CN106780967B - A kind of bank note version recognition methods and device - Google Patents
A kind of bank note version recognition methods and device Download PDFInfo
- Publication number
- CN106780967B CN106780967B CN201710014330.XA CN201710014330A CN106780967B CN 106780967 B CN106780967 B CN 106780967B CN 201710014330 A CN201710014330 A CN 201710014330A CN 106780967 B CN106780967 B CN 106780967B
- Authority
- CN
- China
- Prior art keywords
- image
- binarization
- gray value
- line
- hand feel
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
- G07D7/00—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency
- G07D7/20—Testing patterns thereon
- G07D7/2016—Testing patterns thereon using feature extraction, e.g. segmentation, edge detection or Hough-transformation
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Image Analysis (AREA)
Abstract
The invention belongs to paper money recognition technical field, a kind of bank note version recognition methods and device are provided.It include: the image for obtaining the predeterminated position of bank note, generate binarization of gray value image corresponding with the image of predeterminated position, judge whether binarization of gray value image is feel line image, boundary alignment is carried out to the binarization of gray value image according to the distribution of white pixel point in binarization of gray value image when binarization of gray value image is for feel line image, to obtain hand feel line area image, the hand feel line area image is divided into straight line hand feel line area image and curve hand feel line area image by the distribution further according to white pixel point in hand feel line area image, it is last that the straight line hand feel line area image and the curve hand feel line area image are judged respectively according to default linearity region Parameter Conditions and pre-programmed curve region parameter condition, and determine whether the version of the bank note is target bank note version according to the judging result, to quasi- The version of true identification bank note.
Description
Technical field
The invention belongs to paper money recognition technical field more particularly to a kind of bank note version recognition methods and devices.
Background technique
With the development of economy, circulation of RMB amount is also corresponding increasing.The people for multiple versions that circulate on the market
Coin, that mainly circulates has the RMB of 1999 editions, 2005 editions.Law-breaker is to seek exorbitant profit, and manufactures the RMB of various versions
Counterfeit money endangers financial security.Due between the various versions of RMB may size it is identical, and background texture and rich in color,
Therefore its version can not be determined by paper size or color.For the ease of arranging and identifying a large amount of RMB, bank note
Therefore version automatic identification technology generates.However existing bank note version identification technology is carried out primarily directed to large-denomination bank note
It distinguishes, since difference existing between the various versions of note of small denominations is less, is not available above-mentioned high denomination banknotes version
This identification technology identifies the version of note of small denominations.Therefore, existing bank note version identification technology presence can not be to small
The problem of version of denominations is identified.
Summary of the invention
The present invention provides a kind of bank note version recognition methods and devices, it is intended to solve existing bank note version identification technology
There are problems that not identifying the version of note of small denominations.
The present invention provides a kind of bank note version recognition methods, the bank note version recognition methods includes:
Obtain the image of the predeterminated position of bank note;
Generate binarization of gray value image corresponding with the image of the predeterminated position;
The white pixel point quantity and black pixel point quantity for counting the binarization of gray value image, by the white pixel
Point quantity and the black pixel point quantity are compared with preset white pixel threshold and default black picture element threshold value respectively, and
Judge whether the binarization of gray value image is feel line image according to comparison result;
If so, according to the distribution of white pixel point in the binarization of gray value image to the binarization of gray value image into
Row bound positioning, to obtain hand feel line area image;
The hand feel line area image is divided into directly according to the distribution of white pixel point in the hand feel line area image
Line hand feel line area image and curve hand feel line area image;
According to default linearity region Parameter Conditions and pre-programmed curve region parameter condition respectively to straight line hand feel line area
Area image and the curve hand feel line area image judged, and determines that the version of the bank note is according to the judging result
No is target bank note version.
The present invention also provides a kind of bank note version recognition device, the bank note version recognition device includes:
Obtain module, the image of the predeterminated position for obtaining bank note;
Generation module, for generating binarization of gray value image corresponding with the image of the predeterminated position;
Judgment module, for counting the white pixel point quantity and black pixel point quantity of the binarization of gray value image,
By the white pixel point quantity and the black pixel point quantity respectively with preset white pixel threshold and default black picture element
Threshold value is compared, and judges whether the binarization of gray value image is feel line image according to comparison result;
Boundary alignment module, for the distribution according to white pixel point in the binarization of gray value image to the gray scale two
Value image carries out boundary alignment, to obtain hand feel line area image;
Divide module, for according to the distribution of white pixel point in the hand feel line area image by the hand feel line region
Image segmentation is straight line hand feel line area image and curve hand feel line area image;
Identification module presets linearity region Parameter Conditions and pre-programmed curve region parameter condition respectively to described for basis
Straight line hand feel line area image and the curve hand feel line area image are judged, determine the paper according to the judging result
Whether the version of coin is target bank note version.
The present invention generates gray scale two-value corresponding with the image of predeterminated position by the image of the predeterminated position of acquisition bank note
Change image and judges whether binarization of gray value image is feel line image again, the root when binarization of gray value image is for feel line image
Boundary alignment is carried out to the binarization of gray value image according to the distribution of white pixel point in the binarization of gray value image, to obtain
Hand feel line area image, further according to white pixel point in the hand feel line area image distribution by the hand feel line area image
It is divided into straight line hand feel line area image and curve hand feel line area image, it is last according to default linearity region Parameter Conditions and pre-
If curve regions Parameter Conditions respectively sentence the straight line hand feel line area image and the curve hand feel line area image
It is disconnected, and determine whether the version of the bank note is target bank note version according to the judging result, to accurately identify small amount
The version of bank note.
Detailed description of the invention
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
Needed in attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is some implementations of the invention
Example, for those of ordinary skill in the art, without any creative labor, can also be according to these attached drawings
Obtain other attached drawings.
Fig. 1 is the implementation flow chart for the bank note version recognition methods that the embodiment of the present invention one provides;
Fig. 2 is the front of different bank note versions under visible light in the bank note version recognition methods of the offer of the embodiment of the present invention one
Image;
Fig. 3 is the image of predeterminated position in the bank note version recognition methods of the offer of the embodiment of the present invention one;
Fig. 4 is the schematic diagram of binarization of gray value image in the bank note version recognition methods of the offer of the embodiment of the present invention one;
Fig. 5 is the structural schematic diagram of bank note version recognition device provided by Embodiment 2 of the present invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
The embodiment of the present invention provides a kind of bank note version recognition methods and dress to accurately identify the version of bank note
It sets, wherein the main image by obtaining the predeterminated position of bank note, generates binarization of gray value corresponding with the image of predeterminated position
Image judges whether binarization of gray value image is feel line image again, when binarization of gray value image is for feel line image according to
The distribution of white pixel point carries out boundary alignment to binarization of gray value image in the binarization of gray value image, to obtain hand feel line
Hand feel line area image is divided into straight line feel by area image, the distribution further according to white pixel point in hand feel line area image
Line area image and curve hand feel line area image, it is last according to default linearity region Parameter Conditions and pre-programmed curve region parameter
Condition respectively judges straight line hand feel line area image and curve hand feel line area image, and really states paper according to judging result
Whether the version of coin is target bank note version, to accurately identify the version of bank note.
In order to illustrate above-mentioned bank note version recognition methods and device, carried out specifically below in conjunction with specific embodiment
It is bright:
Embodiment one:
Fig. 1 shows the implementation process of the bank note version recognition methods of the offer of the embodiment of the present invention one, for ease of description,
Only the parts related to this embodiment are shown, and details are as follows:
In step s101, the image of the predeterminated position of bank note is obtained.
It should be noted that the image for obtaining the predeterminated position of bank note includes: the image for obtaining entire paper coin, then therefrom cut
Take the image of predeterminated position.Wherein, predeterminated position refers to the bank note front right hand edge middle-lower part position.As shown in Fig. 2, Fig. 2 a
For lower 1999 editions 10 yuans of the image of visible light, Fig. 2 b is lower 1999 editions 20 yuans of the image of visible light, and Fig. 2 c is
Lower 2005 editions 10 yuans of the image of visible light, Fig. 2 d are lower 2005 editions 20 yuans of the image of visible light, pass through interception
The image of predeterminated position obtains image as shown in Figure 3.Wherein, Fig. 3 a is lower 1999 editions 10 yuans of the default position of visible light
The image set, Fig. 3 b are the image of the lower 1999 editions 20 yuans of predeterminated positions of visible light, and Fig. 3 c is visible light lower 2005 editions 10
The image of yuan predeterminated position, Fig. 3 d are the image of the lower 2005 editions 20 yuans of predeterminated positions of visible light.It needs to illustrate
, obtain entire paper coin image can also be obtained under black light, such as by infrared transmission acquisition entire paper coin image.
In step s 102, binarization of gray value image corresponding with the image of predeterminated position is generated.
Step S102 specifically: to the image line binary conversion treatment of predeterminated position, generate binarization of gray value image.Wherein,
Binary conversion treatment refers to: the gray value of each pixel in the image of predeterminated position being respectively set to 0 or 255, i.e. image is set
It is set to the image for only existing two kinds of color values.Wherein, the threshold value judged when binary conversion treatment can be fixed, can also be according to paper
The type of coin is different and changes.
As shown in figure 4, Fig. 4 a is the corresponding gray scale two of image of lower 1999 editions 10 yuans of the predeterminated position of visible light
Value image, Fig. 4 b are the corresponding binarization of gray value image of image of the lower 1999 editions 20 yuans of predeterminated positions of visible light, figure
4c is the corresponding binarization of gray value image of image of the lower 2005 editions 10 yuans of predeterminated positions of visible light, and Fig. 4 d is under visible light
The corresponding binarization of gray value image of the image of 2005 editions 20 yuans of predeterminated positions.
In step s 103, the white pixel point quantity and black pixel point quantity for counting binarization of gray value image, will be white
Color pixel point quantity and black pixel point quantity are compared with preset white pixel threshold and default black picture element threshold value respectively,
And judge whether binarization of gray value image is feel line image according to comparison result.
Specifically, step S103 the following steps are included:
Count the white pixel point quantity and black pixel point quantity in binary image;
By white pixel point quantity and black pixel point quantity respectively with preset white pixel threshold and default black picture element
Threshold value is compared;
If white pixel point quantity is greater than preset white pixel threshold or black pixel point quantity is less than default black picture element
Threshold value, it is determined that the image of predeterminated position is feel line image;
If white pixel point quantity is not less than default black no more than preset white pixel threshold or black pixel point quantity
Pixel threshold, it is determined that the image of predeterminated position is not feel line image.
It should be noted that preset white pixel threshold and default black picture element threshold value refer to: in order to measure predeterminated position
Image whether be feel line image and preset white pixel point quantitative value and black pixel point quantitative value.It illustrates
Bright, white pixel point quantity is 20000 in the binarization of gray value image in most of 2005 editions 10 yuans of hand feel line region,
Preset white pixel threshold is then set as 20000.
In step S104, if so, according to the distribution of white pixel point in binarization of gray value image to binarization of gray value
Image carries out boundary alignment, to obtain hand feel line area image.
Specifically, step S104 the following steps are included:
Horizontal boundary positioning is carried out to binary image according to the genesis analysis situation of white pixel point in binary image;
Longitudinal boundary positioning is carried out to binary image according to the cross direction profiles situation of white pixel point in binary image;
Hand feel line area image is obtained according to the horizontal boundary of binary image and longitudinal boundary.
Further, binary image is carried out laterally according to the genesis analysis situation of white pixel point in binary image
Boundary alignment the following steps are included:
The quantity of the white pixel point in binary image on same positive value ordinate is calculated in two-dimensional coordinate system
The sum of with obtain corresponding in binary image one or more positive value ordinates column and;
It is compared by the column of one or more positive value ordinates and with the first preset boundary threshold value;
When column and it is greater than the first preset boundary threshold value, and arranging with corresponding positive value ordinate is the vertical seat of one or more positive values
Numerical value is maximum when ordinate in mark, and longitudinal linear position where determining column and corresponding positive value ordinate is binaryzation
The right margin of image;
The left margin of binary image is obtained according to the right margin of the first preset coordinate spacing and binary image.
Herein it should be noted that the first preset boundary threshold value refers to: longitudinal straight where positive value ordinate in order to measure
Line position is the value of the right margin of binary image and preset column sum.First preset coordinate spacing refers to: in order to obtain
The left margin of binary image and preset lateral coordinates spacing, the first preset coordinate spacing can be according to hand feel line
The hand feel line peak width of bank note set.
Further, binary image is indulged according to the cross direction profiles situation of white pixel point in binary image
To boundary alignment the following steps are included:
The quantity of the white pixel point in binary image on same positive value abscissa is calculated in two-dimensional coordinate system
The sum of with obtain in binary image for one or more positive value abscissas row and;
It is compared by the row of one or more positive value abscissas and with the second preset boundary threshold value;
When row and it is greater than the second preset boundary threshold value, and capable and corresponding positive value abscissa is one or more on the occasion of horizontal seat
In mark numerical value it is maximum positive value abscissa when, determine row and for positive value abscissa where lateral linear position be binaryzation
The coboundary of image;
The lower boundary of binary image is obtained according to the coboundary of the second preset coordinate spacing and binary image.
It should be noted that the second preset boundary threshold value refers to: the lateral straight line in order to measure place on positive value abscissa
Position is the value of the coboundary of binary image and preset row sum.First preset coordinate spacing refers to: in order to obtain two
The lower boundary of value image and preset longitudinal coordinate spacing, the second preset coordinate spacing can be according to hand feel line
The hand feel line region height of bank note is set.
In step s105, hand feel line area image is divided according to the distribution of white pixel point in hand feel line area image
For straight line hand feel line area image and curve hand feel line area image.
Specifically, step S105 the following steps are included:
The quantity of the white pixel point in binary image on same positive value ordinate is calculated in two-dimensional coordinate system
The sum of with obtain corresponding in binary image one or more positive value ordinates column and;
It is compared by the column of one or more positive value ordinates and with default segmentation threshold;
When column and it is less than default segmentation threshold, and column and corresponding positive value ordinate are 1 or the vertical seat of multiple positive values
Numerical value is the smallest when ordinate in mark, and longitudinal linear position where determining the column and corresponding positive value ordinate is feel
The partitioning boundary of line area image;
Hand feel line area image is divided into straight line hand feel line area image and curve hand feel line region according to partitioning boundary
Image.
It should be noted that default segmentation threshold refers to: the value in order to determine partitioning boundary and preset column sum.
In step s 106, according to default linearity region Parameter Conditions and pre-programmed curve region parameter condition respectively to straight line
Hand feel line area image and curve hand feel line area image judged, and according to judging result determine bank note version whether be
Target bank note version.
Specifically, step S106 the following steps are included:
It calculates the straight-line segment number of straight line hand feel line area image and obtains starting point coordinate and the end of all straight-line segments
Point coordinate;
Judge whether the number of the straight-line segment of straight line hand feel line region binary image is equal to according to straight-line segment number
Default straight-line segment number threshold value;
If so, judging whether the length of straight-line segment is equal to default straight line according to the starting point coordinate of straight line and terminal point coordinate
Line segment length threshold value;
If so, judging the vertical interval between all straight-line segments according to the starting point coordinate of straight-line segment and terminal point coordinate
Whether preset vertical spacing threshold values is equal to;
If so, on the same line whether the starting point coordinate and terminal point coordinate that judge straight-line segment;
If so, determining that straight line hand feel line area image meets linearity region preset condition;
Whether the pixel matching quantity between judgment curves hand feel line area image and default template image is greater than default picture
Plain matching threshold;
If so, the abscissa of the terminal of the curved segments in judgment curves hand feel line area image and straight line hand feel line area
Whether the abscissa of the starting point of the straight-line segment in area image is equal;
If so, determining that the version of bank note is target bank note version.
It should be noted that default straight-line segment number threshold value refers to: in order to measure whether bank note version is target bank note
The number of version and preset straight-line segment.Default straight-line segment length threshold refers to: in order to whether measure bank note version
The length of preset straight-line segment for target bank note version.Preset vertical spacing threshold values refers to: in order to measure bank note version
This whether be target bank note version and preset straight-line segment vertical interval.Presetted pixel matching threshold refers to, in order to
Measure whether bank note version is target bank note version and preset matched pixel point quantitative value.
It should be noted that between the curve hand feel line area image and template image of the bank note matched pixel point judgement
Judged by the quantity of pixel and the position of pixel.
For example, if target bank note version is 2005 editions, if bank note to be identified is 2005 editions 10 yuans.2005
The straight line hand feel line area image of 10 yuans of version has 12 straight-line segments, and the length of straight-line segment is about 1.2cm, straight line line
Vertical interval between section is about 1.5cm, therefore sets 12 for default straight-line segment number threshold value, presets straight-line segment length threshold
Value is set as 1.2, and preset vertical spacing threshold values is set as 1.5, and setting presetted pixel matching threshold is 25000.Judging the paper
When whether the version of coin is 2005 editions, the straight-line segment number of straight line hand feel line area image is calculated first and obtains all straight lines
The starting point coordinate and terminal point coordinate of line segment, if straight-line segment number is 12, the length of straight-line segment is 1.2cm, straight-line segment it
Between vertical interval be 1.5cm and the abscissa of the starting point of straight-line segment and the abscissa of terminal it is equal, it is determined that the bank note
The straight line hand feel line area image that straight line hand feel line area image is 2005 editions 10 yuans, then statistic curve hand feel line region
The quantitative value of matched pixel point between image and template image, if matching picture between curve hand feel line area image and template image
The quantitative value of vegetarian refreshments is greater than 25000, it is determined that the song that the curve hand feel line area image of the bank note is 2005 editions 10 yuans
Line hand feel line area image, finally, if the abscissa of the terminal of the curved segments in curve hand feel line area image and straight line hand
The abscissa for feeling the starting point of the straight-line segment in line area image is equal, it is determined that the bank note is 2005 editions 10 yuans.
Pass through default straight-line segment number threshold value, default straight-line segment length threshold, preset vertical spacing threshold values and straight
The starting point coordinate and terminal point coordinate of line line segment judge whether straight line hand feel line area image meets default linearity region Parameter Conditions,
Whether pre-programmed curve region parameter condition is met by presetted pixel matching threshold judgment curves hand feel line area image again, finally
Again the abscissa of the terminal of the curved segments in judgment curves hand feel line area image with it is straight in straight line hand feel line area image
Whether the abscissa of the starting point of line line segment is equal, in the case that appeal judging result is to be, determines that bank note version is target
Bank note version.
In the embodiment of the present invention, the image of the predeterminated position by obtaining bank note is generated corresponding with the image of predeterminated position
Binarization of gray value image judge whether binarization of gray value image is feel line image again, when binarization of gray value image is for feel
Boundary alignment is carried out to binarization of gray value image according to the distribution of white pixel point in binarization of gray value image when line image, with
To hand feel line area image, hand feel line area image is divided by the distribution further according to white pixel point in hand feel line area image
Straight line hand feel line area image and curve hand feel line area image pass through default straight-line segment number threshold value, default straight-line segment
Length threshold, the starting point coordinate of preset vertical spacing threshold values and straight-line segment and terminal point coordinate judge straight line hand feel line administrative division map
Seem it is no meet default linearity region Parameter Conditions, then be by presetted pixel matching threshold judgment curves hand feel line area image
It is no to meet pre-programmed curve region parameter condition, the last cross of the terminal of the curved segments in judgment curves hand feel line area image again
Whether coordinate and the abscissa of the starting point of the straight-line segment in straight line hand feel line area image are equal, when appeal judging result is
In the case where being, determine that bank note version is target bank note version, to accurately identify the version of bank note, the bank note of different values of money
The above method can be used to be identified.
Embodiment two:
Bank note version recognition methods shown in FIG. 1 in embodiment one to realize the present invention, present embodiments provides a kind of paper
Coin version recognition device.Fig. 5 shows the structure of the bank note version recognition device, for ease of description, illustrates only and this reality
The relevant part of example is applied, details are as follows:
Bank note version recognition device 50 includes obtaining module 501, generation module 502, judgment module 503, boundary alignment mould
Block 504, segmentation module 505 and identification module 506.
Obtain the image that module 501 is used to obtain the predeterminated position of bank note.Specifically, obtaining module 501 first obtains whole
The image of bank note, then therefrom intercept the image of predeterminated position.
Generation module 502 is for generating binarization of gray value image corresponding with the image of predeterminated position.Generation module 502 is right
The image line binary conversion treatment of predeterminated position generates binarization of gray value image.Binary conversion treatment specifically: by the figure of predeterminated position
The gray value of each pixel as in is respectively set to 0 or 255, i.e. image is set as only existing the image of two kinds of color values.Wherein,
The threshold value judged when binary conversion treatment can be fixed, and can also be changed according to the type difference of bank note.
Judgment module 503 is used to count the white pixel point quantity and black pixel point quantity of binarization of gray value image, will
White pixel point quantity and black pixel point quantity are compared with preset white pixel threshold and default black picture element threshold value respectively
Compared with, and judge whether binarization of gray value image is feel line image according to comparison result.
Boundary alignment module 504 is used for the distribution according to white pixel point in binarization of gray value image to binarization of gray value figure
As carrying out boundary alignment, to obtain hand feel line area image.
Divide module 505 to be used for hand feel line area image minute according to the distribution of white pixel point in hand feel line area image
It is segmented into straight line hand feel line area image and curve hand feel line area image.
Identification module 506 is used for according to default linearity region Parameter Conditions and pre-programmed curve region parameter condition respectively to straight
Line hand feel line area image and curve hand feel line area image judged, according to judging result determine bank note version whether be
Target bank note version.
Specifically, judgment module 503 includes statistic unit, comparing unit and judging unit.
Statistic unit is used to count white pixel point quantity and black picture element in the binary image in binary image
Point quantity;
Comparing unit be used for by white pixel point quantity and black pixel point quantity respectively with preset white pixel threshold and
Default black picture element threshold value is compared;
If judging unit is greater than preset white pixel threshold for white pixel point quantity or black pixel point quantity is less than
Default black picture element threshold value, it is determined that the image of predeterminated position is feel line image;If white pixel point quantity is no more than default
White pixel threshold value or black pixel point quantity are not less than default black picture element threshold value, it is determined that the image of predeterminated position is not
Feel line image.
Further, boundary alignment module 504 includes horizontal boundary positioning unit, longitudinal boundary positioning unit and hand feel line
Area image acquiring unit.
Horizontal boundary positioning unit is used for the genesis analysis situation according to white pixel point in binary image to binaryzation
Image carries out horizontal boundary positioning;
Longitudinal boundary positioning unit is used for the cross direction profiles situation according to white pixel point in binary image to binaryzation
Image carries out longitudinal boundary positioning;
Hand feel line area image acquiring unit is used to obtain feel according to the horizontal boundary and longitudinal boundary of binary image
Line area image.
Further, horizontal boundary positioning unit includes that column and computing unit, first row and comparing unit, the right are defined
Bit location and left margin positioning unit.
Column and computing unit are used to calculate in two-dimensional coordinate system in binary image on same positive value ordinate
The sum of quantity of white pixel point with obtain corresponding in binary image one or more positive value ordinates column and;
First row and comparing unit are used to carry out by the column of one or more positive value ordinates and with the first preset boundary threshold value
Compare;
Right margin positioning unit is used for when column and is greater than the first preset boundary threshold value, and column and corresponding positive value ordinate are
Numerical value is maximum when ordinate in one or more positive value ordinates, determines column and indulging where corresponding positive value ordinate
It is the right margin of binary image to linear position;
Left margin positioning unit is used to obtain binaryzation according to the right margin of the first preset coordinate spacing and binary image
The left margin of image.
Further, longitudinal boundary positioning unit includes capable and computing unit, row and comparing unit, coboundary positioning list
Member and lower boundary positioning unit.
Capable and computing unit is used to calculate in two-dimensional coordinate system in binary image on same positive value abscissa
The sum of quantity of white pixel point with obtain in binary image for one or more positive value abscissas row and;
Capable and comparing unit is used to compare by the row of one or more positive value abscissas and with the second preset boundary threshold value
Compared with;
Coboundary positioning unit is used for when row and is greater than the second preset boundary threshold value, and capable and corresponding positive value abscissa is
One or more positive value abscissas in numerical value it is maximum positive value abscissa when, determine row and for positive value abscissa where cross
It is the coboundary of binary image to linear position;
Lower boundary positioning unit is used to obtain binaryzation according to the coboundary of the second preset coordinate spacing and binary image
The lower boundary of image.
Further, cutting unit 505 includes column and computing unit, secondary series and comparing unit, partitioning boundary positioning list
Member and cutting unit.
Column and computing unit are used to calculate in two-dimensional coordinate system in binary image on same positive value ordinate
The sum of quantity of white pixel point with obtain corresponding in binary image one or more positive value ordinates column and;
Secondary series and comparing unit are used to compare by the column of one or more positive value ordinates and with default segmentation threshold
Compared with;
Partitioning boundary positioning unit is used to work as column and is less than default segmentation threshold, and arranges and be with corresponding positive value ordinate
1 or it is multiple positive value ordinate in numerical value it is the smallest positive value ordinate when, determine column and it is corresponding positive value ordinate where
Longitudinal linear position is the partitioning boundary of hand feel line area image;
Cutting unit is used to that hand feel line area image to be divided into straight line hand feel line area image and song according to partitioning boundary
Line hand feel line area image.
Further, identification module 506 include coordinate acquiring unit, number judging unit, length determining unit, it is vertical between
Away from judging unit, straight line judging unit, curve regions judging unit, horizontal line judging unit and recognition unit.
Coordinate acquiring unit is used to calculate the straight-line segment number of straight line hand feel line area image and obtains all straight line lines
The start-stop point coordinate of section;
Number judging unit is used to judge according to straight-line segment number the straight line line of straight line hand feel line region binary image
Whether the number of section is equal to default straight line number threshold value;
Length determining unit is used for when the judging result of number judging unit is to be, then according to the starting point coordinate of straight line and
Terminal point coordinate judges whether the length of straight-line segment is equal to default straight-line segment length threshold;
Vertical interval judging unit is used for when the judging result of length determining unit is to be, then rising according to straight-line segment
Point coordinate and terminal point coordinate judge whether the vertical interval of all straight-line segments is equal to preset vertical spacing threshold values;
Straight line judging unit is used to then judge that the starting point of straight line is sat when the judging result of vertical interval judging unit is to be
On the same line whether mark and terminal point coordinate;If so, straight line hand feel line area image meets linearity region preset condition;
Curve regions judging unit is used for matched pixel between judgment curves hand feel line area image and default template image
Whether quantity is greater than presetted pixel matching threshold;
Horizontal line judging unit is used for when the judging result of curve regions judging unit is to be, then judgment curves hand feel line
The abscissa of the terminal point coordinate of curved segments in area image and the starting point of the straight-line segment in straight line hand feel line area image
Whether the abscissa of coordinate is equal;
Recognition unit is used for when the judging result of horizontal line judging unit is to be, it is determined that the version of bank note is target basis
Coin version.
Those skilled in the art can be understood that, for convenience of description and succinctly, only with above-mentioned each function
The division progress of module can according to need and for example, in practical application by above-mentioned function distribution by different function lists
Member, module complete, i.e., the internal structure of bank note version recognition device is divided into different functional unit or module, with complete with
The all or part of function of upper description.Each functional module in embodiment can integrate in one processing unit, can also be with
It is that each unit physically exists alone, can also be integrated in one unit with two or more units, above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.In addition, each functional module
Specific name is also only for convenience of distinguishing each other, the protection scope being not intended to limit this application.
It should be noted that bank note version recognition device provided in an embodiment of the present invention, as with side shown in Fig. 1 of the present invention
Method embodiment is based on same design, and bring technical effect is identical as embodiment of the method shown in Fig. 1 of the present invention, and particular content can
Referring to the narration in embodiment of the method shown in Fig. 1 of the present invention, details are not described herein again.
Therefore, safety door prohibition system provided in this embodiment equally can by obtain bank note predeterminated position image,
It generates binarization of gray value image corresponding with the image of predeterminated position and judges whether binarization of gray value image is feel line image again,
When binarization of gray value image is for feel line image according to the distribution of white pixel point in binarization of gray value image to gray scale two
Value image carries out boundary alignment, to obtain hand feel line area image, further according to white pixel point in hand feel line area image
Hand feel line area image is divided into straight line hand feel line area image and curve hand feel line area image by distribution, by presetting straight line
Line segment number threshold value, default straight-line segment length threshold, the starting point coordinate of preset vertical spacing threshold values and straight-line segment and end
Point coordinate judges whether straight line hand feel line area image meets default linearity region Parameter Conditions, then matches threshold by presetted pixel
Whether value judgment curves hand feel line area image meets pre-programmed curve region parameter condition, finally judgment curves hand feel line region again
The abscissa of the starting point of the abscissa and straight-line segment in straight line hand feel line area image of the terminal of curved segments in image
It is whether equal, in the case that appeal judging result is to be, determine that bank note version is target bank note version, to accurately know
The version of other bank note.
The above is merely preferred embodiments of the present invention, be not intended to limit the invention, it is all in spirit of the invention and
Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within principle.
Claims (14)
1. a kind of bank note version recognition methods, which is characterized in that the bank note version recognition methods includes:
Obtain the image of the predeterminated position of bank note;
Generate binarization of gray value image corresponding with the image of the predeterminated position;
The white pixel point quantity and black pixel point quantity for counting the binarization of gray value image, the white pixel is counted
Amount and the black pixel point quantity are compared with preset white pixel threshold and default black picture element threshold value respectively, and according to
Comparison result judges whether the binarization of gray value image is feel line image;
If so, carrying out side to the binarization of gray value image according to the distribution of white pixel point in the binarization of gray value image
Position is defined, to obtain hand feel line area image;
The hand feel line area image is divided into straight line hand according to the distribution of white pixel point in the hand feel line area image
Feel line area image and curve hand feel line area image;
According to default linearity region Parameter Conditions and pre-programmed curve region parameter condition respectively to the straight line hand feel line administrative division map
Picture and the curve hand feel line area image are judged, and determine whether the version of the bank note is target according to judging result
Bank note version.
2. bank note version recognition methods according to claim 1, which is characterized in that count the binarization of gray value image
White pixel point quantity and black pixel point quantity, by the white pixel point quantity and the black pixel point quantity respectively with
Preset white pixel threshold and default black picture element threshold value are compared, and judge the binarization of gray value figure according to comparison result
Seem no includes: for feel line image
Count white pixel point quantity and the black pixel point quantity in the binarization of gray value image;
By the white pixel point quantity and the black pixel point quantity respectively with preset white pixel threshold and default black
Pixel threshold is compared;
If the white pixel point quantity is greater than the preset white pixel threshold or the black pixel point quantity less than default
Black picture element threshold value, it is determined that the image of the predeterminated position is feel line image;
If the white pixel point quantity is not less than no more than the preset white pixel threshold or the black pixel point quantity
Default black picture element threshold value, it is determined that the image of the predeterminated position is not feel line image.
3. bank note version recognition methods according to claim 1, which is characterized in that described according to the binarization of gray value figure
As white pixel point distribution to the binarization of gray value image carry out boundary alignment, obtaining hand feel line area image includes:
The binarization of gray value image is carried out according to the genesis analysis situation of white pixel point in the binarization of gray value image
Horizontal boundary positioning;
The binarization of gray value image is carried out according to the cross direction profiles situation of white pixel point in the binarization of gray value image
Longitudinal boundary positioning;
The hand feel line area image is obtained according to the horizontal boundary of the binarization of gray value image and longitudinal boundary.
4. bank note version recognition methods according to claim 3, which is characterized in that described according to the binarization of gray value figure
The genesis analysis situation of white pixel point includes: to binarization of gray value image progress horizontal boundary positioning as in
The white pixel point in the binarization of gray value image on same positive value ordinate is calculated in two-dimensional coordinate system
The sum of quantity with obtain corresponding in the binarization of gray value image one or more positive value ordinates column and;
It is compared by the column of one or more positive value ordinates and with the first preset boundary threshold value;
When it is described column and be greater than the first preset boundary threshold value, and it is described column and it is corresponding positive value ordinate be described 1 or
Numerical value is maximum when ordinate in multiple positive value ordinates, determines the longitudinal direction where the column and corresponding positive value ordinate
Linear position is the right margin of the binarization of gray value image;
The binarization of gray value image is obtained according to the right margin of the first preset coordinate spacing and the binarization of gray value image
Left margin.
5. bank note version recognition methods according to claim 3, which is characterized in that described according to the binarization of gray value figure
The cross direction profiles situation of white pixel point includes: to binarization of gray value image progress longitudinal boundary positioning as in
The white pixel point in the binarization of gray value image on same positive value abscissa is calculated in two-dimensional coordinate system
The sum of quantity with obtain in the binarization of gray value image for one or more positive value abscissas row and;
It is compared by the row of one or more positive value abscissas and with the second preset boundary threshold value;
When the row and be greater than the second preset boundary threshold value, and the row and it is corresponding positive value abscissa be described 1 or
It is multiple positive value abscissas in numerical value it is maximum positive value abscissa when, determine the row and for positive value abscissa where transverse direction
Linear position is the coboundary of the binarization of gray value image;
The binarization of gray value image is obtained according to the coboundary of the second preset coordinate spacing and the binarization of gray value image
Lower boundary.
6. bank note version recognition methods according to claim 1, which is characterized in that described according to the hand feel line administrative division map
The hand feel line area image is divided into straight line hand feel line area image and curve hand feel line by the distribution of white pixel point as in
Area image includes:
The white pixel point in the binarization of gray value image on same positive value ordinate is calculated in two-dimensional coordinate system
The sum of quantity with obtain corresponding in the binarization of gray value image one or more positive value ordinates column and;
It is compared by the column of one or more positive value ordinates and with default segmentation threshold;
When the column and it is less than the default segmentation threshold, and the column and corresponding positive value ordinate are described 1 or more
In a positive value ordinate when the smallest positive value ordinate of numerical value, determine longitudinal straight where the column and corresponding positive value ordinate
Line position is the partitioning boundary of the hand feel line area image;
The hand feel line area image is divided into straight line hand feel line area image and curve hand feel line according to the partitioning boundary
Area image.
7. bank note version recognition methods according to claim 1, which is characterized in that the basis presets linearity region parameter
Condition and pre-programmed curve region parameter condition are respectively to the straight line hand feel line area image and the curve hand feel line administrative division map
As being judged determine whether the version of the bank note is that target bank note version includes: according to judging result
It calculates the straight-line segment number of the straight line hand feel line area image and obtains starting point coordinate and the end of all straight-line segments
Point coordinate;
According to the straight-line segment number judge the straight-line segment of the straight line hand feel line area image number whether be equal to it is pre-
If straight-line segment number threshold value;
If so, judging whether the length of the straight-line segment is equal to according to the starting point coordinate of the straight-line segment and terminal point coordinate
Default straight-line segment length threshold;
If so, being judged according to the starting point coordinate of the straight-line segment and terminal point coordinate vertical between all straight-line segments
Whether spacing is equal to preset vertical spacing threshold;
If so, on the same line whether the starting point coordinate and terminal point coordinate that judge the straight-line segment;
If so, determining that the straight line hand feel line area image meets linearity region preset condition;
Judge whether the pixel matching quantity between the curve hand feel line area image and default template image is greater than default picture
Plain matching threshold;
If so, judge the terminal of the curved segments in the curve hand feel line area image abscissa and the straight line feel
Whether the abscissa of the starting point of the straight-line segment in line area image is equal;
If so, determining that the version of the bank note is target bank note version.
8. a kind of bank note version recognition device, which is characterized in that the bank note version recognition device includes:
Obtain module, the image of the predeterminated position for obtaining bank note;
Generation module, for generating binarization of gray value image corresponding with the image of the predeterminated position;
Judgment module, for counting the white pixel point quantity and black pixel point quantity of the binarization of gray value image, by institute
State white pixel point quantity and the black pixel point quantity respectively with preset white pixel threshold and default black picture element threshold value
It is compared, and judges whether the binarization of gray value image is feel line image according to comparison result;
Boundary alignment module, for according to the distribution of white pixel point in the binarization of gray value image to the binarization of gray value
Image carries out boundary alignment, to obtain hand feel line area image;
Divide module, for according to the distribution of white pixel point in the hand feel line area image by the hand feel line area image
It is divided into straight line hand feel line area image and curve hand feel line area image;
Identification module presets linearity region Parameter Conditions and pre-programmed curve region parameter condition respectively to the straight line for basis
Hand feel line area image and the curve hand feel line area image are judged, the version of the bank note is determined according to judging result
It whether is target bank note version.
9. bank note version recognition device according to claim 8, which is characterized in that the judgment module includes:
Statistic unit, for counting white pixel point quantity and black pixel point quantity in the binarization of gray value image;
Comparing unit, for by the white pixel point quantity and the black pixel point quantity respectively with preset white pixel threshold
Value and default black picture element threshold value are compared;
Judging unit, if being greater than the preset white pixel threshold or the black picture element for the white pixel point quantity
Point quantity is less than default black picture element threshold value, it is determined that the image of the predeterminated position is feel line image;If the white picture
Vegetarian refreshments quantity is not less than default black picture element threshold no more than the preset white pixel threshold or the black pixel point quantity
Value, it is determined that the image of the predeterminated position is not feel line image.
10. bank note version recognition device according to claim 8, which is characterized in that the boundary alignment module includes:
Horizontal boundary positioning unit, for according to the genesis analysis situation of white pixel point in the binarization of gray value image to institute
It states binarization of gray value image and carries out horizontal boundary positioning;
Longitudinal boundary positioning unit, for according to the cross direction profiles situation of white pixel point in the binarization of gray value image to institute
It states binarization of gray value image and carries out longitudinal boundary positioning;
Hand feel line area image acquiring unit, for being obtained according to the horizontal boundary and longitudinal boundary of the binarization of gray value image
The hand feel line area image.
11. bank note version recognition device according to claim 10, which is characterized in that the horizontal boundary positioning unit packet
It includes:
Column and computing unit, for being calculated in two-dimensional coordinate system in the binarization of gray value image in same positive value ordinate
On the sum of the quantity of white pixel point to obtain the column for corresponding to one or more positive value ordinates in the binarization of gray value image
With;
First row and comparing unit, for by it is described one or more positive value ordinates column and with the first preset boundary threshold value into
Row compares;
Right margin positioning unit, for when it is described column and be greater than the first preset boundary threshold value, and it is described column and it is corresponding just
When to be worth ordinate be the maximum positive value ordinate of numerical value in one or more described positive value ordinates, column and corresponding are determined
Longitudinal linear position where positive value ordinate is the right margin of the binarization of gray value image;
Left margin positioning unit, for obtaining institute according to the right margin of the first preset coordinate spacing and the binarization of gray value image
State the left margin of binarization of gray value image.
12. bank note version recognition device according to claim 10, which is characterized in that the longitudinal boundary positioning unit packet
It includes:
Capable and computing unit, for being calculated in two-dimensional coordinate system in the binarization of gray value image in same positive value abscissa
On the sum of the quantity of white pixel point to obtain in the binarization of gray value image for the row of one or more positive value abscissas
With;
Capable and comparing unit, for comparing by the row of one or more positive value abscissas and with the second preset boundary threshold value
Compared with;
Coboundary positioning unit, for when the row and be greater than the second preset boundary threshold value, and the row and it is corresponding just
When to be worth abscissa be the maximum positive value abscissa of numerical value in one or more described positive value abscissas, determine the row and for
Lateral linear position where positive value abscissa is the coboundary of the binarization of gray value image;
Lower boundary positioning unit, for obtaining institute according to the coboundary of the second preset coordinate spacing and the binarization of gray value image
State the lower boundary of binarization of gray value image.
13. bank note version recognition device according to claim 8, which is characterized in that the segmentation module includes:
Column and computing unit, for being calculated in two-dimensional coordinate system in the binarization of gray value image in same positive value ordinate
On the sum of the quantity of white pixel point to obtain the column for corresponding to one or more positive value ordinates in the binarization of gray value image
With;
Secondary series and comparing unit, for comparing by the column of one or more positive value ordinates and with default segmentation threshold
Compared with;
Partitioning boundary positioning unit, for when it is described column and be less than the default segmentation threshold, and it is described column and it is corresponding just
When to be worth ordinate be the smallest positive value ordinate of numerical value in one or more described positive value ordinates, column and corresponding are determined
Longitudinal linear position where positive value ordinate is the partitioning boundary of the hand feel line area image;
Cutting unit, for the hand feel line area image to be divided into straight line hand feel line area image according to the partitioning boundary
With curve hand feel line area image.
14. bank note version recognition device according to claim 8, which is characterized in that the identification module includes:
Coordinate acquiring unit, for calculating the straight-line segment number of the straight line hand feel line area image and obtaining all straight line lines
The start-stop point coordinate of section;
Number judging unit, for judging the straight-line segment of the straight line hand feel line area image according to the straight-line segment number
Number whether be equal to default straight line number threshold value;
Length determining unit is when being, then according to the straight-line segment for the judging result in the number judging unit
Starting point coordinate and terminal point coordinate judge whether the length of the straight-line segment is equal to default straight-line segment length threshold;
Vertical interval judging unit is when being, then according to the straight line line for the judging result in the length determining unit
The starting point coordinate and terminal point coordinate of section judge whether the vertical interval of all straight-line segments is equal to preset vertical spacing threshold;
Straight line judging unit is when being, then to judge the straight line line for the judging result in the vertical interval judging unit
On the same line whether the starting point coordinate and terminal point coordinate of section;If so, the straight line hand feel line area image meets straight line
Region preset condition;
Curve regions judging unit, for judging matched pixel between the curve hand feel line area image and default template image
Whether quantity is greater than presetted pixel matching threshold;
Horizontal line judging unit is when being, then to judge the curve for the judging result in the curve regions judging unit
Straight line in the abscissa of the terminal point coordinate of curved segments in hand feel line area image and the straight line hand feel line area image
Whether the abscissa of the starting point coordinate of line segment is equal;
Recognition unit is when being, it is determined that the version of the bank note is for the judging result in the horizontal line judging unit
Target bank note version.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710014330.XA CN106780967B (en) | 2017-01-09 | 2017-01-09 | A kind of bank note version recognition methods and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710014330.XA CN106780967B (en) | 2017-01-09 | 2017-01-09 | A kind of bank note version recognition methods and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106780967A CN106780967A (en) | 2017-05-31 |
CN106780967B true CN106780967B (en) | 2019-06-11 |
Family
ID=58948333
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710014330.XA Active CN106780967B (en) | 2017-01-09 | 2017-01-09 | A kind of bank note version recognition methods and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106780967B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112232331A (en) * | 2020-12-17 | 2021-01-15 | 恒银金融科技股份有限公司 | Recognition and classification method for infrared image characteristic region of paper money |
CN117115486B (en) * | 2023-10-19 | 2024-03-19 | 恒银金融科技股份有限公司 | Banknote version identification method and device and electronic equipment |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
GB2414297A (en) * | 2004-05-20 | 2005-11-23 | Enseal Systems Ltd | Assessing the quality of an image of a cheque |
CN102222384A (en) * | 2011-05-27 | 2011-10-19 | 尤新革 | Analysis method of multispectral image of note |
CN104537364A (en) * | 2015-01-29 | 2015-04-22 | 华中科技大学 | Dollar bill denomination and edition identifying method based on texture analysis |
CN105303676A (en) * | 2015-10-27 | 2016-02-03 | 深圳怡化电脑股份有限公司 | Banknote version identification method and banknote version identification system |
CN105957237A (en) * | 2016-04-22 | 2016-09-21 | 深圳怡化电脑股份有限公司 | Method and device for identifying version of banknote |
-
2017
- 2017-01-09 CN CN201710014330.XA patent/CN106780967B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
GB2414297A (en) * | 2004-05-20 | 2005-11-23 | Enseal Systems Ltd | Assessing the quality of an image of a cheque |
CN102222384A (en) * | 2011-05-27 | 2011-10-19 | 尤新革 | Analysis method of multispectral image of note |
CN104537364A (en) * | 2015-01-29 | 2015-04-22 | 华中科技大学 | Dollar bill denomination and edition identifying method based on texture analysis |
CN105303676A (en) * | 2015-10-27 | 2016-02-03 | 深圳怡化电脑股份有限公司 | Banknote version identification method and banknote version identification system |
CN105957237A (en) * | 2016-04-22 | 2016-09-21 | 深圳怡化电脑股份有限公司 | Method and device for identifying version of banknote |
Also Published As
Publication number | Publication date |
---|---|
CN106780967A (en) | 2017-05-31 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP3021259B1 (en) | Banknote recognition and classification method and system | |
CN105303676B (en) | A kind of version recognition methods of bank note and system | |
CN104851184B (en) | A kind of recognition methods of horizontally-spliced bank note and device | |
CN104867225B (en) | A kind of bank note towards recognition methods and device | |
US8385612B2 (en) | Fingerprint image background detection apparatus and detection method | |
CN106952393B (en) | Paper money identification method and device, electronic equipment and storage medium | |
JP2009519532A (en) | Bank note confirmation | |
CN107331031B (en) | A kind of recognition methods of note denomination, device, equipment and storage medium | |
CN107103683A (en) | Paper Currency Identification and device, electronic equipment and storage medium | |
CN106780967B (en) | A kind of bank note version recognition methods and device | |
CN103886309A (en) | Method for identifying dollar denominations through facial recognition | |
CN106558143B (en) | It is a kind of 100 yuans splicing bank note recognition methods and device | |
CN105788064B (en) | Bill identification method and device | |
CN107316373A (en) | A kind of Paper Currency Identification, device, terminal device and readable storage medium storing program for executing | |
CN107578526B (en) | Banknote identification method and device, terminal equipment and readable storage medium | |
CN108510639B (en) | Paper money counterfeit discriminating method and device, currency detector and storage medium | |
CN105551134B (en) | A kind of method and system of the identification of bank note fold | |
CN112307824A (en) | Method, device, system and readable medium for identifying tampering of bill number area | |
CN107134047A (en) | White watermark detection method and device | |
CN106898078B (en) | Port currency version identification method and device | |
CN105957237B (en) | A kind of version recognition methods of bank note and device | |
CN106898081B (en) | A kind of extracting method and device of dog-ear bank note | |
CN106934922A (en) | A kind of paper currency detecting method and device | |
CN106340116A (en) | Banknote identification method and device | |
CN106447908A (en) | Paper money authentic identification method and device |
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 |