WO2008002077A2 - Method for correcting note image in damaged note - Google Patents

Method for correcting note image in damaged note Download PDF

Info

Publication number
WO2008002077A2
WO2008002077A2 PCT/KR2007/003124 KR2007003124W WO2008002077A2 WO 2008002077 A2 WO2008002077 A2 WO 2008002077A2 KR 2007003124 W KR2007003124 W KR 2007003124W WO 2008002077 A2 WO2008002077 A2 WO 2008002077A2
Authority
WO
WIPO (PCT)
Prior art keywords
note image
note
scan lines
image
coordinates
Prior art date
Application number
PCT/KR2007/003124
Other languages
French (fr)
Other versions
WO2008002077A3 (en
Inventor
Hyun-Inn Kang
Han-Seop Kwak
Tae-Wan Choi
Original Assignee
Inpeg 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 Inpeg Co., Ltd. filed Critical Inpeg Co., Ltd.
Publication of WO2008002077A2 publication Critical patent/WO2008002077A2/en
Publication of WO2008002077A3 publication Critical patent/WO2008002077A3/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • 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
    • G07D7/2008—Testing patterns thereon using pre-processing, e.g. de-blurring, averaging, normalisation or rotation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T3/00—Geometric image transformations in the plane of the image
    • G06T3/60—Rotation of whole images or parts thereof
    • G06T3/604—Rotation of whole images or parts thereof using coordinate rotation digital computer [CORDIC] devices
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/97—Determining parameters from multiple pictures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G06T2207/10008—Still image; Photographic image from scanner, fax or copier
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30124—Fabrics; Textile; Paper

Definitions

  • the present invention relates to a method for correcting an image of a damaged note.
  • Image processing can be performed to determine the kind of a note and whether the note is a counterfeit. Such image processing can be performed as follows.
  • FIG 8 illustrates a database containing note images.
  • standard images and corresponding index keys are stored in the database and are used to recognize the kind of a target note.
  • an index key "krlOr” represents the backside image of a Korean ten- thousand-won note
  • an index key "jp5f represents the front side image of a Japanese five-thousand-yen note.
  • a country corresponding to a target note is first selected to limit the search range to the standard note images of the selected country, and then the kind of the target note is determined within the selected search range.
  • the kind of a note included in the query image is recognized by comparing the query image with standard note images stored in the note- image database and selecting the most similar note image to the query image.
  • FIG. 9 is an exemplary view for explaining a method of recognizing the kind of a U.S. note.
  • the image is compared with standard U.S. dollar note images stored in a note- image database for similarity calculation (9d).
  • standard U.S. dollar note images stored in a note- image database for similarity calculation (9d).
  • Extraction of an effective note image from a query image is a first and important step in processing the query image for recognizing the kind of a note included in the query image.
  • a query image of a non-damaged note can be processed.
  • an object of the present invention is to provide a method for correcting a note image of a damaged note.
  • a method for correcting a note image of a damaged note is introduced so that the kind of a target note and whether the target note is forged can be recognized even when the target note is damaged.
  • FIG. 1 is a flowchart for explaining a method of correcting a note image of a damaged note according to the present invention.
  • FIG. 2 is a view for explaining how the coordinate of a center point of a damaged note and the inclined angle of the damaged note are calculated.
  • FIGs. 3 through 6 are views of an exemplary note image for explaining operations of the method of FIG. 1.
  • FIG. 7 is a view illustrating a corrected note image.
  • FIG. 8 is a view illustrating a note-image database.
  • FIG 9 is a view illustrating procedures for recognizing the kind of a target note after limiting a search range to U.S. dollar notes.
  • a method for correcting a note image of a damaged note including: scanning the note image along horizontal scan lines at regular intervals so as to detect coordinates of intersections between the horizontal scan lines and left and right edges of the note image; calculating slopes of sections of the note image divided by the horizontal scan lines using the detected intersection coordinates; selecting horizontal scan lines related to the most frequent slope from the calculated slopes; extracting horizontal scan lines passing through non-damaged sections of the note image from the selected horizontal scan lines; calculating an x-axis coordinate of a center point of the non-damaged sections of the note image using coordinates of intersections between the selected horizontal scan lines and left and right edges of the note image; scanning the note image along vertical scan lines at regular intervals so as to detect coordinates of intersections between the vertical scan lines and upper and lower edges of the note image; calculating a y-axis coordinate of the center point using the detected intersection coordinates, and determining a coordinate of the center point using the calculated x-axis and y-axis
  • the calculating of the x-axis coordinate of the center point may include: selecting an effective scan line that is mostly close to a bitmap center of the note image from the horizontal scan lines passing through the non-damaged sections of the note image; and calculating the x-axis coordinate of the center point by using coordinates of intersections between the effective scan line and left and right edges of the non-damaged section and Equation below:
  • x r is an x-axis coordinate of the intersection between the effective scan line and the right edge of the non-damage section
  • X 1 is a x-axis coordinate of the intersection between the effective scan line and the left edge of the non-damage section.
  • ⁇ x is the interval between the vertical scan lines
  • Y pm and Y pn are y-axis coordinates of intersections at which two vertical scan lines located at both sides of a vertical scan line passing through the center point of the note image are across the upper or lower edge of the note image.
  • FIG. 1 is a flowchart for explaining a method of correcting a note image of a damaged note according to the present invention
  • FIG. 2 is a view for explaining how the coordinate of a center point of a damaged note and the inclined angle of the damaged note are calculated.
  • a damaged note image 100 is scanned along horizontal scan lines at regular intervals, and coordinates of intersections between the horizontal scan lines and left and right edges of the note image 100 are detected (Sl 1).
  • the horizontal scan lines are denoted by SX 1 , SX 2 , SX 3 , SX 4 , SX 5 , SX 6 , SX 7 , SX 8 , and SX 9 .
  • the coordinates of intersections between the horizontal scan lines and the left edge of the note image 100 are (x « i, y ⁇ , (x 1 2 , y 2 ), ..., (x i N, YN), and the coordinates of intersections between the horizontal scan lines and the right edge of the note image 100 are (x rl , y ⁇ ), (X ⁇ , y 2 ), ..., (X ⁇ N, YN)-
  • N denotes scan numbers.
  • left and right edge slopes are calculated from each section of the note image 100 divided by the horizontal scan lines.
  • the left and right edges are calculated using the detected intersection coordinates and the following equations (S 12).
  • N denotes scan numbers
  • slopes of non-damaged edges of the sections of the note image 100 are almost the same, and slopes of damaged edges of the sections of the note image 100 are different. Therefore, intersection coordinates related to the most frequent slope are selected, and horizontal scan lines SX 1 , SX 2 , SX 3 , SX 4 , SX 7 , and SX 8 including the selected intersection coordinates are extracted as indicated by green lines in FIG. 4.
  • intersection coordinates of non-damaged sections of the note image 100 are extracted (S 14).
  • the non-damaged sections are sections having no damaged edge on both sides.
  • scan lines SX 1 , SX 2 , SX 7 , and SX 8 are removed from the extracted scan lines SX 1 , SX 2 , SX 3 , SX 4 , SX 7 , and SX 8 indicated by green lines in FIG. 4.
  • the remaining scan lines SX 3 and SX 4 are scan lines passing through non-damaged sections of the note image 100 and will now be also referred to as effective scan lines.
  • Each of the effective scan lines passes through both coordinates expressed by the following equations.
  • left intersection coordinates (x ⁇ 3 , y 3 ) and (x « 4 , y 4 ), and right intersection coordinates (x r3 , y 3 ) and (x r4 , y 4 ) of non-damaged sections of the note image 100 can be extracted.
  • the scan lines SX 3 and SX 4 passing through the extracted left and right intersection coordinates are selected (S 15).
  • the scan lines SX 3 and SX 4 passing through the non-damaged sections of the note image are selected in Step S 15.
  • the coordinate of a center point of the note image 100 is calculated using the intersecion coordinates between the selected horizontal scan lines SX 3 and SX 4 and left and right edges of the note image;(S16).
  • the scan line SX 4 is selected from the scan lines SX 3 and SX 4 as a scan line that is mostly close to a bitmap center of the note image 100, and an x-axis coordinate of the center point of the note image 100 is calculated using the left and right intersection coordinates (x 1 4 , y 4 ) and (x r4 , y 4 ) of the scan line SX 4 and the following equation.
  • X-axis coordinate of center point :
  • a coordinate (PC X(4) , Y 4 ) is calculated (indicated by a blue cross in FIG. 6)
  • the note image 100 is scanned along vertical scan lines at regular intervals to detect coordinates of intersections between the vertical scan lines and the upper and lower edges of the note image 100 (S 17).
  • the note image 100 is scanned along a vertical scan line SYc passing through the coordinate (PC X(4) , Y 4 ) and scan lines SYl and SYr spaced apart from the scan line SYc by ⁇ x.
  • the scan lines SYc, SYl, and SYr meet with upper and lower edges of the note image 100 at coordinates (x pl , y pl ), (x p2 , y p2 ), (x p3 , y p3 ), (Xp4, y P 4), (x P 5, y P s), and
  • an y-axis coordinate of the center point of the note image 100 is calculated using the following equation (S 18) .
  • the coordinate of the center point of the note image 100 is (PC X(4) , PC y ) as indicated by a red cross in FIG. 6.
  • the inclined angle of the note image 100 is calculated (S 19).
  • the inclined angle of the note image 100 can be calculated from a region located between the vertical scan lines SYl and SYr by using the following equation.
  • the note image 100 is rotated using the center-point coordinate and the inclined angle for correcting the note image 100 (S20).
  • the note image 100 is rotated by - ⁇ (counterclockwise) on the center point (PC X(4) , PC y ). Therefore, a rotation-corrected image can be obtained as shown in FIG. 7.
  • the processing load can be reduced by rotation- correcting the note image 100 as follows: the note image 100 is first moved such that an imaginary left lower coner of the note image 100 indicated by a dashed line is moved to the center point (PC x(4) , PC y ); the note image 100 is rotated about the center point (PC X(4) , PC y ); and the note image 100 is moved back.
  • the note image 100 when the note image 100 has a MxN size, the note image 100 is moved to the first quadrant of a coordinate system having the center point (PC X ( 4) , PC y )as an origin. Then, the coordinate (x m , y n ) of each pixel of the note image 100 is converted to (x' m , y' n ) by rotation correction using the following equation.
  • the rotation-corrected note image 100 is shown in FIG 7.
  • a corrected image can be provided for a damaged note so that the kind of the damaged note and whether the damaged note is a counterfeit can be recognized.
  • Each operation of the method of the present invention can be programmed for performing the operation using a computer.
  • the method of the present invention can be realized in the form of an integrated circuit partially or entirely including a functional module executing each operation of the method.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Image Analysis (AREA)
  • Editing Of Facsimile Originals (AREA)
  • Image Processing (AREA)
  • Inspection Of Paper Currency And Valuable Securities (AREA)

Abstract

Provided is a method for correcting a note image of a damaged note. In the method, the note image is scanned along horizontal scan lines at regular intervals to detect coordinates of intersections between the horizontal scan lines and left and right edges of the note image, and slopes of sections of the note image divided by the horizontal scan lines are calculated using the intersection coordinates. Horizontal scan lines related to the most frequent slope is selected from the calculated slopes. Horizontal scan lines passing through non-damaged sections of the note image is extracted from the selected horizontal scan lines. An x-axis coordinate of a center point of the non-damaged sections is calculated using coordinates of intersections between the selected horizontal scan lines and left and right edges of the note image. The note image is scanned along vertical scan lines at regular intervals to detect coordinates of intersections between the vertical scan lines and upper and lower edges of the note image. A y-axis coordinate of the center point is calculated using the detected intersection coordinates to determine a coordinate of the center point. An inclined angle of the note image is calculated. The note image is rotated on the center point by the calculated inclined angle in a direction opposite to the inclined direction of the note image to obtain a rotation-corrected note image.

Description

METHOD FORCORRECTINGNOTE IMAGE INDAMAGEDNOTE
Technical Field
The present invention relates to a method for correcting an image of a damaged note.
Background Art
Image processing can be performed to determine the kind of a note and whether the note is a counterfeit. Such image processing can be performed as follows.
FIG 8 illustrates a database containing note images. Referring to FIG. 8, standard images and corresponding index keys are stored in the database and are used to recognize the kind of a target note. For example, an index key "If represents the front side image of a one-dollar note, an index key "krlOr" represents the backside image of a Korean ten- thousand-won note, and an index key "jp5f represents the front side image of a Japanese five-thousand-yen note. In the case of using the note-image database, it may take too much time to perform similarity comparison and it may be difficult to precisely recognize a target note since a number of standard note images are registered in the note-image database for each country. To address these problems, a country corresponding to a target note is first selected to limit the search range to the standard note images of the selected country, and then the kind of the target note is determined within the selected search range.
When a query image is input, the kind of a note included in the query image is recognized by comparing the query image with standard note images stored in the note- image database and selecting the most similar note image to the query image.
FIG. 9 is an exemplary view for explaining a method of recognizing the kind of a U.S. note.
In detail, when a query image of an inclined ten-dollar note is input (a), the inclined angle and center coordinate of the ten-dollar note are calculated, and the effective region of the ten-dollar note is extracted from the query image (9b). Next, the extracted image is rotated and newly positioned using the calculated inclination angle and center coordinate (9c).
Then, the image is compared with standard U.S. dollar note images stored in a note- image database for similarity calculation (9d). In this way, the kind of a target note included in the query image can be recognized by determining a standard U.S. note image having the most similarity to the target note.
Disclosure
Technical Problem
Extraction of an effective note image from a query image is a first and important step in processing the query image for recognizing the kind of a note included in the query image. In the related art, a query image of a non-damaged note can be processed.
However, a method of extracting an effective image from a query image of a damaged note is not introduced.
Technical Solution Accordingly, an object of the present invention is to provide a method for correcting a note image of a damaged note.
Additional advantages, objects, and features of the invention will become apparent to those having ordinary skill in the related art upon examination of the following embodiments of the invention.
Advantageous Effects
According to the present invention, a method for correcting a note image of a damaged note is introduced so that the kind of a target note and whether the target note is forged can be recognized even when the target note is damaged.
Description of Drawings
FIG. 1 is a flowchart for explaining a method of correcting a note image of a damaged note according to the present invention.
FIG. 2 is a view for explaining how the coordinate of a center point of a damaged note and the inclined angle of the damaged note are calculated.
FIGs. 3 through 6 are views of an exemplary note image for explaining operations of the method of FIG. 1.
FIG. 7 is a view illustrating a corrected note image. FIG. 8 is a view illustrating a note-image database.
FIG 9 is a view illustrating procedures for recognizing the kind of a target note after limiting a search range to U.S. dollar notes.
Best Mode
According to the present invention, there is provided a method for correcting a note image of a damaged note, the method including: scanning the note image along horizontal scan lines at regular intervals so as to detect coordinates of intersections between the horizontal scan lines and left and right edges of the note image; calculating slopes of sections of the note image divided by the horizontal scan lines using the detected intersection coordinates; selecting horizontal scan lines related to the most frequent slope from the calculated slopes; extracting horizontal scan lines passing through non-damaged sections of the note image from the selected horizontal scan lines; calculating an x-axis coordinate of a center point of the non-damaged sections of the note image using coordinates of intersections between the selected horizontal scan lines and left and right edges of the note image; scanning the note image along vertical scan lines at regular intervals so as to detect coordinates of intersections between the vertical scan lines and upper and lower edges of the note image; calculating a y-axis coordinate of the center point using the detected intersection coordinates, and determining a coordinate of the center point using the calculated x-axis and y-axis coordinates; calculating an inclined angle of the note image; and rotating the note image on the center point by the calculated inclined angle in a direction opposite to the inclined direction of the note image so as to obtain a rotation-corrected note image. Coordinates of the horizontal scan lines passing through the non-damaged sections of the note image satisfy both of Equations below: <Left intersection coordinate>
<Right intersection coordinate> (*m,>O= (*ri+te -(«- l)>;yri+AK«- l)) where Δx = xn - Xn-1 is an x-axis displacement, xn is an x-axis coordinate of a point which a horizontal scan line related to the most frequent slope passes through, xn-i is an x- axis coordinate of a neighboring point, Δy = yn - yn-1 = yn-1 - yn-2 .... is a y-axis displacement between the two neighboring points, and n denotes the number of scanned lines. The calculating of the x-axis coordinate of the center point may include: selecting an effective scan line that is mostly close to a bitmap center of the note image from the horizontal scan lines passing through the non-damaged sections of the note image; and calculating the x-axis coordinate of the center point by using coordinates of intersections between the effective scan line and left and right edges of the non-damaged section and Equation below:
Pf* = ■*- -f-
^ ^- x x I^ 2
where xr is an x-axis coordinate of the intersection between the effective scan line and the right edge of the non-damage section, and X1 is a x-axis coordinate of the intersection between the effective scan line and the left edge of the non-damage section. The inclined angle of the note image may be calculated using Equation below:
'X where Δx is the interval between the vertical scan lines, and Ypm and Ypn are y-axis coordinates of intersections at which two vertical scan lines located at both sides of a vertical scan line passing through the center point of the note image are across the upper or lower edge of the note image.
Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings.
FIG. 1 is a flowchart for explaining a method of correcting a note image of a damaged note according to the present invention, and FIG. 2 is a view for explaining how the coordinate of a center point of a damaged note and the inclined angle of the damaged note are calculated.
As shown in FIG. 3 by blue lines, a damaged note image 100 is scanned along horizontal scan lines at regular intervals, and coordinates of intersections between the horizontal scan lines and left and right edges of the note image 100 are detected (Sl 1). The horizontal scan lines are denoted by SX1, SX2, SX3, SX4, SX5, SX6, SX7, SX8, and SX9.
In detail, referring to FIG. 2, the coordinates of intersections between the horizontal scan lines and the left edge of the note image 100 are (x « i, yθ, (x 12, y2), ..., (x i N, YN), and the coordinates of intersections between the horizontal scan lines and the right edge of the note image 100 are (xrl, yϊ), (Xβ, y2), ..., (XΓN, YN)- Here, N denotes scan numbers.
Next, left and right edge slopes are calculated from each section of the note image 100 divided by the horizontal scan lines. The left and right edges are calculated using the detected intersection coordinates and the following equations (S 12).
<Slopes of left edges of sections>
Figure imgf000007_0001
<Slopes of right edges of sections>
T1= r2= rn=-
Xr2'Xrl Xr3~Xr2 X m+l"X rn
where N denotes scan numbers, and n denotes the number of horizontal scan lines (n=l, 2, ..., N-l). Next, horizontal scan lines related to the most frequent slope is selected from the calculated slopes (S 13).
In detail, slopes of non-damaged edges of the sections of the note image 100 are almost the same, and slopes of damaged edges of the sections of the note image 100 are different. Therefore, intersection coordinates related to the most frequent slope are selected, and horizontal scan lines SX1, SX2, SX3, SX4, SX7, and SX8 including the selected intersection coordinates are extracted as indicated by green lines in FIG. 4.
Thereafter, intersection coordinates of non-damaged sections of the note image 100 are extracted (S 14). Here, the non-damaged sections are sections having no damaged edge on both sides.
Although both left and right edges of a section of the note image 100 have the most frequent slope, the section can be a damaged section. Therefore, it is necessary to select non-damaged sections from the sections of which both left and right edges have the most frequent slope. For this, scan lines SX1, SX2, SX7, and SX8 are removed from the extracted scan lines SX1, SX2, SX3, SX4, SX7, and SX8 indicated by green lines in FIG. 4.
The remaining scan lines SX3 and SX4 are scan lines passing through non-damaged sections of the note image 100 and will now be also referred to as effective scan lines. Each of the effective scan lines passes through both coordinates expressed by the following equations.
<Left intersection coordinate>
<Right intersection coordinate>
Figure imgf000008_0001
where Δx = xn - Xn-1 is an x-axis displacement (xn is an x-axis coordinate of a point which a horizontal scan line related to the most frequent slope passes through, and Xn-1 is an neighboring x-axis coordinate of Xn), and Δy = yn - yn-1 = yn-1 - yn-2 .... is a y-axis displacement between neighboring two y-coordinates (Δy is equal to the interval between the horizontal scan lines).
Left and right intersection coordinates of the scan lines SX1, SX2, SX3, SX4, SX7, and SX8 are input to the equations, and coordinates satisfying the equations are extracted.
In this way, left intersection coordinates (x ^ 3, y3) and (x « 4, y4), and right intersection coordinates (xr3, y3) and (xr4, y4) of non-damaged sections of the note image 100 can be extracted.
Therefore, as shown by red lines in FIG. 5, the scan lines SX3 and SX4 passing through the extracted left and right intersection coordinates are selected (S 15). In other words, the scan lines SX3 and SX4 passing through the non-damaged sections of the note image are selected in Step S 15. Next, the coordinate of a center point of the note image 100 is calculated using the intersecion coordinates between the selected horizontal scan lines SX3 and SX4 and left and right edges of the note image;(S16).
The scan line SX4 is selected from the scan lines SX3 and SX4 as a scan line that is mostly close to a bitmap center of the note image 100, and an x-axis coordinate of the center point of the note image 100 is calculated using the left and right intersection coordinates (x 1 4, y4) and (xr4, y4) of the scan line SX4 and the following equation. X-axis coordinate of center point:
PC = JC , <**-**)
Therefore, as shown in FIG. 2, a coordinate (PCX(4), Y4) is calculated (indicated by a blue cross in FIG. 6)
Next, referring to FIGs. 2 and 6, the note image 100 is scanned along vertical scan lines at regular intervals to detect coordinates of intersections between the vertical scan lines and the upper and lower edges of the note image 100 (S 17).
In detail, the note image 100 is scanned along a vertical scan line SYc passing through the coordinate (PCX(4), Y4) and scan lines SYl and SYr spaced apart from the scan line SYc by Δx. Here, the scan lines SYc, SYl, and SYr meet with upper and lower edges of the note image 100 at coordinates (xpl, ypl), (xp2, yp2), (xp3, yp3), (Xp4, yP4), (xP5, yPs), and
(xP6, yPό).
Next, an y-axis coordinate of the center point of the note image 100 is calculated using the following equation (S 18) .
Figure imgf000009_0001
Therefore, as shown in FIG 2, the coordinate of the center point of the note image 100 is (PCX(4), PCy) as indicated by a red cross in FIG. 6. Next, the inclined angle of the note image 100 is calculated (S 19). The inclined angle of the note image 100 can be calculated from a region located between the vertical scan lines SYl and SYr by using the following equation.
Θ = tan ( p p )
Next, the note image 100 is rotated using the center-point coordinate and the inclined angle for correcting the note image 100 (S20). Referring to FIG. 2, the note image 100 is rotated by -Θ (counterclockwise) on the center point (PCX(4), PCy). Therefore, a rotation-corrected image can be obtained as shown in FIG. 7.
Here, when the note image 100 is rotation-corrected using a program, it is interpreted that the note image 100 is centered on the center point (PCX(4), PCy) across the first to fourth quadrants. Therefore, the processing load can be reduced by rotation- correcting the note image 100 as follows: the note image 100 is first moved such that an imaginary left lower coner of the note image 100 indicated by a dashed line is moved to the center point (PCx(4), PCy); the note image 100 is rotated about the center point (PCX(4), PCy); and the note image 100 is moved back.
In detail, when the note image 100 has a MxN size, the note image 100 is moved to the first quadrant of a coordinate system having the center point (PCX(4), PCy)as an origin. Then, the coordinate (xm, yn) of each pixel of the note image 100 is converted to (x'm, y'n) by rotation correction using the following equation.
M N M N x'm =(χ m-γ )cosθ+Ov— >™-*>y',T - O,*-^- )sinθ+(^M-— )cosθ ,
Where β>0 and0<xm<M, 0<yιι<N
x'm =(χ m — )cosθ- (y_-— )sinθ, >>'„ = (>,,,-— )sinθ+O_— )<κ»β ,
Where θ<θ andθ<Λ:w<A4",O<>>Λ<.Λ/
Then, the note image 100 is moved back using the following equation. x"m=x'm+rCx,y"n=y>n+PCy Where 0<*' m<M,0<y' n<N,0<x" m<M,0<y" »<N,
In this way, a rotation-corrected coordinate (x"m, y"n) is calculated for each pixel of the note image 100.
The rotation-corrected note image 100 is shown in FIG 7.
As described above, a corrected image can be provided for a damaged note so that the kind of the damaged note and whether the damaged note is a counterfeit can be recognized. Each operation of the method of the present invention can be programmed for performing the operation using a computer. In addition, the method of the present invention can be realized in the form of an integrated circuit partially or entirely including a functional module executing each operation of the method.
The method of the present invention can be embodied in various forms and manners without departing from the spirit and scope of the invention as defined by the appended claims.

Claims

Claims:
1. A method for correcting a note image of a damaged note, the method comprising: scanning the note image along horizontal scan lines at regular intervals so as to detect coordinates of intersections between the horizontal scan lines and left and right edges of the note image; calculating slopes of sections of the note image divided by the horizontal scan lines using the detected intersection coordinates; selecting horizontal scan lines related to the most frequent slope from the calculated slopes; extracting horizontal scan lines passing through non-damaged sections of the note image from the selected horizontal scan lines; calculating an x-axis coordinate of a center point of the non-damaged sections of the note image using coordinates of intersections between the selected horizontal scan lines and left and right edges of the note image; scanning the note image along vertical scan lines at regular intervals so as to detect coordinates of intersections between the vertical scan lines and upper and lower edges of the note image; calculating a y-axis coordinate of the center point using the detected intersection coordinates, and determining a coordinate of the center point using the calculated x-axis and y-axis coordinates; calculating an inclined angle of the note image; and rotating the note image on the center point by the calculated inclined angle in a direction opposite to the inclined direction of the note image so as to obtain a rotation- corrected note image.
2. The method of claim 1, wherein coordinates of the horizontal scan lines passing through the non-damaged sections of the note image satisfy both of Equations below:
<Left intersection coordinate>
Figure imgf000011_0001
-(«- l)) <Right intersection coordinate>
where Δx = xn - Xn-1 is an x-axis displacement, xn is an x-axis coordinate of a point which a horizontal scan line related to the most frequent slope passes through, Xn-1 is an x- axis coordinate of a neighboring point, Δy = yn - yn-1 = yn-1 - yn-2 .... is a y-axis displacement between the two neighboring points, and n denotes the number of scan lines.
3. The method of claim 1, wherein the calculating of the x-axis coordinate of the center point comprises: selecting an effective scan line that is mostly close to a bitmap center of the note image from the horizontal scan lines passing through the non-damaged sections of the note image; and calculating the x-axis coordinate of the center point by using coordinates of intersections between the effective scan line and left and right edges of the non-damaged section and Equation below:
PC x,Xι l +^p 2 L
where xr is an x-axis coordinate of the intersection between the effective scan line and the right edge of the non-damage section, and X1 is a x-axis coordinate of the intersection between the effective scan line and the left edge of the non-damage section.
4. The method of claim 1, wherein the inclined angle of the note image is calculated using Equation below:
Figure imgf000012_0001
where Δx is the interval between the vertical scan lines, and Ypm and Ypn are y-axis coordinates of intersections at which two vertical scan lines located at both sides of a vertical scan line passing through the center point of the note image are across the upper or lower edge of the note image.
PCT/KR2007/003124 2006-06-30 2007-06-27 Method for correcting note image in damaged note WO2008002077A2 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
KR10-2006-0061131 2006-06-30
KR1020060061131A KR100768134B1 (en) 2006-06-30 2006-06-30 Method for correcting note image in damaged note

Publications (2)

Publication Number Publication Date
WO2008002077A2 true WO2008002077A2 (en) 2008-01-03
WO2008002077A3 WO2008002077A3 (en) 2008-02-21

Family

ID=38815121

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2007/003124 WO2008002077A2 (en) 2006-06-30 2007-06-27 Method for correcting note image in damaged note

Country Status (2)

Country Link
KR (1) KR100768134B1 (en)
WO (1) WO2008002077A2 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103606221A (en) * 2013-12-04 2014-02-26 广州广电运通金融电子股份有限公司 Fault automatic diagnostic method of counter and device
CN107481393A (en) * 2017-06-26 2017-12-15 广州广电运通金融电子股份有限公司 A kind of detection method for splicing bill and the detection means for splicing bill

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101276063B1 (en) * 2008-09-09 2013-06-14 삼성전자주식회사 Image forming apparatus, image forming system and control method thereof

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5751841A (en) * 1995-09-29 1998-05-12 Ncr Corporation Method and apparatus for scanning bank notes
EP1093084A1 (en) * 1998-06-30 2001-04-18 Sharp Kabushiki Kaisha Image correction device
JP2001331839A (en) * 2000-05-22 2001-11-30 Glory Ltd Method and device for discriminating paper money
US20050133339A1 (en) * 2002-08-30 2005-06-23 Fujitsu Limited Device, method and program for paper discrimination

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5751841A (en) * 1995-09-29 1998-05-12 Ncr Corporation Method and apparatus for scanning bank notes
EP1093084A1 (en) * 1998-06-30 2001-04-18 Sharp Kabushiki Kaisha Image correction device
JP2001331839A (en) * 2000-05-22 2001-11-30 Glory Ltd Method and device for discriminating paper money
US20050133339A1 (en) * 2002-08-30 2005-06-23 Fujitsu Limited Device, method and program for paper discrimination

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103606221A (en) * 2013-12-04 2014-02-26 广州广电运通金融电子股份有限公司 Fault automatic diagnostic method of counter and device
CN103606221B (en) * 2013-12-04 2016-01-20 广州广电运通金融电子股份有限公司 Fault automatic diagnostic method of counter and device
US9947164B2 (en) 2013-12-04 2018-04-17 Grg Banking Equipment Co., Ltd. Automatic fault diagnosis method and device for sorting machine
CN107481393A (en) * 2017-06-26 2017-12-15 广州广电运通金融电子股份有限公司 A kind of detection method for splicing bill and the detection means for splicing bill
CN107481393B (en) * 2017-06-26 2019-10-11 广州广电运通金融电子股份有限公司 A kind of detection method for splicing bill and the detection device for splicing bill

Also Published As

Publication number Publication date
WO2008002077A3 (en) 2008-02-21
KR100768134B1 (en) 2007-10-17

Similar Documents

Publication Publication Date Title
US8170368B2 (en) Correcting device and method for perspective transformed document images
US6754369B1 (en) License plate reading apparatus and method
CN106934795B (en) A kind of automatic testing method and prediction technique of glue into concrete beam cracks
US8401333B2 (en) Image processing method and apparatus for multi-resolution feature based image registration
US8483489B2 (en) Edge based template matching
US8144989B2 (en) Methods and systems for identifying text orientation in a digital image
US7613355B2 (en) Image processing device and registration data generation method in image processing
US20120294528A1 (en) Method of Detecting and Correcting Digital Images of Books in the Book Spine Area
JPH11219407A (en) Document image recognizing device and storage medium for document image recognizing program
US20120093434A1 (en) Edge detection
CN111340701A (en) Circuit board image splicing method for screening matching points based on clustering method
TWI522934B (en) Gyro sensor license plate recognition system for smart phone and method thereof
US20080317343A1 (en) Methods and Systems for Identifying Text Orientation in a Digital Image
CN107423735B (en) License plate positioning method utilizing horizontal gradient and saturation
KR20100112155A (en) Feature-based signatures for image identification
EP2579129A2 (en) Apparatus and method of detecting an input position with display pattern recognition
CN110533036B (en) Rapid inclination correction method and system for bill scanned image
JP2011043969A (en) Method for extracting image feature point
JP2008217833A (en) Document image recognition system and storage medium for document image recognition program
JP2002228423A (en) Tire detecting method and device
WO2008002077A2 (en) Method for correcting note image in damaged note
CN114140794A (en) General image correction method based on OCR
CN112800824A (en) Processing method, device and equipment for scanning file and storage medium
JP5010627B2 (en) Character recognition device and character recognition method
JP3303246B2 (en) Image processing device

Legal Events

Date Code Title Description
DPE2 Request for preliminary examination filed before expiration of 19th month from priority date (pct application filed from 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 07747146

Country of ref document: EP

Kind code of ref document: A2

NENP Non-entry into the national phase

Ref country code: DE

NENP Non-entry into the national phase

Ref country code: RU

122 Ep: pct application non-entry in european phase

Ref document number: 07747146

Country of ref document: EP

Kind code of ref document: A2