out.It is understood that license plate is usually It is mounted or placed on vehicle, therefore above-mentioned bit image undetermined is it is to be understood that will carry out License Plate comprising vehicle Image.Based on this, above-mentioned bit image undetermined can be the image comprising vehicle captured on road, can be and shoots in parking lot The image etc. comprising vehicle.Certainly, above-mentioned bit image undetermined can also obtain by other means, and the application is to figure to be positioned The acquisition pattern of picture is not specifically limited.
Initial license plate area can be understood as the region obtained when carrying out Primary Location to image to be positioned, alternatively referred to as fixed Position block.Since the license plate for including in bit image undetermined in practice can be single layer license plate, or the double-deck license plate, therefore it is initial License plate area may be the image-region of single layer license plate or the image-region of the double-deck license plate.For example, each license plate image in Fig. 1 The initial license plate area that can be.Certainly, when there is the more layers license plates such as three layers of license plate, four layers of license plate in practice, just Beginning license plate area may be the image-region of multilayer license plate.
Wherein, single layer license plate is that characters on license plate is distributed in the license plate type in one layer.The double-deck license plate is characters on license plate distribution License plate type in upper and lower two layers, the double-deck license plate include upper layer character and lower layer's character.Three layers of license plate are characters on license plate distribution License plate type in upper, middle and lower-ranking.
Image to be positioned is detected, when obtaining initial license plate area, a variety of specific embodiments, packet can be used It includes but is not limited to following two:
One is according to the jump feature of license plate pixel value, detect image to be positioned, obtain initial license plate area Domain.
In this embodiment, when being detected according to the jump feature of license plate pixel value to image to be positioned, specifically It can be to be progressively scanned to image to be positioned, detect the adjacent pixel that pixel value in bit image undetermined is greater than threshold value Point, as pixel jump point；According to the region of pixel jump point dense distribution, initial license plate area is determined from bit image undetermined Domain.It is understood that the value differences in license plate between character and background are larger, therefore exist largely in license plate area The pixel jump point of dense distribution can be from bit image undetermined according to the dense distribution feature and the pixel jump point determined Determine initial license plate area.
Another kind is, according to the pixel value tag at license plate edge, to detect to image to be positioned, obtains initial license plate area Domain.
In this embodiment, according to the pixel value tag at license plate edge treat positioning image detected when, specifically It can be that gray proces are carried out to image to be positioned, obtain gray level image, edge extracting is carried out to the gray level image, obtains side Edge image is determined for compliance with the connected region of preset license plate edge shape from the edge image, and the connected region is corresponding Image-region is as initial license plate area.
Step S202: each character block in initial license plate area is determined.
In this step, when determining each character block in initial license plate area, 1~step of following steps can be specifically used 3:
Step 1: binary conversion treatment being carried out to the pixel value of initial license plate area using preset different threshold values, is obtained each Treated initial license plate image.
Wherein, above-mentioned threshold value can be selected from pixel value 0~255, such as can take the different pixels between 60~80 Value.The quantity of threshold value can be multiple, for example, 10,15,20 or other values, this is all feasible.
It may include: needle when carrying out binary conversion treatment to the pixel value of initial license plate area using preset different threshold values To each pixel in initial license plate area, judge whether the pixel value of the pixel is greater than above-mentioned threshold value, if it is greater, then The pixel value of the pixel is set to the first presetted pixel value, if it is less, it is pre- that the pixel value of the pixel is set to second If pixel value.Above embodiment is to carry out the process of binaryzation, needle to the pixel value of initial license plate area using a threshold value To each different threshold value, binary conversion treatment can be carried out using pixel value of the aforesaid way to initial license plate area.Wherein, Second presetted pixel value is different from the first presetted pixel value.For example, the first presetted pixel value and the second presetted pixel value can divide 0 and 255 or 255 and 0 is not taken, naturally it is also possible to take other two different pixel values.The application does not do specific limit to this It is fixed.
Step 2: determining the constant each connected region of pixel value from each treated initial license plate image.
It is understood that character portion and background parts are generally comprised in license plate, in order to make the character of license plate be easy to distinguish Recognize, the discrimination between this two-part pixel value is usually all arranged bigger.Therefore, when the different threshold values of use are to initial vehicle When the pixel value of board image carries out binary conversion treatment, the background parts of character portion and surrounding are in each binary conversion treatment result In may each be and stablize constant region.That is, pixel value does not become in each initial license plate image after treatment There are corresponding relationships with character portion for the connected region of change.
Step 3: image-region corresponding to the external frame by each connected region is determined as in initial license plate area Each character block.
Wherein, external frame can be understood as to surround the minimum frame of connected region.As an implementation, outside Edge fit frame can be boundary rectangle, and corresponding, identified each character block is also rectangular image area；Or external frame can Think external circle, corresponding, identified each character block is also circular image regions.Since boundary rectangle frame is easier to It is determined, therefore processing complexity when determining character block can reduce as external frame using boundary rectangle frame.
As an example, the black font part in left side is a connected region in Fig. 2 a, which does not include Background area, the rectangle frame in Fig. 2 a in image right are the boundary rectangle of the connected region.The image district that the rectangle frame includes Domain is character block, and wherein character block includes grid background (assuming that there are grids around character B in initial license plate image).
Determine each character block in initial license plate area, it can be understood as, determine each character in initial license plate area The coordinate of block.
As an example, when determining the character block in number is 1 in Fig. 1 initial license plate area, available Fig. 2 b Shown in result.It wherein, is character block with the region that dotted line frames in Fig. 2 b.
Since there may be many interference characters, such as mud point, rivet or other image noises in initial license plate image Deng, can be when determining the constant each connected region of pixel value, by these noise words in order to improve the accuracy of image procossing The connected region of symbol is removed.When determining each character block, preset size threshold and/or area threshold pair can use Each connected region is screened, and removal size is less than size threshold, area is less than the connected region of area threshold, will be remaining Connected region is determined as character block.
As a kind of specific embodiment, step 3, image district corresponding to the external frame by each connected region Domain when each character block being determined as in initial license plate area, is specifically as follows:
Judge whether each connected region meets preset condition, if it is, by the external frame institute of each connected region Corresponding image-region, each character block being determined as in initial license plate area.
For example, preset condition can be at least one of following situations:
The size of connected region is greater than pre-set dimension threshold value；
The area of connected region is greater than preset area threshold value.
Wherein, the size of connected region includes the height of connected region and the width of connected region, corresponding pre-set dimension Threshold value includes preset height threshold value and predetermined width threshold value.
For example, when the character in license plate is made of different parts, for example, include in license plate " Shanghai " this by difference The word that part is constituted.Four connected regions, i.e. " family " of three points in left side and right side may be corresponded to by being somebody's turn to do " Shanghai ".Since " Rui " is right The area for three connected regions answered is less than area threshold or size is less than size threshold, then the connected region meeting of these three points It is removed, " Shanghai " is only determined " family " this character block.For another example there are a rivets beside character, if rivet and word Symbol is connected, then rivet and character can be confirmed as a character block；If rivet is not attached to character, since rivet is corresponding Connected region area is less than area threshold or size is less than size threshold and can be removed.
That is, each character block in the initial license plate area determined in the present embodiment, can be characters on license plate A part, be also possible to complete characters on license plate.
It, can will be initial when determining each character block in initial license plate area as another specific embodiment License plate area inputs preset character block and determines in network, determines from the character block and determines each character block in network.Wherein, should Character block determines network, and the license plate area obtained when for determining that network training is completed according to the character block is corresponding with character block Relationship determines each character block from received initial license plate area.The character block determines that network can be for using machine learning The network that algorithm is trained great amount of samples license plate area.
Step S203: according to determining each character block, judging whether initial license plate area is multilayer license plate area, if It is to then follow the steps S204.
Wherein, multiple license plate area may include the double-deck license plate area, three layers of license plate area, four layers of license plate area etc..Mesh Preceding common multilayer license plate is the double-deck license plate, therefore the application is introduced by taking the double-deck license plate area as an example.
In this step, according to determining each character block, when judging whether initial license plate area is multilayer license plate area, It can specifically include numerous embodiments.Such as:
Mode one, judges whether each character block is located in one layer, if it is not, then determining that initial license plate area is multilayer vehicle Board region.
It in the present embodiment, specifically can be according to each character block when judging that whether each character block is located in one layer Coordinate judged.For example, judging whether the preset coordinate point in each character block is respectively positioned in a line, if it is not, then sentencing Fixed each character block is not located in one layer.Wherein, preset coordinate point can centered on point, the upper left corner, the upper right corner, the lower left corner or the right side Inferior horn, certainly, as long as the coordinate points where character block in image-region, are feasible.
When judging whether each character block is located in one layer, or, the coordinate of each character block central point is obtained, it should Coordinate can be ordinate, the slope of straight line where determining the central point of each character block, judge the slope absolute value whether Less than default slope threshold value, if it is not, then determining that each character block is not located in one layer.Wherein, it is contemplated that vehicle in real image There are certain inclinations for board character, therefore default slope threshold value can be taken as the value close to 0, such as 0.01~0.1.
It when judging that whether each character block is located in one layer, can also be to calculate the vertical seat of every two character block central point Absolute difference between mark obtains the distance between each central point value, judge each distance value whether be all larger than it is default away from From threshold value, if it is, in determining that each character block is located on the same floor.
Mode two determines the layer where each character block, judges whether the quantity of identified layer is greater than one, if so, Then determine that initial license plate area is multilayer license plate area.
In the present embodiment, it when determining the layer where each character block, can be determined according to the position of each character block Layer where each character block.For example, can determine each character block institute according to the position of preset coordinate point in each character block Layer.Wherein, preset coordinate point can centered on point, the upper left corner, the upper right corner, the lower left corner or the lower right corner, certainly, as long as word Coordinate points where according with block in image-region, are feasible.
When judging initial license plate area not is multilayer license plate area, it is believed that the license plate in initial license plate image is single layer Initial license plate area can be determined as the license plate area of bit image undetermined at this time, can also do other processing by license plate.
Step S204: according at least one layer of license plate area, the license plate area of bit image undetermined is determined.
When initial license plate area is multilayer license plate area, according at least one layer of license plate area, bit image undetermined is determined When license plate area, it specifically can be, but not limited to include following implementation:
Mode one, each layer license plate area for including by initial license plate area, is identified as the license plate area of bit image undetermined Domain.
If initial license plate area is two layers of license plate area, two layers of license plate area can be determined as bitmap undetermined respectively The license plate area of picture.For example, for initial license plate area shown in Fig. 2 b, can by where R, V, B upper layer license plate area and 7,2, lower layer's license plate area where X, G, P is identified as the license plate area of bit image undetermined.
Mode two, by comprising the most floor license plate area of character number of blocks be determined as the license plate area of bit image undetermined Domain.
If initial license plate area is two layers of license plate area, one layer of most license plate area of character number of blocks can be determined For the license plate area of bit image undetermined.For example, for initial license plate area shown in Fig. 2 b, the upper deck of the initial license plate area Board region include tri- characters of R, V, B, lower layer's license plate area include 7, five 2, X, G, P characters, can by comprising character block The most lower layer's license plate area of quantity is determined as the license plate area of bit image undetermined.
It is right for multilayer license plate in relevant Car license recognition process in the alternative embodiment of the application It, can also be to the vehicle after image-region to be positioned carries out License Plate, and while being split to obtained license plate area The region of board overlying regions and lower section carries out Text RegionDetection, obtains the license plate area of upper layer or lower layer.Then, to getting All license plate areas carry out Character segmentation.In order to avoid modifying relevant Car license recognition process, the step of the present embodiment too much One layer of most license plate area of the character number of blocks that positioning obtains is determined as to the license plate area of bit image undetermined, energy in rapid S204 It enough avoids modifying relevant Car license recognition process too much, the efficiency of Car license recognition process is improved with the smallest modification cost.Together When, the accuracy of Recognition of License Plate Characters process can also be improved.
In the alternative embodiment of the application, determine the license plate area of bit image undetermined, it can be understood as, determine to Position the upper and lower boundary and left and right boundary of the license plate area of image.
The present embodiment defines each character block from initial license plate area, therefore can be according to each word determined The position for according with block, determines the upper and lower boundary and left and right boundary of the license plate area of bit image undetermined.
Specifically, when determining the upper and lower boundary of license plate area of bit image undetermined, it can be directly by least one layer of license plate The upper and lower boundary in region is determined as the corresponding upper and lower boundary of license plate area of bit image undetermined, can also be by least one layer of vehicle The coboundary in board region or one in lower boundary, carry out group with one in the coboundary of initial license plate area or lower boundary It closes, obtains the upper and lower boundary of the license plate area of bit image undetermined.
It, can be directly by the left and right of initial license plate area when determining the left and right boundary of license plate area of bit image undetermined Boundary is determined as the corresponding left and right boundary of license plate area of bit image undetermined, can also be by the left side of at least one layer of license plate area At least one of boundary or right margin obtain the license plate of bit image undetermined in conjunction with the left margin or right margin of initial license plate area The left and right boundary in region.
Wherein, the coboundary of at least one layer of license plate area can be understood as each character block at least one layer of license plate area The mean place of coboundary, the lower boundary of at least one layer of license plate area can be understood as the following of each character block in target area The mean place on boundary.The left margin of at least one layer of license plate area can be understood as leftmost side character block at least one layer of license plate area Left margin, the right margin of at least one layer of license plate area can be understood as the right margin of rightmost side character block in target area.
As shown in the above, the present embodiment can determine each character block from initial license plate area, according to each word The position distribution for according with block, judges whether initial license plate area is multilayer license plate area, without a large amount of double-deck license plates of matching one by one Template, therefore can be improved the efficiency of License Plate process.
Meanwhile scheme provided in this embodiment generalization ability in positioning licence plate is stronger, versatility is more preferable.Due to this implementation The not complicated treatment process of the License Plate scheme of example, efficiency is higher when execution, therefore can be by this License Plate scheme It applies in the electronic equipment for not having higher operational capability, such as applies in the monitoring camera etc. comprising embedded system.This Sample can quickly and efficiently positioning licence plate, and then identify license plate number.
In a kind of embodiment of embodiment shown in Fig. 2, after the license plate area for determining bit image undetermined, this method It can also include: to export the apex coordinate of the boundary rectangle comprising same layer character block as positioning result, in order to Other equipment carry out information processing according to the positioning result.
Wherein, the apex coordinate of boundary rectangle can be four apex coordinates of boundary rectangle, or boundary rectangle Two diagonal apex coordinates.
For example, the boundary rectangle comprising same layer character block can be with when the license plate in bit image undetermined is single layer license plate It is interpreted as the boundary rectangle of all character blocks.In another example including same layer when the license plate in bit image undetermined is the double-deck license plate The boundary rectangle of character block can be understood as the boundary rectangle comprising upper layer character block and/or the external square comprising lower layer's character block Shape.In another example when the license plate in bit image undetermined is three layers or more of license plate, the external square comprising same layer character block Shape can be the boundary rectangle comprising all character blocks in same horizontal line.
In the alternative embodiment of the application, when the license plate in bit image undetermined is single layer license plate, determining After the license plate area of bit image undetermined, Character segmentation can also be carried out to the license plate area, obtain each word that segmentation obtains Region is accorded with, character recognition is carried out to each character zone, obtains the license plate number of bit image undetermined.
When the license plate in bit image undetermined is the double-deck license plate, after the license plate area for determining bit image undetermined, by In the license plate area include top area and lower region, therefore can be to the top area and lower region point in license plate area Not carry out Character segmentation, the character zone of top area and the character zone of lower region are obtained, to the character area of top area The character zone of domain and lower region carries out character recognition, obtains the character identification result of top area and the character of lower region Recognition result synthesizes two parts character identification result according to the positional relationship of character zone, obtains bit image undetermined License plate number.
In addition, due in the license plate of part there may be pattern, rivet, regions, the character block determined such as be stained and may wrap Containing these interference regions.In order to improve accuracy, character block can be screened using following embodiment.
In another embodiment of embodiment shown in Fig. 2, step S202 determines each character in initial license plate area The step of block, can carry out according to flow diagram shown in Fig. 3 a, specifically include step S202a~step S202b:
Step S202a: each doubtful character block in initial license plate area is determined.Wherein, doubtful character block includes character At least one of block or non-character block.
In this step, when determining each doubtful character block in initial license plate area, following steps 1~step can be used Rapid 3:
Step 1: binary conversion treatment being carried out to the pixel value of initial license plate area using preset different threshold values, is obtained each Treated initial license plate image.
Step 2: determining the constant each connected region of pixel value from each treated initial license plate image.
Step 3: image-region corresponding to the external frame by each connected region is determined as in initial license plate area Each doubtful character block.
It should be noted that the specific embodiment of 1~step 3 of above-mentioned steps and step S202 in embodiment illustrated in fig. 2 Similar, particular content may refer to the explanation of embodiment illustrated in fig. 2, and details are not described herein again.
Step S202b: according to the distance between the size characteristic of preset characters on license plate or preset characters on license plate feature, Character block is filtered out from each doubtful character block.
In this step, it when filtering out character block from each doubtful character block, can be determined from doubtful character block non- Character block.Non-character block, is to filter out character block from each doubtful character block determined by removing.Determining doubtful character May include following several embodiments when non-character block in block:
Mode one, the doubtful character block by height less than the first height threshold are determined as non-character block.Wherein, the first height Threshold value can be the height of 1/3~1/2 initial license plate area.
Mode two, by between the abscissa of the abscissa of central point doubtful character block central point adjacent with left and right directions away from From the doubtful character block for being greater than the first width threshold value, it is determined as non-character block.Wherein, the first width threshold value is preset value, alternatively, First width threshold value is the mean breadth of each doubtful character block.
The ratio of width to height is located at the doubtful character block in [x1, x2] range and is determined as non-character block by mode three.Wherein, x1 is Preset the ratio of width to height minimum value, x2 are preset the ratio of width to height maximum value.For example, x1 can be taken as the ratio of width to height of number 1, x2 can be with It is taken as the ratio of width to height etc. of letter w.
Wherein, the first height threshold and the first width threshold value can be the threshold determined according to the size characteristic of characters on license plate Value, the first width threshold value are also possible to the threshold value determined according to the distance between characters on license plate feature.
To sum up, the present embodiment can screen doubtful character block, remove non-character block, improve License Plate process Accuracy.
In a kind of embodiment of embodiment shown in Fig. 2, step S202 is determined each in the initial license plate area When character block, it can be realized using following implementation, including step 1 and step 2:
Step 1: being calculated according to maximum extreme value stability region (Maximally Stable Extremal Regions, MSER) Method detects the region MSER in initial license plate area.
Wherein, the region MSER is the foreground parts such as text, number, pattern in the corresponding binary image of initial license plate image The connected region of composition, and do not include surrounding background area.
Step 2: image-region corresponding to the boundary rectangle in the above-mentioned region MSER is determined as in initial license plate area Character block.
It is understood that the region MSER is the connected region that the foreground parts such as text, number, pattern are constituted, it is terrible To the image-region in the corresponding initial license plate image in the region, the boundary rectangle in the region MSER can be determined, by the external square Shape corresponding image-region in initial license plate image is determined as the character block in initial license plate area.In this way, obtained word Symbol block just contains foreground area and background area.
As it can be seen that the present embodiment can determine each character block in initial license plate area using MSER algorithm, due to MSER Algorithm has affine-invariant features, stability and multiple dimensioned characteristic, and the region MSER detected is only by pixel value (i.e. gray value) Size relation determine, it is insensitive to illumination variation, thus utilize the algorithm determine character block accuracy it is higher.
In order to further increase the accuracy of position fixing process, in a kind of embodiment of embodiment shown in Fig. 2, in step After S202, i.e., after each character block in the initial license plate area of determination, this method can also include:
Judge whether the total quantity of each character block is greater than preset quantity threshold value, if it is, illustrating this word determined It is effective to accord with block, executes step S203；If it is not, then the character block for illustrating that this is determined is invalid, it can be directly by initial license plate area Domain is determined as the license plate area of bit image undetermined, can also do other processing.
Wherein, preset quantity threshold value can be taken as in the double-deck license plate that least number of characters magnitude, the value can be according to a large amount of Characters on license plate total quantity determines in sample license plate.For example, since the least double-deck license plate of character is " 1, upper layer character+lower layer 2 The structure of a character " or " 2, upper layer character+lower layer, 3 characters ", therefore preset quantity threshold value can be taken as 3.
It should be noted that whether the total quantity of detection character block is greater than preset quantity threshold value, it can be to determining character block The validity of process judged.When the process for determining character block is effective, then step S203 is executed, it is fixed to can be improved license plate The validity of position process, avoids useless operation.
Further, if determining that character block, the present embodiment can determine each character block using MSER algorithm Total quantity is greater than preset quantity threshold value, detects the validity of MSER algorithm.
As it can be seen that the present embodiment can judge determine whether the process of character is effective according to the total quantity of each character block, when When judging result is effective, then step S203 is executed, the accuracy of position fixing process can be further increased.
In order to further increase positioning license plate area accuracy, in another embodiment of embodiment shown in Fig. 2, Step S204 the step of determining the license plate area of bit image undetermined, may refer to shown in Fig. 3 b according at least one layer of license plate area Flow diagram can specifically include following steps S204A~step S204C:
Step S204A: true respectively by the coboundary for each character block for including at least one layer of license plate area and lower boundary It is set to coboundary and the lower boundary of the license plate area of bit image undetermined.
Specifically, the average value of the coboundary of character block each at least one layer of license plate area can be determined as undetermined The coboundary of the license plate area of bit image determines the average value of the lower boundary of character block each at least one layer of license plate area For the lower boundary of the license plate area of bit image undetermined.
Step S204B: judge that the left margin of leftmost side character block and initial license plate area are left at least one layer of license plate area Whether the distance between boundary is greater than pre-determined distance threshold value, if it is, by leftmost side character block at least one layer of license plate area Left margin be determined as bit image undetermined license plate area left margin；If it is not, then the left margin of initial license plate area is true It is set to the left margin of the license plate area of bit image undetermined.
Wherein, pre-determined distance threshold value can be default value, and the default value can character width obtains in a large amount of license plates according to The statistical value arrived.
When the judging result of step S204B, which is, is, show the left margin and real vehicle of determining initial license plate area The left margin distance in board region includes more disturbing factor on the left of initial license plate area farther out.It therefore, at this time can will at least The left margin of leftmost side character block is determined as the left margin of the license plate area of bit image undetermined in one layer of license plate area, fixed to improve The accuracy of the license plate area left margin of position.
When the judging result of step S204B is no, show the left margin and real vehicle of determining initial license plate area The left margin in board region is closer, can be directly using the left margin of initial license plate area as the license plate area of bit image undetermined Left margin, it is not necessary that the left margin of leftmost side character block at least one layer of license plate area to be determined as to the license plate area of bit image undetermined The left margin in domain reduces operating process.
Step S204C: judge that the right margin of rightmost side character block and initial license plate area are right at least one layer of license plate area Whether the distance between boundary is greater than pre-determined distance threshold value, if it is, by rightmost side character block at least one layer of license plate area Right margin be determined as bit image undetermined license plate area right margin；If it is not, then the right margin of initial license plate area is true It is set to the right margin of the license plate area of bit image undetermined.
When the judging result of step S204C, which is, is, show the right margin and real vehicle of determining initial license plate area The right margin distance in board region includes more disturbing factor on the right side of initial license plate area farther out.It therefore, at this time can will at least The right margin of rightmost side character block is determined as the right margin of the license plate area of bit image undetermined in one layer of license plate area, fixed to improve The accuracy of the license plate area right margin of position.
When the judging result of step S204C is no, show the right margin and real vehicle of determining initial license plate area The right margin in board region is closer, can be directly using the right margin of initial license plate area as the license plate area of bit image undetermined Right margin, it is not necessary that the right margin of rightmost side character block at least one layer of license plate area to be determined as to the license plate area of bit image undetermined The right margin in domain reduces operating process.
As it can be seen that the present embodiment can be according to the coboundary of each character block for including at least one layer of license plate area and following Boundary, left margin and right margin determine coboundary and lower boundary, the left margin and right margin of the license plate area of bit image undetermined, energy Enough accuracys for improving identified license plate area.
In addition, in the case that the license plate in initial license plate area is single layer license plate, initial license plate area is directly true When being set to the license plate area of bit image undetermined, can also according to the left and right boundary of one layer of license plate area where each character block, With the left and right boundary of initial license plate area, the left and right boundary of the license plate area of bit image undetermined is determined respectively.
According to the left and right boundary of one layer of license plate area where each character block, and the left and right side of initial license plate area Boundary is specifically as follows when determining the left and right boundary of the license plate area of bit image undetermined respectively, judges in one layer of license plate area most Whether the distance between the left margin of left side character block and initial license plate area left margin are greater than pre-determined distance threshold value, if so, The left margin of leftmost side character block in one layer of license plate area is then determined as to the left margin of the license plate area of bit image undetermined；If It is no, then the left margin of initial license plate area is determined as to the left margin of the license plate area of bit image undetermined.
Judge the distance between the right margin of rightmost side character block and initial license plate area right margin in one layer of license plate area Whether pre-determined distance threshold value is greater than, if it is, the right margin of rightmost side character block in one layer of license plate area is determined as undetermined The right margin of the license plate area of bit image；If it is not, then the right margin of initial license plate area to be determined as to the vehicle of bit image undetermined The right margin in board region.
As it can be seen that in the present embodiment, it, can also be according to each in the case that the license plate in initial license plate area is single layer license plate The left and right boundary of one layer of license plate area where a character block, determines the left and right boundary of the license plate area of bit image undetermined, this Sample can be improved the accuracy of positioned license plate area.
Fig. 4 is a kind of structural schematic diagram of license plate positioning device provided by the embodiments of the present application, is applied to electronic equipment.It should Embodiment is corresponding with embodiment of the method shown in Fig. 2.The apparatus may include:
Initial alignment module 401 is examined for obtaining the bit image undetermined comprising license plate, and to the bit image undetermined It surveys, obtains initial license plate area；
Character block determining module 402, for determining each character block in the initial license plate area；
License plate judgment module 403, for judging whether the initial license plate area is more according to determining each character block Layer license plate area；
First locating module 404, for determining the license plate area of the bit image undetermined according at least one layer of license plate area Domain.
In a kind of embodiment of embodiment shown in Fig. 4, the initial alignment module 401 is specifically used for according to license plate The jump feature of pixel value detects the bit image undetermined, obtains initial license plate area；Alternatively,
The initial alignment module, specifically for the pixel value tag according to license plate edge, to the bit image undetermined into Row detection, obtains initial license plate area.
In a kind of embodiment of embodiment shown in Fig. 4, the character block determining module 402 may include:
Binaryzation submodule (not shown), for the picture using preset different threshold values to the initial license plate area Element value carries out binary conversion treatment, obtains each treated initial license plate image；
Region determines submodule (not shown), for determining pixel value from each treated initial license plate image Constant each connected region；
Character block determines submodule (not shown), for image corresponding to the external frame by each connected region Region, each character block being determined as in the initial license plate area.
In a kind of embodiment of embodiment shown in Fig. 4, the character block determining module 402 is specifically used for:
Determine each doubtful character block in the initial license plate area, according to the size characteristic of preset characters on license plate or The distance between preset characters on license plate feature filters out character block from each doubtful character block.
In a kind of embodiment of embodiment shown in Fig. 4, which can also include:
Quantity judgment module (not shown), for determine each character block in the initial license plate area it Afterwards, judge whether the total quantity of each character block is greater than preset quantity threshold value, if it is, executing the license plate judgment module.
In a kind of embodiment of embodiment shown in Fig. 4, the first locating module 404 may include:
Up-and-down boundary determines submodule (not shown), each character for that will include at least one layer of license plate area The coboundary of block and lower boundary are identified as coboundary and the lower boundary of the license plate area of the bit image undetermined；
Left margin determines submodule (not shown), for judging leftmost side character block at least one layer of license plate area Whether the distance between left margin and the initial license plate area left margin are greater than pre-determined distance threshold value；If it is, will at least The left margin of leftmost side character block is determined as the left margin of the license plate area of the bit image undetermined in one layer of license plate area；If It is no, then the left margin of the initial license plate area is determined as to the left margin of the license plate area of the bit image undetermined；
Right margin determines submodule (not shown), for judging rightmost side character block at least one layer of license plate area Whether the distance between right margin and the initial license plate area right margin are greater than pre-determined distance threshold value；If it is, will at least The right margin of rightmost side character block is determined as the right margin of the license plate area of the bit image undetermined in one layer of license plate area；If It is no, then the right margin of the initial license plate area is determined as to the right margin of the license plate area of the bit image undetermined.
In a kind of embodiment of embodiment shown in Fig. 4, first locating module 404 is specifically used for:
Each layer license plate area for including by the initial license plate area, is identified as the license plate area of the bit image undetermined Domain；Alternatively,
First locating module 404 is determined specifically for one layer of most license plate area of the character number of blocks that will include For the license plate area of the bit image undetermined.
In a kind of embodiment of embodiment shown in Fig. 4, described device can also include:
Second locating module (not shown) judges that the initial license plate area is not multilayer license plate area for working as When, the initial license plate area is determined as to the license plate area of the bit image undetermined.
In a kind of embodiment of embodiment shown in Fig. 4, described device can also include:
As a result output module (not shown), for will wrap after the license plate area for determining the bit image undetermined The apex coordinate of the boundary rectangle of the character block containing same layer is exported as positioning result.
Since above-mentioned apparatus embodiment is obtained based on embodiment of the method, and this method technical effect having the same, Therefore details are not described herein for the technical effect of Installation practice.For device embodiment, since it is substantially similar to method Embodiment, so describing fairly simple, the relevent part can refer to the partial explaination of embodiments of method.
Fig. 5 is a kind of structural schematic diagram of electronic equipment provided by the embodiments of the present application.The electronic equipment includes processor 501, memory 502；
The memory 502, for storing executable code；
The processor 501, by reading the executable code stored in memory 502, for executing the application reality The license plate locating method of example offer is provided.This method comprises:
The bit image undetermined comprising license plate is obtained, and the bit image undetermined is detected, obtains initial license plate area；
Determine each character block in the initial license plate area；
According to determining each character block, judge whether the initial license plate area is multilayer license plate area；
If it is, determining the license plate area of the bit image undetermined according at least one layer of license plate area.
The memory that above-mentioned electronic equipment is mentioned may include random access memory (Random Access Memory, It RAM), also may include nonvolatile memory (Non-Volatile Memory, NVM), for example, at least a disk storage Device.Optionally, memory can also be that at least one is located remotely from the storage device of aforementioned processor.
Above-mentioned processor can be general processor, including central processing unit (Central Processing Unit, CPU), network processing unit (Network Processor, NP) etc.；It can also be digital signal processor (Digital Signal Processing, DSP), it is specific integrated circuit (Application Specific Integrated Circuit, ASIC), existing It is field programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic device, discrete Door or transistor logic, discrete hardware components.
The present embodiment is the embodiment obtained based on the same inventive concept with embodiment of the method shown in Fig. 2, the present embodiment Specific embodiment can be using the content in embodiment of the method shown in Fig. 2.
As it can be seen that the present embodiment can determine each character block from initial license plate area, according to the position of each character block Distribution, judges whether initial license plate area is multilayer license plate area, without the largely double-deck plate templates, therefore energy of matching one by one Enough improve the efficiency of License Plate process.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or any other variant be intended to it is non- It is exclusive to include, so that the process, method, article or equipment for including a series of elements not only includes those elements, It but also including other elements that are not explicitly listed, or further include solid by this process, method, article or equipment Some elements.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including There is also other identical elements in the process, method, article or equipment of the element.
Each embodiment in this specification is all made of relevant mode and describes, same and similar portion between each embodiment Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.
The foregoing is merely the preferred embodiments of the application, are not intended to limit the protection scope of the application.It is all Any modification, equivalent substitution, improvement and etc. done within spirit herein and principle are all contained in the protection scope of the application It is interior.

Claims (19)

1. a kind of license plate locating method, which is characterized in that the described method includes:
The bit image undetermined comprising license plate is obtained, and the bit image undetermined is detected, obtains initial license plate area；
Determine each character block in the initial license plate area；
According to determining each character block, judge whether the initial license plate area is multilayer license plate area；
If it is, determining the license plate area of the bit image undetermined according at least one layer of license plate area.
2. being obtained just the method according to claim 1, wherein described detect the bit image undetermined The step of beginning license plate area, comprising:
According to the jump feature of license plate pixel value, the bit image undetermined is detected, initial license plate area is obtained；Alternatively,
According to the pixel value tag at license plate edge, the bit image undetermined is detected, initial license plate area is obtained.
3. the method according to claim 1, wherein each character in the determination initial license plate area The step of block, comprising:
Binary conversion treatment is carried out to the pixel value of the initial license plate area using preset different threshold values, after obtaining each processing Initial license plate image；
The constant each connected region of pixel value is determined from each treated initial license plate image；
Image-region corresponding to external frame by each connected region, each word being determined as in the initial license plate area Accord with block.
4. the method according to claim 1, wherein each character in the determination initial license plate area The step of block, comprising:
Determine each doubtful character block in the initial license plate area；
According to the distance between the size characteristic of preset characters on license plate or preset characters on license plate feature, from each doubtful character Character block is filtered out in block.
5. the method according to claim 1, wherein determining each character block in the initial license plate area Later, the method also includes:
Judge whether the total quantity of each character block is greater than preset quantity threshold value；
If it is, execution is described according to determining each character block, judge whether the initial license plate area is multilayer license plate The step of region.
6. the method according to claim 1, wherein described according at least one layer of license plate area, determine it is described to The step of positioning the license plate area of image, comprising:
By the coboundary for each character block for including at least one layer of license plate area and lower boundary, it is identified as described to be positioned The coboundary of the license plate area of image and lower boundary；
Judge at least one layer of license plate area between the left margin and the initial license plate area left margin of leftmost side character block Whether distance is greater than pre-determined distance threshold value；If it is, the left margin of leftmost side character block at least one layer of license plate area is true It is set to the left margin of the license plate area of the bit image undetermined；If it is not, then the left margin of the initial license plate area is determined For the left margin of the license plate area of the bit image undetermined；
Judge at least one layer of license plate area between the right margin and the initial license plate area right margin of rightmost side character block Whether distance is greater than pre-determined distance threshold value；If it is, the right margin of rightmost side character block at least one layer of license plate area is true It is set to the right margin of the license plate area of the bit image undetermined；If it is not, then the right margin of the initial license plate area is determined For the right margin of the license plate area of the bit image undetermined.
7. the method according to claim 1, wherein described according at least one layer of license plate area, determine it is described to The step of positioning the license plate area of image, comprising:
Each layer license plate area for including by the initial license plate area, is identified as the license plate area of the bit image undetermined； Alternatively,
By comprising the most one layer of license plate area of character number of blocks be determined as the license plate area of the bit image undetermined.
8. described in any item methods according to claim 1~7, which is characterized in that when judging the initial license plate area not When for multilayer license plate area, the method also includes:
The initial license plate area is determined as to the license plate area of the bit image undetermined.
9. described in any item methods according to claim 1~7, which is characterized in that in the license plate for determining the bit image undetermined After region, the method also includes:
The apex coordinate of boundary rectangle comprising same layer character block is exported as positioning result.
10. a kind of license plate positioning device, which is characterized in that described device includes:
Initial alignment module is detected for obtaining the bit image undetermined comprising license plate, and to the bit image undetermined, is obtained Initial license plate area；
Character block determining module, for determining each character block in the initial license plate area；
License plate judgment module, for judging whether the initial license plate area is multilayer license plate according to determining each character block Region；
First locating module, for determining the license plate area of the bit image undetermined according at least one layer of license plate area.
11. device according to claim 10, which is characterized in that the initial alignment module is specifically used for:
According to the jump feature of license plate pixel value, the bit image undetermined is detected, initial license plate area is obtained；Alternatively,
The initial alignment module, is specifically used for:
According to the pixel value tag at license plate edge, the bit image undetermined is detected, initial license plate area is obtained.
12. device according to claim 10, which is characterized in that the character block determining module, comprising:
Binaryzation submodule, for being carried out at binaryzation using preset different threshold values to the pixel value of the initial license plate area Reason obtains each treated initial license plate image；
Region determines submodule, for determining the constant each connected region of pixel value from each treated initial license plate image Domain；
Character block determines submodule, for image-region corresponding to the external frame by each connected region, is determined as described Each character block in initial license plate area.
13. device according to claim 10, which is characterized in that the character block determining module is specifically used for:
It determines each doubtful character block in the initial license plate area, according to the size characteristic of preset characters on license plate or presets The distance between characters on license plate feature, filter out character block from each doubtful character block.
14. device according to claim 10, which is characterized in that further include:
Quantity judgment module, for judging each character block after determining each character block in the initial license plate area Total quantity whether be greater than preset quantity threshold value, if it is, executing the license plate judgment module.
15. device according to claim 10, which is characterized in that first locating module, comprising:
Up-and-down boundary determines submodule, the coboundary of each character block for that will include at least one layer of license plate area and following Boundary is identified as coboundary and the lower boundary of the license plate area of the bit image undetermined；
Left margin determines submodule, for judging at least one layer of license plate area the left margin of leftmost side character block and described initial Whether the distance between license plate area left margin is greater than pre-determined distance threshold value；If it is, by least one layer of license plate area most The left margin of left side character block is determined as the left margin of the license plate area of the bit image undetermined；If it is not, then by described initial The left margin of license plate area is determined as the left margin of the license plate area of the bit image undetermined；
Right margin determines submodule, for judging at least one layer of license plate area the right margin of rightmost side character block and described initial Whether the distance between license plate area right margin is greater than pre-determined distance threshold value；If it is, by least one layer of license plate area most The right margin of right side character block is determined as the right margin of the license plate area of the bit image undetermined；If it is not, then by described initial The right margin of license plate area is determined as the right margin of the license plate area of the bit image undetermined.
16. the device according to claim, which is characterized in that first locating module is specifically used for:
Each layer license plate area for including by the initial license plate area, is identified as the license plate area of the bit image undetermined； Alternatively,
First locating module, is specifically used for:
By comprising the most one layer of license plate area of character number of blocks be determined as the license plate area of the bit image undetermined.
17. 0~16 described in any item devices according to claim 1, which is characterized in that described device further include:
Second locating module, for when judging the initial license plate area is not multilayer license plate area, by the initial vehicle Board region is determined as the license plate area of the bit image undetermined.
18. 0~16 described in any item methods according to claim 1, which is characterized in that described device further include:
As a result output module, for same layer character block will to be included after the license plate area for determining the bit image undetermined The apex coordinate of boundary rectangle is exported as positioning result.
19. a kind of electronic equipment, which is characterized in that including processor, memory；
The memory, for storing executable code；
The processor, by reading the executable code stored in memory, to require any one of 1-9 institute for perform claim The license plate locating method stated.
CN201710656000.0A 2017-08-03 2017-08-03 License plate positioning method and device Active CN109389122B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710656000.0A CN109389122B (en) 2017-08-03 2017-08-03 License plate positioning method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710656000.0A CN109389122B (en) 2017-08-03 2017-08-03 License plate positioning method and device

Publications (2)

Publication Number Publication Date
CN109389122A true CN109389122A (en) 2019-02-26
CN109389122B CN109389122B (en) 2021-03-12

Family

ID=65412383

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710656000.0A Active CN109389122B (en) 2017-08-03 2017-08-03 License plate positioning method and device

Country Status (1)

Country Link
CN (1) CN109389122B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111027544A (en) * 2019-11-29 2020-04-17 武汉虹信技术服务有限责任公司 MSER license plate positioning method and system based on visual saliency detection
CN112446375A (en) * 2019-09-03 2021-03-05 上海高德威智能交通系统有限公司 License plate recognition method, device, equipment and storage medium
CN112651417A (en) * 2019-10-12 2021-04-13 杭州海康威视数字技术股份有限公司 License plate recognition method, device, equipment and storage medium
CN112686246A (en) * 2019-10-18 2021-04-20 深圳市优必选科技股份有限公司 License plate character segmentation method and device, storage medium and terminal equipment
CN115909526A (en) * 2022-11-29 2023-04-04 广州柏瀚信息科技有限公司 Highway toll collection method and device

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080212837A1 (en) * 2007-03-02 2008-09-04 Canon Kabushiki Kaisha License plate recognition apparatus, license plate recognition method, and computer-readable storage medium
US20120275653A1 (en) * 2011-04-28 2012-11-01 Industrial Technology Research Institute Method for recognizing license plate image, and related computer program product, computer-readable recording medium, and image recognizing apparatus using the same
CN103279736A (en) * 2013-04-27 2013-09-04 电子科技大学 License plate detection method based on multi-information neighborhood voting
CN103530608A (en) * 2013-10-10 2014-01-22 智博锐视（北京）科技有限公司 Vehicle type judgment method and vehicle type judgment device
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
CN105335743A (en) * 2015-10-28 2016-02-17 重庆邮电大学 Vehicle license plate recognition method
CN105528609A (en) * 2014-09-28 2016-04-27 江苏省兴泽实业发展有限公司 Vehicle license plate location method based on character position
CN105894004A (en) * 2014-12-16 2016-08-24 中防通用电信技术有限公司 Complement technology of locating incomplete license plate aiming at license plate identification system

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080212837A1 (en) * 2007-03-02 2008-09-04 Canon Kabushiki Kaisha License plate recognition apparatus, license plate recognition method, and computer-readable storage medium
US20120275653A1 (en) * 2011-04-28 2012-11-01 Industrial Technology Research Institute Method for recognizing license plate image, and related computer program product, computer-readable recording medium, and image recognizing apparatus using the same
CN103279736A (en) * 2013-04-27 2013-09-04 电子科技大学 License plate detection method based on multi-information neighborhood voting
CN103530608A (en) * 2013-10-10 2014-01-22 智博锐视（北京）科技有限公司 Vehicle type judgment method and vehicle type judgment device
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
CN105528609A (en) * 2014-09-28 2016-04-27 江苏省兴泽实业发展有限公司 Vehicle license plate location method based on character position
CN105894004A (en) * 2014-12-16 2016-08-24 中防通用电信技术有限公司 Complement technology of locating incomplete license plate aiming at license plate identification system
CN105335743A (en) * 2015-10-28 2016-02-17 重庆邮电大学 Vehicle license plate recognition method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
王云龙: "机动车号牌识别系统的算法研究及实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112446375A (en) * 2019-09-03 2021-03-05 上海高德威智能交通系统有限公司 License plate recognition method, device, equipment and storage medium
CN112651417A (en) * 2019-10-12 2021-04-13 杭州海康威视数字技术股份有限公司 License plate recognition method, device, equipment and storage medium
CN112651417B (en) * 2019-10-12 2024-06-14 杭州海康威视数字技术股份有限公司 License plate recognition method, device, equipment and storage medium
CN112686246A (en) * 2019-10-18 2021-04-20 深圳市优必选科技股份有限公司 License plate character segmentation method and device, storage medium and terminal equipment
CN112686246B (en) * 2019-10-18 2024-01-02 深圳市优必选科技股份有限公司 License plate character segmentation method and device, storage medium and terminal equipment
CN111027544A (en) * 2019-11-29 2020-04-17 武汉虹信技术服务有限责任公司 MSER license plate positioning method and system based on visual saliency detection
CN111027544B (en) * 2019-11-29 2023-09-29 武汉虹信技术服务有限责任公司 MSER license plate positioning method and system based on visual saliency detection
CN115909526A (en) * 2022-11-29 2023-04-04 广州柏瀚信息科技有限公司 Highway toll collection method and device
CN115909526B (en) * 2022-11-29 2023-10-20 广州柏瀚信息科技有限公司 Expressway charging method and device

Also Published As

Publication number Publication date
CN109389122B (en) 2021-03-12

Similar Documents

Publication Publication Date Title
CN109389122A (en) A kind of license plate locating method and device
CN111382704B (en) Vehicle line pressing violation judging method and device based on deep learning and storage medium
CN107545239B (en) Fake plate detection method based on license plate recognition and vehicle characteristic matching
CN102722704B (en) Method and system for recognizing vehicle license plate by integrating video dynamic tracking
CN109711264B (en) Method and device for detecting occupation of bus lane
KR101403876B1 (en) Method and Apparatus for Vehicle License Plate Recognition
CN108268867B (en) License plate positioning method and device
CN109993056A (en) A kind of method, server and storage medium identifying vehicle violation behavior
Asif et al. Multinational vehicle license plate detection in complex backgrounds
CN111027535B (en) License plate recognition method and related equipment
Paunwala et al. A novel multiple license plate extraction technique for complex background in Indian traffic conditions
CN106650553A (en) License plate recognition method and system
JPH11306283A (en) Number plate reader
Khalifa et al. Malaysian Vehicle License Plate Recognition.
CN108229466B (en) License plate recognition method and device
CN110378351B (en) Seal identification method and device
CN114387591A (en) License plate recognition method, system, equipment and storage medium
CN109034158B (en) License plate recognition method and device and computer equipment
Jagannathan et al. License plate character segmentation using horizontal and vertical projection with dynamic thresholding
Cao et al. An adaptive approach to vehicle license plate localization
CN107578011A (en) The decision method and device of key frame of video
CN112686252A (en) License plate detection method and device
Ullah et al. An approach of locating Korean vehicle license plate based on mathematical morphology and geometrical features
CN108205670B (en) License plate recognition method and device
CN107506777A (en) A kind of real-time more licence plate recognition methods and device based on Wavelet transformation and SVMs

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant