WO2018082540A1 - 一种票据一维信号的检测方法及装置 - Google Patents

一种票据一维信号的检测方法及装置 Download PDF

Info

Publication number
WO2018082540A1
WO2018082540A1 PCT/CN2017/108559 CN2017108559W WO2018082540A1 WO 2018082540 A1 WO2018082540 A1 WO 2018082540A1 CN 2017108559 W CN2017108559 W CN 2017108559W WO 2018082540 A1 WO2018082540 A1 WO 2018082540A1
Authority
WO
WIPO (PCT)
Prior art keywords
dimensional
abnormal point
abnormal
point
ticket
Prior art date
Application number
PCT/CN2017/108559
Other languages
English (en)
French (fr)
Inventor
龚岩
王荣秋
王佳
孙燕
Original Assignee
广州广电运通金融电子股份有限公司
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 广州广电运通金融电子股份有限公司 filed Critical 广州广电运通金融电子股份有限公司
Publication of WO2018082540A1 publication Critical patent/WO2018082540A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G07CHECKING-DEVICES
    • G07DHANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
    • G07D7/00Testing 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
    • G07D7/20Testing patterns thereon
    • G07D7/202Testing patterns thereon using pattern matching
    • G07D7/2041Matching statistical distributions, e.g. of particle sizes orientations

Definitions

  • the invention relates to the field of banknote identification, and in particular to a method and a device for detecting a one-dimensional signal of a bill.
  • the ticket identification technology is changing with each passing day.
  • the existing commonly used mainstream ticket identification technology is one-dimensional signal detection.
  • One-dimensional signal detection includes magnetic signal recognition and thickness signal recognition.
  • the traditional one-dimensional signal detection method of bills has certain limitations.
  • 1 is not sensitive to the position information determination of the bill feature points, especially the positional relationship on the vertical component of the bill running direction is prone to errors;
  • Dimensional signal processing is susceptible to interference and instability, and the positive and negative samples are less distinguishable;
  • 3 one-dimensional signal processing has fewer algorithms and has certain limitations. Therefore, the existing one-dimensional signal detection method of the bill easily causes the abnormal sample to be difficult to be distinguished, and the normal sample is rejected, and the like, especially for the banknote with the transparent tape attached, the bank needs to recycle the banknote. It is often difficult to accurately identify by the prior art, for example, the conventional thickness detection cannot judge both the banknote adhesive tape and the banknote wrinkle.
  • the embodiment of the invention provides a method and a device for detecting a one-dimensional signal of a bill, which converts the one-dimensional signal data of the bill into a two-dimensional gray value, and then preprocesses the two-dimensional gray value, and finally passes the two-dimensional data processing.
  • the method detects the two-dimensional gray value and detects whether the note has adhesive tape, authenticity and face value, and solves the problem that the abnormal sample caused by the existing one-dimensional signal detection method is difficult to be distinguished and the normal sample is rejected. Especially for technical problems in which it is difficult to accurately identify banknotes with transparent tape.
  • the two-dimensional gray value I(x, n) is detected by a two-dimensional data processing method and it is determined whether the ticket is abnormal.
  • the two-dimensional data processing method After the one-dimensional signal data of the at least one ticket is converted into a two-dimensional gray value I(x, n) by the first formula and sequentially mapped to the grayscale image according to the number of signal channels, the two-dimensional data processing method is adopted. Before detecting the two-dimensional gray value and determining whether the ticket is abnormal, it also includes:
  • the two-dimensional gray value I(x, n) is subjected to tilt correction and/or interpolation processing.
  • Detecting the two-dimensional gray value I(x, n) by the two-dimensional data processing method and determining whether the ticket is abnormal includes:
  • the information of the abnormal point is compared with the information of the normal ticket to determine whether the ticket is stuck with tape or wrinkles.
  • the gray scale g (x, n) Comparing the gray level g (x, n) of each point in the gray scale map mapped by the two-dimensional gray value I(x, n) with the magnitude of the threshold value t, the gray scale g (x, n) The information that is greater than the threshold t is marked as an abnormal point and then the information of the abnormal point is calculated to include:
  • the abnormal point is marked with a number
  • the abnormal point is marked as the number of the left abnormal point
  • the abnormal point is marked as the number of the abnormal point above;
  • the abnormal point is marked as the left abnormal point number and the smaller number in the upper abnormal point number, and the larger number is modified to be smaller. Numbering;
  • Detecting the two-dimensional gray value I(x, n) by the two-dimensional data processing method and determining whether the ticket is abnormal includes:
  • the two-dimensional gray value of the normal ticket is respectively established by taking at least one normal ticket as a sample, and the normal reference two-dimensional gray value I n is obtained by averaging the two-dimensional gray values of the same position of at least one normal ticket. , y);
  • the normal reference two-dimensional gray value I n (x, y) is matched with the two-dimensional gray value I(x, n) and the ticket authenticity and the bill face value are determined by the matching degree.
  • the embodiment of the invention provides a device for detecting a one-dimensional signal of a bill, comprising:
  • the acquiring module is configured to acquire one-dimensional signal data of the ticket collected by at least one sensor in real time;
  • a two-dimensional data conversion module configured to convert at least one of the one-dimensional signal data of the ticket into a two-dimensional gray value I(x, n) by a first formula and sequentially map to a grayscale image according to the number of signal channels,
  • the first formula is Where g (x, n) is one of the gray levels of the two-bit gray value, M(x, n) is the one-dimensional signal data, and M min is the minimum value of the one-dimensional signal data, M max For the maximum value of the one-dimensional signal data, the range of x is the total length of the one-dimensional signal data, n represents the first channel currently located in the one-dimensional signal, and Chanel max is the number of the signal channels;
  • the two-dimensional data processing module is configured to detect the two-dimensional gray value I(x, n) by a two-dimensional data processing method and determine whether the ticket is abnormal.
  • the apparatus for detecting a one-dimensional signal of the bill further includes:
  • a tilt correction and/or interpolation processing module for performing tilt correction and/or interpolation processing on the two-dimensional gray value I(x, n).
  • the two-dimensional data processing module specifically includes:
  • a threshold calculation unit configured to calculate a threshold t between the foreground and the background of the image
  • the abnormal point information calculation unit compares the gradation g (x, n) of each point in the gradation map mapped by the two-dimensional gradation value I (x, n) with the magnitude of the threshold t, and the ash A point where the degree g (x, n) is greater than the threshold t is marked as an abnormal point and then the information of the abnormal point is calculated;
  • the comparison judging unit compares the information of the abnormal point with the information of the normal ticket to determine whether the ticket is stuck with tape.
  • the threshold calculation unit is specifically configured to:
  • the abnormal point information calculation unit specifically includes:
  • An abnormal point is marked from the unit for scanning the grayscale map mapped by the two-dimensional gray value I(x, n) from left to right and from top to bottom and comparing the mapped grayscale maps
  • the gradation g (x, n) of each point and the magnitude of the threshold t if the gradation g (x, n) is greater than the threshold t , mark the point corresponding to the gradation g (x, n) as An abnormal point, and then determining whether there is an abnormal point on the left side and the upper side in the four neighborhoods of the abnormal point;
  • the abnormal point is marked with a number
  • the abnormal point is marked as the number of the left abnormal point
  • the abnormal point is marked as the number of the abnormal point above;
  • the abnormal point is marked as the left abnormal point number and the smaller number in the upper abnormal point number, and the larger number is modified to be smaller. Numbering;
  • a connected domain information calculation subunit for scanning the grayscale map mapped by the two-dimensional grayscale value I(x, n) of the abnormal point mark and combining the different numbers of connected numbers by using a summation algorithm The abnormal point area, and then calculate the connected domain information of the abnormal point.
  • the two-dimensional data processing module specifically includes:
  • the normal reference two-dimensional data establishing module is configured to respectively establish a two-dimensional gray value of the normal ticket by using at least one normal ticket as a sample, and obtain a normal value by averaging the two-dimensional gray values of the same position of at least one normal ticket.
  • the matching judging module is configured to match the normal reference two-dimensional gray value I n (x, y) with the two-dimensional gray value I(x, n) and determine the bill authenticity and the bill face value by the matching degree.
  • the embodiment of the present invention is completely based on two-dimensional signals, so that the positional relationship of the bill features can be better reflected, and the processing means that can be used is more abundant.
  • the one-dimensional signal conversion is mapped to the two-dimensional signal, and the data is converted and mapped based on the data of a single ticket, which better retains the feature information of the ticket, and eliminates the interference of noise.
  • the dimension-increasing processing can make the detection more intuitive, retaining the characteristics of the one-dimensional signal and reducing the noise carried by the one-dimensional signal.
  • the distinction between positive and negative samples of the bill The degree is also more obvious, and the detection effect is better.
  • FIG. 1 is a schematic flowchart of a first embodiment of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention
  • FIG. 2 is a schematic flowchart of a second embodiment of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention
  • FIG. 3 is a schematic structural diagram of a first embodiment of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present disclosure
  • FIG. 4 is a schematic structural diagram of a second embodiment of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention
  • FIG. 5 is a diagram of a test banknote with a vertical tape attached to a first application example of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention
  • FIG. 6 is a schematic diagram of a thickness signal of a test banknote to which a vertical adhesive tape is pasted in a first application example of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention
  • FIG. 7 is an image of a first application example of a method for detecting a one-dimensional signal of a bill according to an embodiment of the present invention, in which a one-dimensional signal of a test bank is upgraded into a two-dimensional gray image;
  • FIG. 8 is an image of a first application example of a method for detecting a one-dimensional signal of a ticket after performing multi-interpolation on two-dimensional gray data of a two-dimensional gray map according to an embodiment of the present disclosure
  • FIG. 9 is a schematic diagram of an image after a test banknote is detected in a first application example of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention.
  • FIG. 10 and FIG. 11 are respectively a sample view of a folded banknote and a corner portion of a banknote in a second application example of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention
  • FIG. 12 and FIG. 13 are schematic diagrams showing an original one-dimensional signal of a folded banknote and a corner portion of a banknote in a second application example of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention
  • FIG. 14 and FIG. 15 are respectively a two-dimensional grayscale diagram of a folded banknote and a corner portion of a banknote in a second application example of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention
  • FIG. 16 and FIG. 17 are respectively a final detection image of a folded banknote and a corner portion of a banknote in a second application example of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention
  • FIG. 19 and FIG. 20 are respectively a sample view of a pleated banknote, a first irregular tape, and a second irregular tape in a third application example of a method for detecting a one-dimensional signal of a bill according to an embodiment of the present invention
  • FIG. 22 and FIG. 23 are respectively a first example of a method for detecting a one-dimensional signal of a bill according to an embodiment of the present invention, in which a pleated banknote, a first irregular tape, and a second irregular tape are sampled.
  • Dimensional thickness signal diagram ;
  • FIG. 25 and FIG. 26 are respectively a second application example of a method for detecting a one-dimensional signal of a bill according to an embodiment of the present invention, in which two types of pleated banknotes, first irregular tapes, and second irregular tapes are used.
  • Dimensional grayscale image
  • FIG. 28 and FIG. 29 respectively illustrate the detection of the pleated banknote, the first irregular tape, and the second irregular tape in the third application example of the method for detecting the one-dimensional signal of the ticket according to an embodiment of the present invention.
  • FIG. 30 and FIG. 31 are sample diagrams of US$2 and US$5 in a fourth application example of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention
  • FIG. 32 and FIG. 33 are schematic diagrams showing a one-dimensional magnetic signal of US$2 and US$5 in a fourth application example of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention
  • FIG. 34 and FIG. 35 are two-dimensional magnetic signal images of US$2 and US$5, respectively, in a fourth application example of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention.
  • the embodiment of the invention provides a method and a device for detecting a one-dimensional signal of a bill, which converts the one-dimensional signal data of the bill into a two-dimensional gray value, and then preprocesses the two-dimensional gray value, and finally passes the two-dimensional data processing.
  • the method detects the two-dimensional gray value and detects whether the note has adhesive tape, authenticity and face value, and solves the problem that the abnormal sample caused by the existing one-dimensional signal detection method is difficult to be distinguished and the normal sample is rejected, especially for the paste. Tape banknotes are difficult to accurately identify and other technical issues
  • a first embodiment of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention includes:
  • one-dimensional signal data of the ticket collected by at least one sensor is acquired in real time.
  • 102. Convert one-way signal data of at least one ticket into a two-dimensional gray value I(x, n) by using a first formula, and sequentially map to a grayscale image according to the number of signal channels, and the first formula is Where g (x, n) is one of the gray levels of the two-bit gray value, M(x, n) is the one-dimensional signal data, M min is the minimum value of the one-dimensional signal data, and M max is the one-dimensional signal data. The maximum value, the range of x is the total length of the one-dimensional signal data, n represents the first channel of the current one-dimensional signal, and Chanel max is the number of signal channels;
  • the one-dimensional signal data of the at least one ticket is further converted into the two-dimensional gray value I(x, n) by the first formula. And according to the number of signal channels, it is mapped into the grayscale image in turn, the first formula is Where g (x, n) is one of the gray levels of the two-bit gray value, M(x, n) is the one-dimensional signal data, M min is the minimum value of the one-dimensional signal data, and M max is the one-dimensional signal data.
  • the maximum value, the range of x is the total length of the one-dimensional signal data, n represents the first channel of the current one-dimensional signal, and the Chanel max is the number of signal channels. It should be noted that the number of signal channels and the one-dimensional signal data of the ticket The number corresponds.
  • the one-dimensional signal data of at least one bill is converted into a two-dimensional gray value I(x, n) by the first formula and sequentially mapped to the gray scale image according to the number of signal channels, and the first formula is After that, it is also necessary to detect the two-dimensional gray value I(x, n) by the two-dimensional data processing method and determine whether the ticket is abnormal.
  • a second embodiment of a method for detecting a one-dimensional signal of a ticket according to an embodiment of the present invention includes:
  • the one-dimensional signal data of the ticket collected by at least one sensor is obtained in real time.
  • the one-dimensional signal data of the ticket may be a thickness signal, a magnetic and other authentication signals.
  • 202 Convert at least one ticket one-dimensional signal data into a two-dimensional gray value I(x, n) by a first formula and sequentially map to a grayscale image according to the number of signal channels, and the first formula is Where g (x, n) is one of the gray levels of the two-bit gray value, M(x, n) is the one-dimensional signal data, M min is the minimum value of the one-dimensional signal data, and M max is the one-dimensional signal data. The maximum value, the range of x is the total length of the one-dimensional signal data, n represents the first channel of the current one-dimensional signal, and Chanel max is the number of signal channels;
  • the one-dimensional signal data of the at least one ticket is further converted into the two-dimensional gray value I(x, n) by the first formula. And according to the number of signal channels, it is mapped into the grayscale image in turn, the first formula is Where g (x, n) is one of the gray levels of the two-bit gray value, M(x, n) is the one-dimensional signal data, M min is the minimum value of the one-dimensional signal data, and M max is the one-dimensional signal data.
  • the maximum value, the range of x is the total length of the one-dimensional signal data, n represents the first channel of the current one-dimensional signal, and the Chanel max is the number of signal channels. It should be noted that the number of signal channels and the one-dimensional signal data of the ticket The number corresponds.
  • the one-dimensional signal data of at least one ticket is converted into a two-dimensional gray value I(x, n) by the first formula and sequentially mapped to the grayscale image according to the number of signal channels
  • the two-dimensional gray value I(x, n) is subjected to tilt correction and/or interpolation processing.
  • the interpolation processing is specifically to insert an average value of two adjacent elements in the longitudinal direction of the matrix, for example, if Doing an interpolation process will get
  • the otsu algorithm and the second formula t Max[w0(t)*(u0() are also needed.
  • t)-u) 2 + w1(t)*(u1(t)-u) 2 )] Calculate the threshold t between the foreground and the back of the image, where w0 is the background scale, u0 is the background mean, and w1 is the foreground scale.
  • U1 is the foreground mean and u is the gray mean of the entire image.
  • the abnormal point is marked with a number
  • the marked abnormal point is the number of the left abnormal point
  • the marked abnormal point is the number of the above abnormal point
  • the target abnormal point is the left abnormal point number and the smaller number in the upper abnormal point number, and the larger number is changed to a smaller number;
  • the grayscale map mapped by the two-dimensional gray value I(x, n) needs to be scanned from left to right and from top to bottom, and each of the mapped grayscale maps is compared.
  • the gray level g (x, n) is greater than the threshold value t , the point corresponding to the gray level g (x, n) is marked as an abnormal point, and then the abnormal point is determined. Whether there are abnormal points on the left and above in the four neighborhoods;
  • the abnormal point is marked with a number
  • the marked abnormal point is the number of the left abnormal point
  • the abnormal point is marked as The number of the abnormal point of the surface
  • the target abnormal point is the left abnormal point number and the smaller number in the upper abnormal point number, and the larger number is changed to a smaller number.
  • the grayscale map mapped by the two-dimensional gray value I(x, n) is scanned from left to right and from top to bottom, and the gray scale g of each point in the mapped grayscale image is compared ( x, n) and the magnitude of the threshold t, if the gradation g (x, n) is greater than the threshold t , mark the point corresponding to the gradation g (x, n) as an abnormal point, and then determine the left side of the four points in the abnormal point After there is an abnormal point above, it is also necessary to scan the gray scale map mapped by the two-dimensional gray value I(x, n) of the abnormal point mark and use the summation algorithm to merge the different number of abnormal point regions connected and then calculate The connection domain information of the abnormal point needs to be described. It is a prior art for the person skilled in the art to check the algorithm, and details are not described herein again.
  • the grayscale map mapped by the two-dimensional gray value I(x, n) of the abnormal point mark is scanned, and the different number of abnormal point regions connected by the parallel algorithm are combined by the parallel search algorithm, and then the abnormal point is calculated. After the connected domain information, it is also necessary to compare the information of the abnormal point with the information of the normal ticket to determine whether the ticket is stuck with tape or wrinkles.
  • the two-dimensional gray value of the normal ticket needs to be established by using at least one normal ticket as a sample.
  • the normal reference two-dimensional gray value I n (x, y) is obtained by averaging the two-dimensional gray values of the same position points of at least one normal ticket.
  • the two-dimensional gray value of the normal ticket is respectively established by using at least one normal ticket as a sample, and the two-dimensional gray value of the same position of at least one normal ticket is averaged to obtain a normal reference two-dimensional.
  • the gray value I n (x, y) it is also necessary to match the normal reference two-dimensional gray value I n (x, y) with the two-dimensional gray value I (x, n) and determine the authenticity of the ticket by the matching degree.
  • the normal reference two-dimensional gray value I n (x, y) is matched with the two-dimensional gray value I (x, n) and the matching degree is judged by the matching degree and the note face value is applied to Detection of magnetic and other forging features.
  • a first embodiment of a device for detecting a one-dimensional signal of a ticket includes:
  • the acquiring module 301 is configured to acquire one-dimensional signal data of the ticket collected by the at least one sensor in real time;
  • the two-dimensional data conversion module 302 is configured to convert at least one ticket one-dimensional signal data into a two-dimensional gray value I(x, n) by the first formula and sequentially map to the grayscale image according to the number of signal channels, first Formula is Where g (x, n) is one of the gray levels of the two-bit gray value, M(x, n) is the one-dimensional signal data, M min is the minimum value of the one-dimensional signal data, and M max is the one-dimensional signal data. The maximum value, the range of x is the total length of the one-dimensional signal data, n represents the first channel of the current one-dimensional signal, and Chanel max is the number of signal channels;
  • the two-dimensional data processing module 303 is configured to detect the two-dimensional gray value I(x, n) by the two-dimensional data processing method and determine whether the ticket is abnormal.
  • a second embodiment of a device for detecting a one-dimensional signal of a ticket includes:
  • the obtaining module 401 is configured to acquire one-dimensional signal data of the ticket collected by the at least one sensor in real time;
  • the two-dimensional data conversion module 402 is configured to convert at least one ticket one-dimensional signal data into a two-dimensional gray value I(x, n) by a first formula and sequentially map to a grayscale image according to the number of signal channels, first Formula is Where g (x, n) is one of the gray levels of the two-bit gray value, M(x, n) is the one-dimensional signal data, M min is the minimum value of the one-dimensional signal data, and M max is the one-dimensional signal data. The maximum value, the range of x is the total length of the one-dimensional signal data, n represents the first channel of the current one-dimensional signal, and Chanel max is the number of signal channels;
  • the tilt correction and/or interpolation processing module 403 is configured to perform tilt correction and/or interpolation processing on the two-dimensional gray value I(x, n).
  • the two-dimensional data processing module 404 is configured to detect the two-dimensional gray value I(x, n) by the two-dimensional data processing method and determine whether the ticket is abnormal.
  • the two-dimensional data processing module 404 specifically includes:
  • the threshold calculation unit 4041 is configured to calculate a threshold t between the foreground and the background of the image
  • the abnormal point information calculation unit 4042 compares the magnitudes of the gradations g (x, n) and the thresholds t of the points in the gradation map mapped by the two-dimensional gradation value I(x, n), and sets the gradation g (x, n) a point larger than the threshold t is marked as an abnormal point and then calculates information of the abnormal point;
  • the comparison judging unit 4043 compares the information of the abnormal point with the information of the normal ticket to determine whether the ticket is stuck with tape.
  • the threshold calculation unit 4041 is specifically configured to:
  • the abnormal point information calculation unit 4042 specifically includes:
  • the abnormal point is marked from the unit 40421, and is used for scanning the grayscale map mapped by the two-dimensional gray value I(x, n) from left to right and from top to bottom and comparing the gray scales of the points in the mapped grayscale image.
  • g (x, n) and the magnitude of the threshold t if the gradation g (x, n) is greater than the threshold t , mark the point corresponding to the gradation g (x, n) as an abnormal point, and then determine the four points in the abnormal point Whether there are abnormal points on the left and above;
  • the abnormal point is marked with a number
  • the marked abnormal point is the number of the left abnormal point
  • the marked abnormal point is the number of the above abnormal point
  • the target abnormal point is the left abnormal point number and the smaller number in the upper abnormal point number, and the larger number is changed to a smaller number;
  • the connected domain information calculation subunit 40422 is configured to scan the grayscale map mapped by the two-dimensional grayscale value I(x, n) of the abnormal point mark and combine the different number of abnormal point regions connected by using the parallel search algorithm, and then Calculate the connected domain information of the abnormal point.
  • the two-dimensional data processing module 404 specifically includes:
  • the normal reference two-dimensional data establishing module 4044 is configured to respectively establish a two-dimensional gray value of the normal ticket by using at least one normal ticket as a sample, and obtain an average value of the two-dimensional gray value of the same position of the at least one normal ticket.
  • the matching judging module 4045 is configured to match the normal reference two-dimensional gray value I n (x, y) with the two-dimensional gray value I (x, n) and determine the bill authenticity and the bill face value by the matching degree.
  • the first application example includes: detecting and identifying a test banknote with a vertical adhesive tape attached thereto.
  • the test bank image with the vertical tape attached is shown in Figure 5.
  • the one-dimensional thickness signal of the banknote is obtained, and then the data of the multi-channel thickness sensor is collected.
  • the data of the thickness signal should be as follows:
  • T 1 [0,0,180,181,180,182...]
  • T 2 [0,181,183,180,180,182...]
  • T 3 [180, 180, 180, 181, 180, 180...]
  • T n [0, 182, 180, 181, 180, 182...],
  • the one-dimensional signal is upgraded into a two-dimensional grayscale image.
  • the grayscale image can be obtained as shown in FIG. 8. Then, the grayscale image after multiple interpolation is compared with the calculated threshold value, and the point corresponding to the grayscale value larger than the threshold is marked as abnormal. Point, finally use the algorithm and merge the algorithm to merge the connected abnormal point area, assign the abnormal point area to the new gray value (not 255), and assign the abnormal area to the gray value 255, as shown in Figure 9, where the gray in Figure 9
  • the part is the detected abnormal point area, that is, the tape area
  • the white part (gray value 255) is the banknote area
  • the black part is the invalid data area.
  • the second application example includes: detecting and identifying the more difficult to distinguish the folded banknotes and the corners of the banknotes.
  • the more difficult to distinguish the angled banknotes and the corners of the tapes are shown in Figure 10 and Figure 11, respectively, to obtain the original one-dimensional signal of the folded banknotes and the corners of the tapes.
  • the existing one-dimensional thickness signal detecting method is more based on the calculation of the abnormal portion of the single-path data.
  • the abnormal portion of the data of the single path is as shown in FIG. It is difficult to distinguish between folded-angle banknotes and tape banknotes.
  • multi-channel signals are analyzed simultaneously in one-dimensional thickness signal detection, one-dimensional detection means are less, multi-channel signals are not enough, and the detection effect will be very bad.
  • the algorithm for multi-channel detection of multi-path signals is more complicated and difficult to implement, which will make the detection effect worse.
  • the original one-dimensional signal of the folded banknote and the corner portion of the tape is converted into two-dimensional gray data and mapped into the image, as shown in FIG. 14 and FIG. 15, the gray portion of the figure is the valid data portion, and the black portion is invalid.
  • the data section it can be seen that in the two-dimensional image, the characteristics of the folded-angle banknotes and the tape-type banknotes have been significantly restored, and the general-purpose image detection algorithm can distinguish the folded-angle banknotes and the tape-type banknotes;
  • Fig. 16 and Fig. 17 The final image of the detection of the folded banknotes and the corners of the banknotes is shown in Fig. 16 and Fig. 17, in which the gray part is the detected tape area, the white part is the banknote area, the black part is the invalid data area, and the folded side is the invalid data area.
  • the corner portion of the banknote the corner of the corner of the banknote is defective, and the tape is not, and the distinguishability is made higher by the description of the difference between the two.
  • a third application example includes: detecting and distinguishing between a wrinkled banknote and two irregular tapes.
  • the pleated banknote, the first irregular tape banknote, and the second irregular tape banknote sample diagram are as shown in FIG. 18 to FIG. 20, and the one-dimensional thickness signal of the pleated banknote, the first irregular tape banknote, and the second irregular tape banknote is as shown in FIG.
  • the first A one-dimensional thickness signal of an irregular tape banknote and a second irregular tape banknote is converted into two-dimensional gray scale data and mapped into an image as shown in FIG. 24, FIG. 25, FIG. 26, for the wrinkled banknote, the first irregularity
  • the final image of the tape banknote and the second irregular tape banknote is shown in Fig. 27, Fig. 28, and Fig. 29.
  • the gray portion in the figure is the detected tape region
  • the white portion is the banknote region
  • the black portion is the invalid data region.
  • the tape banknotes and the pleated banknotes can be easily distinguished by describing the area, shape, and size of the connected areas of the image.
  • a fourth application example includes magnetic detection of dollars in different denominations. Samples of $2 and $5 are shown in Figures 31 and 32. The one-dimensional magnetic signals of $2 and $5 are shown in Figures 33 and 34. It can be seen that the magnetic features on the banknotes will be different. Waveform reaction, so there are different one-dimensional magnetic signal waveforms between different denominations, real banknotes and counterfeit banknotes, but in one-dimensional inspection, The positional information response to the feature is not accurate enough.
  • the final image 35 and image 36 of the detection of $2 and $5 are shown in the gray image as the valid data portion and the black portion as the invalid data portion. It can be seen that the two-dimensional image is used for detection. The means detects the above magnetic signals, and the accuracy of the detection can be significantly improved for distinguishing the face value version of the banknote and identifying the counterfeit banknote.
  • the disclosed system, apparatus, and method may be implemented in other manners.
  • the device embodiments described above are merely illustrative.
  • the division of the unit is only a logical function division.
  • there may be another division manner for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored or not executed.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
  • each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
  • the above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.
  • the integrated unit if implemented in the form of a software functional unit and sold or used as a standalone product, may be stored in a computer readable storage medium.
  • the technical solution of the present invention which is essential or contributes to the prior art, or all or part of the technical solution, may be embodied in the form of a software product stored in a storage medium.
  • a number of instructions are included to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention.
  • the foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), A variety of media that can store program code, such as random access memory (RAM), disk, or optical disk.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Inspection Of Paper Currency And Valuable Securities (AREA)

Abstract

一种票据一维信号的检测方法及装置,通过将票据一维信号数据转化为二维灰度值,然后将二维灰度值作预处理,最后通过二维数据处理方法检测二维灰度值从而检测票据是否粘有胶带、真伪以及面值,解决了现有的一维信号检测方法造成的异常样本难以被区分而正常样本又误拒、尤其对于粘贴有透明胶带的钞票很难准确识别等技术问题。

Description

一种票据一维信号的检测方法及装置
本申请要求于2016年11月07号提交中国专利局、申请号为201610975164.5、发明名称为“一种票据一维信号的检测方法及装置”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本发明涉及钞票识别领域,尤其涉及一种票据一维信号的检测方法及装置。
背景技术
经济高速发展,钞票的发行量日益增大,假钞的出现,轻则影响人们的正常生活,重则会造成国家经济不稳定,干扰货币流通的正常秩序,破坏社会信用原则。
票据识别技术日新月异,现有的比较常用的主流票据鉴伪识别技术为一维信号检测,一维信号检测包括磁信号识别和厚度信号识别等。
然而,传统的票据一维信号检测方法存在着一定的局限性,例如:1对于票据特征点的位置信息判定不敏感,尤其在票据运行方向的垂直向分量上的位置关系容易存在误差;2一维信号处理易受干扰,不稳定,正样本与负样本可区分度小;3一维信号处理的算法手段少,并且存在有一定局限性。因此,现有的票据一维信号检测方法容易造成异常样本难以被区分,而正常样本又误拒等问题,尤其对于粘贴有透明胶带的钞票,银行对这种钞票需要进行回收,而这种钞票往往通过现有技术很难进行准确识别,例如常规的厚度检测无法判断钞票粘贴胶带和钞票褶皱两种情况。
发明内容
本发明实施例提供了一种票据一维信号的检测方法及装置,通过将票据一维信号数据转化为二维灰度值,然后将二维灰度值作预处理,最后通过二维数据处理方法检测二维灰度值从而检测票据的是否粘有胶带、真伪以及面值,解决了现有的一维信号检测方法造成的异常样本难以被区分而正常样本又误拒、 尤其对于粘贴有透明胶带的钞票很难准确识别等技术问题。
本发明实施例提供的一种票据一维信号的检测方法,包括:
实时获取至少一路传感器采集到的票据一维信号数据;
将至少一路所述票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中,所述第一公式为
Figure PCTCN2017108559-appb-000001
其中g(x,n)为所述二位灰度值的其中一个灰度,M(x,n)是所述一维信号数据,Mmin为所述一维信号数据中最小值,Mmax为所述一维信号数据中最大值,x的范围是一维信号数据总长,n表示当前所处在所述一维信号的第几通道,Chanelmax为所述信号通道数;
通过二维数据处理方法检测所述二维灰度值I(x,n)并判断票据是否异常。
优选地,
在将至少一路所述票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中之后,在通过二维数据处理方法检测二维灰度值并判断票据是否异常之前还包括:
对所述二维灰度值I(x,n)作倾斜矫正和/或插值处理。
优选地,
通过二维数据处理方法检测所述二维灰度值I(x,n)并判断票据是否异常具体包括:
计算图像前景与后景区分的阈值t;
比较所述二维灰度值I(x,n)映射出的灰度图中各点的灰度g(x,n)与所述阈值t的大小,将所述灰度g(x,n)大于阈值t的点标记为异常点然后计算所述异常点的信息;
将所述异常点的信息与正常票据的信息对比判断所述票据是否粘有胶带或是否褶皱。
优选地,
计算图像前景与后景区分的阈值t具体为:
通过otsu算法和第二公式t=Max[w0(t)*(u0(t)-u)2+w1(t)*(u1(t)-u)2)]计算图像前景与后景区分的阈值t,其中w0为背景比例,u0为背景均值,w1为前景比例,u1为前景均值,u为整幅图像的灰度均值。
优选地,
比较所述二维灰度值I(x,n)映射出的灰度图中各点的灰度g(x,n)与所述阈值t的大小,将所述灰度g(x,n)大于阈值t的点标记为异常点然后计算所述异常点的信息具体包括:
从左向右、自上而下扫描所述二维灰度值I(x,n)映射出的所述灰度图并比较映射出的所述灰度图中各点的灰度g(x,n)与所述阈值t的大小,若所述灰度g(x,n)大于阈值t则将所述灰度g(x,n)对应的点标记为异常点,然后判断所述异常点四邻域中的左面和上面是否有异常点;
若所述异常点四邻域中的左面和上面都没有异常点,则将所述异常点标记一个编号;
若所述异常点四邻域中的左面有异常点,上面没有异常点,则标记所述异常点为左面异常点的编号;
若所述异常点四邻域中的上面有异常点,左面没有异常点,则标记所述异常点为上面异常点的编号;
若所述异常点四邻域中的左面和上面都有异常点,则标所述异常点为左面异常点编号和上面异常点编号中较小的编号,并将较大的编号修改为较小的编号;
扫描经过异常点标记的所述二维灰度值I(x,n)映射出的所述灰度图并利用并查集算法合并连通的不同编号的所述异常点区域,然后计算所述异常点的连通域信息。
优选地,
通过二维数据处理方法检测所述二维灰度值I(x,n)并判断票据是否异常具体包括:
以至少一张正常票据为样本分别建立正常票据的二维灰度值,通过将至少一张正常票据相同位置点的二维灰度值取平均值得到正常参考二维灰度值In(x,y);
将所述正常参考二维灰度值In(x,y)与所述二维灰度值I(x,n)匹配并通过匹配度判断票据真伪和票据面值。
本发明实施例提供了一种票据一维信号的检测装置,包括:
获取模块,用于实时获取至少一路传感器采集到的票据一维信号数据;
二维数据转化模块,用于将至少一路所述票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中,所述第一公式为
Figure PCTCN2017108559-appb-000002
其中g(x,n)为所述二位灰度值的其中一个灰度,M(x,n)是所述一维信号数据,Mmin为所述一维信号数据中最小值,Mmax为所述一维信号数据中最大值,x的范围是一维信号数据总长,n表示当前所处在所述一维信号的第几通道,Chanelmax为所述信号通道数;
二维数据处理模块,用于通过二维数据处理方法检测所述二维灰度值I(x,n)并判断票据是否异常。
优选地,
所述票据一维信号的检测装置还包括:
倾斜矫正和/或插值处理模块,用于对所述二维灰度值I(x,n)作倾斜矫正和/或插值处理。
优选地,
所述二维数据处理模块具体包括:
阈值计算单元,用于计算图像前景与后景区分的阈值t;
异常点信息计算单元,比较所述二维灰度值I(x,n)映射出的灰度图中各点的灰度g(x,n)与所述阈值t的大小,将所述灰度g(x,n)大于阈值t的点标记为异常点然后计算所述异常点的信息;
比较判断单元,将所述异常点的信息与正常票据的信息对比判断所述票据是否粘有胶带。
优选地,
所述阈值计算单元具体用于:
通过otsu算法和第二公式t=Max[w0(t)*(u0(t)-u)2+w1(t)*(u1(t)-u)2)]计算图像前景与后景区分的阈值t,其中w0为背景比例,u0为背景均值,w1为前景比例,u1为前景均值,u为整幅图像的灰度均值。
优选地,
所述异常点信息计算单元具体包括:
异常点标记自单元,用于从左向右、自上而下扫描所述二维灰度值I(x,n)映射出的所述灰度图并比较映射出的所述灰度图中各点的灰度g(x,n)与所述阈值t的大小,若所述灰度g(x,n)大于阈值t则将所述灰度g(x,n)对应的点标记为异常点,然后判断所述异常点四邻域中的左面和上面是否有异常点;
若所述异常点四邻域中的左面和上面都没有异常点,则将所述异常点标记一个编号;
若所述异常点四邻域中的左面有异常点,上面没有异常点,则标记所述异常点为左面异常点的编号;
若所述异常点四邻域中的上面有异常点,左面没有异常点,则标记所述异常点为上面异常点的编号;
若所述异常点四邻域中的左面和上面都有异常点,则标所述异常点为左面异常点编号和上面异常点编号中较小的编号,并将较大的编号修改为较小的编号;
连通域信息计算子单元,用于扫描经过异常点标记的所述二维灰度值I(x,n)映射出的所述灰度图并利用并查集算法合并连通的不同编号的所述异常点区域,然后计算所述异常点的连通域信息。
优选地,
所述二维数据处理模块具体包括:
正常参考二维数据建立模块,用于以至少一张正常票据为样本分别建立正常票据的二维灰度值,通过将至少一张正常票据相同位置点的二维灰度值取平均值得到正常参考二维灰度值In(x,y);
匹配判断模块,用于将所述正常参考二维灰度值In(x,y)与所述二维灰度值I(x,n)匹配并通过匹配度判断票据真伪和票据面值。
从以上技术方案可以看出,本发明实施例具有以下优点:
1、通过将票据一维信号数据转化为二维灰度值,然后将二维灰度值作预处理,最后通过二维数据处理方法检测二维灰度值,从而检测票据的是否粘有胶带、真伪以及面值,解决了现有的一维信号检测方法造成的异常样本难以被区分而正常样本又误拒、尤其对于粘贴有透明胶带的钞票很难准确识别等技术问题,大幅度提高票据鉴伪的能力和识别的稳定性。
2、本发明实施例,完全是基于二维信号做处理,从而能更好的反映票据特征的位置关系,并且可以使用的处理的手段更丰富。
3、在一维信号转换映射至二维信号中,基于单张票据的数据进行转换和映射,更好的保留了票据的特征信息,而又排除了噪声的干扰。
4、对由一维信号数据转化来的二维灰度值作插值处理,有效的丰富了灰度图纵向上的数据,解决了现有的一维信号检测方法对于票据特征点的位置信息判定不敏感,尤其在票据运行方向的垂直向分量上的位置关系容易存在误差的技术问题,同时可以降低真正的噪声。
5、相比于单纯的一维信号处理,通过增维处理,可以使检测更直观,既保留一维信号的特征,又降低了一维信号所携带的噪声,同时,票据正负样本的区分度也更加明显,检测效果更好。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动性的前提下,还可以根据这些附图获得其它的附图。
图1为本发明实施例提供的一种票据一维信号的检测方法的第一实施例的流程示意图;
图2为本发明实施例提供的一种票据一维信号的检测方法的第二实施例的流程示意图;
图3为本发明实施例提供的一种票据一维信号的检测方法的第一实施例的结构示意图;
图4为本发明实施例提供的一种票据一维信号的检测方法的第二实施例的结构示意图;
图5为本发明实施例提供的一种票据一维信号的检测方法的第一应用例中粘贴了竖贴胶带的测试钞图像;
图6为本发明实施例提供的一种票据一维信号的检测方法的第一应用例中粘贴了竖贴胶带的测试钞的厚度信号示意图;
图7为本发明实施例提供的一种票据一维信号的检测方法的第一应用例中将测试钞一维信号升维成二维灰度图后的图像;
图8为本发明实施例提供的一种票据一维信号的检测方法的第一应用例中对二维灰度图的二维灰度数据作多次插值之后的图像;
图9为本发明实施例提供的一种票据一维信号的检测方法的第一应用例中测试钞经过检测之后的图像;
图10和图11分别为本发明实施例提供的一种票据一维信号的检测方法的第二应用例中折角钞票和边角部分胶带钞样本图;
图12和图13分别为本发明实施例提供的一种票据一维信号的检测方法的第二应用例中折角钞票和边角部分胶带钞的原始一维信号示意图;
图14和图15分别为本发明实施例提供的一种票据一维信号的检测方法的第二应用例中折角钞票和边角部分胶带钞的二维灰度图;
图16和图17分别为本发明实施例提供的一种票据一维信号的检测方法的第二应用例中折角钞票和边角部分胶带钞的检测最终图像;
图18、图19和图20分别为本发明实施例提供的一种票据一维信号的检测方法的第三应用例中褶皱钞票、第一不规则胶带钞、第二不规则胶带钞样本图;
图21、图22和图23分别为本发明实施例提供的一种票据一维信号的检测方法的第三应用例中褶皱钞票、第一不规则胶带钞、第二不规则胶带钞样的一维厚度信号图;
图24、图25和图26分别为本发明实施例提供的一种票据一维信号的检测方法的第三应用例中褶皱钞票、第一不规则胶带钞、第二不规则胶带钞样的二维灰度图;
图27、图28和图29分别为本发明实施例提供的一种票据一维信号的检测方法的第三应用例中褶皱钞票、第一不规则胶带钞、第二不规则胶带钞样的检测最终图像;
图30和图31分别为本发明实施例提供的一种票据一维信号的检测方法的第四应用例中2美元与5美元的样本图;
图32和图33分别为本发明实施例提供的一种票据一维信号的检测方法的第四应用例中2美元与5美元的一维磁信号示意图;
图34和图35分别为本发明实施例提供的一种票据一维信号的检测方法的第四应用例中2美元与5美元的二维磁信号图像。
具体实施方式
本发明实施例提供了一种票据一维信号的检测方法及装置,通过将票据一维信号数据转化为二维灰度值,然后将二维灰度值作预处理,最后通过二维数据处理方法检测二维灰度值从而检测票据的是否粘有胶带、真伪以及面值,解决了现有的一维信号检测方法造成的异常样本难以被区分而正常样本又误拒、尤其对于粘贴有透明胶带的钞票很难准确识别等技术问题
为使得本发明的发明目的、特征、优点能够更加的明显和易懂,下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,下面所描述的实施例仅仅是本发明一部分实施例,而非全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其它实施例,都属于本发明保护的范围。
请参阅图1,本发明实施例提供的一种票据一维信号的检测方法的第一实施例,包括:
101,实时获取至少一路传感器采集到的票据一维信号数据;
本发明实施例中,首先需要实时获取至少一路传感器采集到的票据一维信号数据。
102,将至少一路票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中,第一公式为
Figure PCTCN2017108559-appb-000003
其中g(x,n)为二位灰度值的其中一个灰度,M(x,n)是一维信号数据,Mmin为一维信号数据中最小值,Mmax为一维信号数据中最大值,x的范围是一维信号数据总长,n表示当前所处在一维信号的第几通道,Chanelmax为信号通道数;
本发明实施例中,在实时获取至少一路传感器采集到的票据一维信号数据之后,还需要将至少一路票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中,第一公式为
Figure PCTCN2017108559-appb-000004
其中g(x,n)为二位灰度值的其中一个灰度,M(x,n)是一维信号数据,Mmin为一维信号数据中最小值,Mmax为一维信号数据中最大值,x的范围是一维信号数据总长,n表示当前所处在一维信号的第几通道,Chanelmax为信号通道数,需要说明的是,信号通道数与票据一维信号数据的路数相对应。
103,通过二维数据处理方法检测二维灰度值I(x,n)并判断票据是否异常;
本发明实施例中,在将至少一路票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中,第一公式为
Figure PCTCN2017108559-appb-000005
之后,还需要通过二维数据处理方法检测二维灰度值I(x,n)并判断票据是否异常。
请参阅图2,本发明实施例提供的一种票据一维信号的检测方法的第二实施例,包括:
201,实时获取至少一路传感器采集到的票据一维信号数据;
本发明实施例中,首先实时获取至少一路传感器采集到的票据一维信号数据,需要说明的是,票据一维信号数据可以是厚度信号、磁等鉴伪信号。
202,将至少一路票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中,第一公式为
Figure PCTCN2017108559-appb-000006
其中g(x,n)为二位灰度值的其中一个灰度,M(x,n)是一维信号数据,Mmin为一维信号数据中最小值,Mmax为一维信号数据中最大值,x的范围是一维信号数据总长,n表示当前所处在一维信号的第几通道,Chanelmax为信号通道数;
本发明实施例中,在实时获取至少一路传感器采集到的票据一维信号数据之后,还需要将至少一路票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中,第一公式为
Figure PCTCN2017108559-appb-000007
其中g(x,n)为二位灰度值的其中一个灰度,M(x,n)是一维信号数据,Mmin为一维信号数据中最小值,Mmax为一维信号数据中最大值,x的范围是一维信号数据总长,n表示当前所处在一维信号的第几通道,Chanelmax为信号通道数,需要说明的是,信号通道数与票据一维信号数据的路数相对应。
203,对二维灰度值I(x,n)作倾斜矫正和/或插值处理;
本发明实施例中,在将至少一路票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中之后,还需要对二维灰度值I(x,n)作倾斜矫正和/或插值处理,需要说明的是,插值处理具体为在矩阵的纵向上插入临近两个元素的平均值,例如若对
Figure PCTCN2017108559-appb-000008
作一次插值处理,将会得到
Figure PCTCN2017108559-appb-000009
204,通过otsu算法和第二公式t=Max[w0(t)*(u0(t)-u)2+w1(t)*(u1(t)-u)2)]计 算图像前景与后景区分的阈值t,其中w0为背景比例,u0为背景均值,w1为前景比例,u1为前景均值,u为整幅图像的灰度均值;
本发明实施例中,在对二维灰度值I(x,n)作倾斜矫正和/或插值处理之后,还需要通过otsu算法和第二公式t=Max[w0(t)*(u0(t)-u)2+w1(t)*(u1(t)-u)2)]计算图像前景与后景区分的阈值t,其中w0为背景比例,u0为背景均值,w1为前景比例,u1为前景均值,u为整幅图像的灰度均值。
205,从左向右、自上而下扫描二维灰度值I(x,n)映射出的灰度图并比较映射出的灰度图中各点的灰度g(x,n)与阈值t的大小,若灰度g(x,n)大于阈值t则将灰度g(x,n)对应的点标记为异常点,然后判断异常点四邻域中的左面和上面是否有异常点;
若异常点四邻域中的左面和上面都没有异常点,则将异常点标记一个编号;
若异常点四邻域中的左面有异常点,上面没有异常点,则标记异常点为左面异常点的编号;
若异常点四邻域中的上面有异常点,左面没有异常点,则标记异常点为上面异常点的编号;
若异常点四邻域中的左面和上面都有异常点,则标异常点为左面异常点编号和上面异常点编号中较小的编号,并将较大的编号修改为较小的编号;
本发明实施例中,在通过otsu算法和第二公式t=Max[w0(t)*(u0(t)-u)2+w1(t)*(u1(t)-u)2)]计算图像前景与后景区分的阈值t之后,还需要从左向右、自上而下扫描二维灰度值I(x,n)映射出的灰度图并比较映射出的灰度图中各点的灰度g(x,n)与阈值t的大小,若灰度g(x,n)大于阈值t则将灰度g(x,n)对应的点标记为异常点,然后判断异常点四邻域中的左面和上面是否有异常点;
若异常点四邻域中的左面和上面都没有异常点,则将异常点标记一个编号;
若异常点四邻域中的左面有异常点,上面没有异常点,则标记异常点为左面异常点的编号;
若异常点四邻域中的上面有异常点,左面没有异常点,则标记异常点为上 面异常点的编号;
若异常点四邻域中的左面和上面都有异常点,则标异常点为左面异常点编号和上面异常点编号中较小的编号,并将较大的编号修改为较小的编号。
206,扫描经过异常点标记的二维灰度值I(x,n)映射出的灰度图并利用并查集算法合并连通的不同编号的异常点区域,然后计算异常点的连通域信息。
本发明实施例中,在从左向右、自上而下扫描二维灰度值I(x,n)映射出的灰度图并比较映射出的灰度图中各点的灰度g(x,n)与阈值t的大小,若灰度g(x,n)大于阈值t则将灰度g(x,n)对应的点标记为异常点,然后判断异常点四邻域中的左面和上面是否有异常点之后,还需要扫描经过异常点标记的二维灰度值I(x,n)映射出的灰度图并利用并查集算法合并连通的不同编号的异常点区域,然后计算异常点的连通域信息,需要说明的是,对于本领域技术人员来说并查集算法是现有技术,在此不再赘述。
207,将异常点的信息与正常票据的信息对比判断票据是否粘有胶带或是否褶皱;
本发明实施例中,在扫描经过异常点标记的二维灰度值I(x,n)映射出的灰度图并利用并查集算法合并连通的不同编号的异常点区域,然后计算异常点的连通域信息之后,还需要将异常点的信息与正常票据的信息对比判断票据是否粘有胶带或是否褶皱。
208,以至少一张正常票据为样本分别建立正常票据的二维灰度值,通过将至少一张正常票据相同位置点的二维灰度值取平均值得到正常参考二维灰度值In(x,y);
本发明实施例中,在对二维灰度值I(x,n)作倾斜矫正和/或插值处理之后,还需要以至少一张正常票据为样本分别建立正常票据的二维灰度值,通过将至少一张正常票据相同位置点的二维灰度值取平均值得到正常参考二维灰度值In(x,y)。
209,将正常参考二维灰度值In(x,y)与二维灰度值I(x,n)匹配并通过匹配度判断票据真伪和票据面值;
本发明实施例中,在以至少一张正常票据为样本分别建立正常票据的二维灰度值,通过将至少一张正常票据相同位置点的二维灰度值取平均值得到正常 参考二维灰度值In(x,y)之后,还需要将正常参考二维灰度值In(x,y)与二维灰度值I(x,n)匹配并通过匹配度判断票据真伪和票据面值,需要说明的是将正常参考二维灰度值In(x,y)与二维灰度值I(x,n)匹配并通过匹配度判断票据真伪和票据面值适用于基于磁等鉴伪特征的检测。
请参阅图3,本发明实施例提供的一种票据一维信号的检测装置的第一实施例,包括:
获取模块301,用于实时获取至少一路传感器采集到的票据一维信号数据;
二维数据转化模块302,用于将至少一路票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中,第一公式为
Figure PCTCN2017108559-appb-000010
其中g(x,n)为二位灰度值的其中一个灰度,M(x,n)是一维信号数据,Mmin为一维信号数据中最小值,Mmax为一维信号数据中最大值,x的范围是一维信号数据总长,n表示当前所处在一维信号的第几通道,Chanelmax为信号通道数;
二维数据处理模块303,用于通过二维数据处理方法检测二维灰度值I(x,n)并判断票据是否异常。
请参阅图4,本发明实施例提供的一种票据一维信号的检测装置的第二实施例,包括:
获取模块401,用于实时获取至少一路传感器采集到的票据一维信号数据;
二维数据转化模块402,用于将至少一路票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中,第一公式 为
Figure PCTCN2017108559-appb-000011
其中g(x,n)为二位灰度值的其中一个灰度,M(x,n)是一维信号数据,Mmin为一维信号数据中最小值,Mmax为一维信号数据中最大值,x的范围是一维信号数据总长,n表示当前所处在一维信号的第几通道,Chanelmax为信号通道数;
倾斜矫正和/或插值处理模块403,用于对二维灰度值I(x,n)作倾斜矫正和/或插值处理。
二维数据处理模块404,用于通过二维数据处理方法检测二维灰度值I(x,n)并判断票据是否异常。
二维数据处理模块404具体包括:
阈值计算单元4041,用于计算图像前景与后景区分的阈值t;
异常点信息计算单元4042,比较二维灰度值I(x,n)映射出的灰度图中各点的灰度g(x,n)与阈值t的大小,将灰度g(x,n)大于阈值t的点标记为异常点然后计算异常点的信息;
比较判断单元4043,将异常点的信息与正常票据的信息对比判断票据是否粘有胶带。
阈值计算单元4041具体用于:
通过otsu算法和第二公式t=Max[w0(t)*(u0(t)-u)2+w1(t)*(u1(t)-u)2)]计算图像前景与后景区分的阈值t,其中w0为背景比例,u0为背景均值,w1为前景比例,u1为前景均值,u为整幅图像的灰度均值。
异常点信息计算单元4042具体包括:
异常点标记自单元40421,用于从左向右、自上而下扫描二维灰度值I(x,n)映射出的灰度图并比较映射出的灰度图中各点的灰度g(x,n)与阈值t的大小,若灰度g(x,n)大于阈值t则将灰度g(x,n)对应的点标记为异常点,然后判断异常点四 邻域中的左面和上面是否有异常点;
若异常点四邻域中的左面和上面都没有异常点,则将异常点标记一个编号;
若异常点四邻域中的左面有异常点,上面没有异常点,则标记异常点为左面异常点的编号;
若异常点四邻域中的上面有异常点,左面没有异常点,则标记异常点为上面异常点的编号;
若异常点四邻域中的左面和上面都有异常点,则标异常点为左面异常点编号和上面异常点编号中较小的编号,并将较大的编号修改为较小的编号;
连通域信息计算子单元40422,用于扫描经过异常点标记的二维灰度值I(x,n)映射出的灰度图并利用并查集算法合并连通的不同编号的异常点区域,然后计算异常点的连通域信息。
二维数据处理模块404具体包括:
正常参考二维数据建立模块4044,用于以至少一张正常票据为样本分别建立正常票据的二维灰度值,通过将至少一张正常票据相同位置点的二维灰度值取平均值得到正常参考二维灰度值In(x,y);
匹配判断模块4045,用于将正常参考二维灰度值In(x,y)与二维灰度值I(x,n)匹配并通过匹配度判断票据真伪和票据面值。
上面是对一种票据一维信号的检测方法及装置进行的详细说明,为便于理解,下面将以一具体应用场景对一种票据一维信号的检测方法及装置的应用进行说明;
请参阅图5至图9第一应用例包括:对一张粘贴了竖贴胶带的测试钞进行检测识别。粘贴了竖贴胶带的测试钞图像如图5所示,首先获取钞票的一维厚度信号,然后通过多路厚度传感器的数据采集,其厚度信号的数据应该如下:
第一路:T1=[0,0,180,181,180,182......]
第二路:T2=[0,181,183,180,180,182......]
第三路:T3=[180,180,180,181,180,180......]
第四路:T4=[0,0,0,181,180,182,180......]
……
第n路:Tn=[0,182,180,181,180,182......],
直观的示意图如图6所示;
然后将一维信号升维成二维灰度图
Figure PCTCN2017108559-appb-000012
实际的图像效果如图7所示;
接着对图像进行倾斜矫正得到
Figure PCTCN2017108559-appb-000013
之后对图像插值,丰富数据得到
Figure PCTCN2017108559-appb-000014
重复多次图像插值后可以得到灰度图如图8所示;然后对多次插值后的灰度图作灰度值与计算的阈值比较,将大于阈值的灰度值对应的点标记为异常点,最后利用并查集算法合并连通的异常点区域,将异常点区域赋予新灰度值(非255),将异常区域赋予灰度值255,如图9所示,其中,图9中灰色部分为检测出的异常点区域即胶带区域,白色部分(灰度值255)为钞票区域,黑色部分为无效数据区域。
请参阅图10至图17,第二应用例包括:对较难区分的折角钞票和边角部分胶带钞进行检测识别。较难区分的折角钞票和边角部分胶带钞样图分别如图10和图11所示,分别获取折角钞票和边角部分胶带钞的原始一维信号示意图 如图12如13所示,现有的一维厚度信号检测方法更多的是基于单一路数据异常部分的计算,照此,从图12和图13中可以看出,单一路的数据异常部分很难区分折角钞和胶带钞,如果在一维厚度信号检测中基于多路信号同时分析,一维检测的手段比较少,多路信号联系性又不够,检测效果会很不好,如果使用一维检测的算法综合计算法多路信号时比较复杂,不易实现,将会使检测效果更加不好;
将折角钞票和边角部分胶带钞的原始一维信号转化成二维灰度数据并映射到图像中去,如图14和图15所示,图中灰色部分为有效数据部分,黑色部分为无效数据部分,可以看出,二维图像中,折角钞和胶带钞的特征已经明显还原,通过一般的图像检测算法即可区分出折角钞和胶带钞;
对折角钞票和边角部分胶带钞的检测最终图像如图16和图17所示,图中灰色部分为检测出的胶带区域,白色部分为钞票区域,黑色部分为无效数据区域,折角钞与边角部分胶带钞相比,折角钞折角区域有缺损,而胶带钞没有,通过对二者区别的描述,可区分性变得更高。
请参阅图18至图29第三应用例包括:对褶皱钞票与两种不规则胶带钞进行检测区分。褶皱钞票、第一不规则胶带钞、第二不规则胶带钞样本图如图18至图20所示,褶皱钞票、第一不规则胶带钞、第二不规则胶带钞的一维厚度信号如图21至图23所示,从图中所框示区域看,几乎每一路信号都存在异常部分,但如果不是整体考虑信号各突出部分,很难区分出褶皱钞和胶带钞,将褶皱钞票、第一不规则胶带钞、第二不规则胶带钞的一维厚度信号转化成二维灰度数据并映射到图像中去如图24、图25、图26所示,对褶皱钞票、第一不规则胶带钞、第二不规则胶带钞的检测最终图像如图27、图28、图29所示,图中灰色部分为检测出的胶带区域,白色部分为钞票区域,黑色部分为无效数据区域。通过对图像的连通区域的面积、形状和大小的描述,可以很容易区分出胶带钞与褶皱钞。
请参阅图30至图29第四应用例包括:对不同面值的美元的磁检测。2美元与5美元的样本如图31和图32所示,2美元与5美元的一维磁信号示意图如图33和图34所示,可以看出,钞票上的磁特征部分会有不同的波形反应,因此不同面值,真钞和假钞之间有着不同的一维磁信号波形,但是一维检测中, 对于特征的位置信息反应不够精确,2美元与5美元的检测最终图像35和图像36所示,其中灰色部分为有效数据部分,黑色部分为无效数据部分,可以看出,使用二维图像检测的手段对以上磁信号进行检测,对于区分钞票的面值版本和鉴别伪钞,检测的精度可以明显提升。
所属领域的技术人员可以清楚地了解到,为描述的方便和简洁,上述描述的系统,装置和单元的具体工作过程,可以参考前述方法实施例中的对应过程,在此不再赘述。
在本申请所提供的几个实施例中,应该理解到,所揭露的系统,装置和方法,可以通过其它的方式实现。例如,以上所描述的装置实施例仅仅是示意性的,例如,所述单元的划分,仅仅为一种逻辑功能划分,实际实现时可以有另外的划分方式,例如多个单元或组件可以结合或者可以集成到另一个系统,或一些特征可以忽略,或不执行。另一点,所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口,装置或单元的间接耦合或通信连接,可以是电性,机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的,作为单元显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外,在本发明各个实施例中的各功能单元可以集成在一个处理单元中,也可以是各个单元单独物理存在,也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现,也可以采用软件功能单元的形式实现。
所述集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时,可以存储在一个计算机可读取存储介质中。基于这样的理解,本发明的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的全部或部分可以以软件产品的形式体现出来,该计算机软件产品存储在一个存储介质中,包括若干指令用以使得一台计算机设备(可以是个人计算机,服务器,或者网络设备等)执行本发明各个实施例所述方法的全部或部分步骤。而前述的存储介质包括:U盘、移动硬盘、只读存储器(ROM,Read-Only Memory)、 随机存取存储器(RAM,Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述,以上实施例仅用以说明本发明的技术方案,而非对其限制;尽管参照前述实施例对本发明进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分技术特征进行等同替换;而这些修改或者替换,并不使相应技术方案的本质脱离本发明各实施例技术方案的精神和范围。

Claims (12)

  1. 一种票据一维信号的检测方法,其特征在于,包括:
    实时获取至少一路传感器采集到的票据一维信号数据;
    将至少一路所述票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中,所述第一公式为
    Figure PCTCN2017108559-appb-100001
    其中g(x,n)为所述二位灰度值的其中一个灰度,M(x,n)是所述一维信号数据,Mmin为所述一维信号数据中最小值,Mmax为所述一维信号数据中最大值,x的范围是一维信号数据总长,n表示当前所处在所述一维信号的第几通道,Chanelmax为所述信号通道数;
    通过二维数据处理方法检测所述二维灰度值I(x,n)并判断票据是否异常。
  2. 根据权利要求1所述的票据一维信号的检测方法,其特征在于,在将至少一路所述票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中之后,在通过二维数据处理方法检测二维灰度值并判断票据是否异常之前还包括:
    对所述二维灰度值I(x,n)作倾斜矫正和/或插值处理。
  3. 根据权利要求1所述的票据一维信号的检测方法,其特征在于,通过二维数据处理方法检测所述二维灰度值I(x,n)并判断票据是否异常具体包括:
    计算图像前景与后景区分的阈值t;
    比较所述二维灰度值I(x,n)映射出的灰度图中各点的灰度g(x,n)与所述阈值t的大小,将所述灰度g(x,n)大于阈值t的点标记为异常点然后计算所述异常点的信息;
    将所述异常点的信息与正常票据的信息对比判断所述票据是否粘有胶带或是否褶皱。
  4. 根据权利要求3所述的票据一维信号的检测方法,其特征在于,计算图像前景与后景区分的阈值t具体为:
    通过otsu算法和第二公式t=Max[w0(t)*(u0(t)-u)2+w1(t)*(u1(t)-u)2)]计算图像前景与后景区分的阈值t,其中w0为背景比例,u0为背景均值,w1为前景比例,u1为前景均值,u为整幅图像的灰度均值。
  5. 根据权利要求3所述的票据一维信号的检测方法,其特征在于,比较所述二维灰度值I(x,n)映射出的灰度图中各点的灰度g(x,n)与所述阈值t的大小,将所述灰度g(x,n)大于阈值t的点标记为异常点然后计算所述异常点的信息具体包括:
    从左向右、自上而下扫描所述二维灰度值I(x,n)映射出的所述灰度图并比较映射出的所述灰度图中各点的灰度g(x,n)与所述阈值t的大小,若所述灰度g(x,n)大于阈值t则将所述灰度g(x,n)对应的点标记为异常点,然后判断所述异常点四邻域中的左面和上面是否有异常点;
    若所述异常点四邻域中的左面和上面都没有异常点,则将所述异常点标记一个编号;
    若所述异常点四邻域中的左面有异常点,上面没有异常点,则标记所述异常点为左面异常点的编号;
    若所述异常点四邻域中的上面有异常点,左面没有异常点,则标记所述异常点为上面异常点的编号;
    若所述异常点四邻域中的左面和上面都有异常点,则标所述异常点为左面异常点编号和上面异常点编号中较小的编号,并将较大的编号修改为较小的编号;
    扫描经过异常点标记的所述二维灰度值I(x,n)映射出的所述灰度图并利用并查集算法合并连通的不同编号的所述异常点区域,然后计算所述异常点的连通域信息。
  6. 根据权利要求1所述的票据一维信号的检测方法,其特征在于,通过二维数据处理方法检测所述二维灰度值I(x,n)并判断票据是否异常具体包括:
    以至少一张正常票据为样本分别建立正常票据的二维灰度值,通过将 至少一张正常票据相同位置点的二维灰度值取平均值得到正常参考二维灰度值In(x,y);
    将所述正常参考二维灰度值In(x,y)与所述二维灰度值I(x,n)匹配并通过匹配度判断票据真伪和票据面值。
  7. 一种票据一维信号的检测装置,其特征在于,包括:
    获取模块,用于实时获取至少一路传感器采集到的票据一维信号数据;
    二维数据转化模块,用于将至少一路所述票据一维信号数据通过第一公式分别转化为二维灰度值I(x,n)并按信号通道数依次映射到灰度图中,所述第一公式为
    Figure PCTCN2017108559-appb-100002
    其中g(x,n)为所述二位灰度值的其中一个灰度,M(x,n)是所述一维信号数据,Mmin为所述一维信号数据中最小值,Mmax为所述一维信号数据中最大值,x的范围是一维信号数据总长,n表示当前所处在所述一维信号的第几通道,Chanelmax为所述信号通道数;
    二维数据处理模块,用于通过二维数据处理方法检测所述二维灰度值I(x,n)并判断票据是否异常。
  8. 根据权利要求7所述的票据一维信号的检测装置,其特征在于,还包括:
    倾斜矫正和/或插值处理模块,用于对所述二维灰度值I(x,n)作倾斜矫正和/或插值处理。
  9. 根据权利要求7所述的票据一维信号的检测装置,其特征在于,所述二维数据处理模块具体包括:
    阈值计算单元,用于计算图像前景与后景区分的阈值t;
    异常点信息计算单元,比较所述二维灰度值I(x,n)映射出的灰度图中各点的灰度g(x,n)与所述阈值t的大小,将所述灰度g(x,n)大于阈值t的点标记为异常点然后计算所述异常点的信息;
    比较判断单元,将所述异常点的信息与正常票据的信息对比判断所述票据是否粘有胶带。
  10. 根据权利要求9所述的票据一维信号的检测装置,其特征在于,所述阈值计算单元具体用于:
    通过otsu算法和第二公式t=Max[w0(t)*(u0(t)-u)2+w1(t)*(u1(t)-u)2)]计算图像前景与后景区分的阈值t,其中w0为背景比例,u0为背景均值,w1为前景比例,u1为前景均值,u为整幅图像的灰度均值。
  11. 根据权利要求9所述的票据一维信号的检测装置,其特征在于,所述异常点信息计算单元具体包括:
    异常点标记自单元,用于从左向右、自上而下扫描所述二维灰度值I(x,n)映射出的所述灰度图并比较映射出的所述灰度图中各点的灰度g(x,n)与所述阈值t的大小,若所述灰度g(x,n)大于阈值t则将所述灰度g(x,n)对应的点标记为异常点,然后判断所述异常点四邻域中的左面和上面是否有异常点;
    若所述异常点四邻域中的左面和上面都没有异常点,则将所述异常点标记一个编号;
    若所述异常点四邻域中的左面有异常点,上面没有异常点,则标记所述异常点为左面异常点的编号;
    若所述异常点四邻域中的上面有异常点,左面没有异常点,则标记所述异常点为上面异常点的编号;
    若所述异常点四邻域中的左面和上面都有异常点,则标所述异常点为左面异常点编号和上面异常点编号中较小的编号,并将较大的编号修改为较小的编号;
    连通域信息计算子单元,用于扫描经过异常点标记的所述二维灰度值I(x,n)映射出的所述灰度图并利用并查集算法合并连通的不同编号的所述异常点区域,然后计算所述异常点的连通域信息。
  12. 根据权利要求7所述的票据一维信号的检测装置,其特征在于,所述二维数据处理模块具体包括:
    正常参考二维数据建立模块,用于以至少一张正常票据为样本分别建立正常票据的二维灰度值,通过将至少一张正常票据相同位置点的二维灰度值取平均值得到正常参考二维灰度值In(x,y);
    匹配判断模块,用于将所述正常参考二维灰度值In(x,y)与所述二维灰 度值I(x,n)匹配并通过匹配度判断票据真伪和票据面值。
PCT/CN2017/108559 2016-11-07 2017-10-31 一种票据一维信号的检测方法及装置 WO2018082540A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610975164.5 2016-11-07
CN201610975164.5A CN106548558B (zh) 2016-11-07 2016-11-07 一种票据一维信号的检测方法及装置

Publications (1)

Publication Number Publication Date
WO2018082540A1 true WO2018082540A1 (zh) 2018-05-11

Family

ID=58394267

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/108559 WO2018082540A1 (zh) 2016-11-07 2017-10-31 一种票据一维信号的检测方法及装置

Country Status (2)

Country Link
CN (1) CN106548558B (zh)
WO (1) WO2018082540A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110146589A (zh) * 2019-06-17 2019-08-20 东北大学 一种管道漏磁数据的高清可视化方法
CN111832428A (zh) * 2020-06-23 2020-10-27 北京科技大学 一种应用于冷轧轧机断带故障诊断的数据增强方法

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106548558B (zh) * 2016-11-07 2019-07-23 广州广电运通金融电子股份有限公司 一种票据一维信号的检测方法及装置
CN107134046B (zh) * 2017-05-02 2019-08-23 深圳怡化电脑股份有限公司 一种纸币厚度异常检测方法及装置
CN110222695B (zh) * 2019-06-19 2021-11-02 拉扎斯网络科技(上海)有限公司 一种证件图片处理方法及装置、介质、电子设备
CN111476937B (zh) * 2020-04-18 2021-11-16 武汉卓目科技有限公司 金融机具二维信号辅助一维信号鉴伪的方法及装置

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3390146B2 (ja) * 1999-05-10 2003-03-24 冨士色素株式会社 偽札検知またはセキュリティ認証インキ及び検知認証方法
US20100144291A1 (en) * 2008-12-08 2010-06-10 Georgios Stylianou Vision assistance using mobile telephone
CN102324134A (zh) * 2011-09-19 2012-01-18 广州广电运通金融电子股份有限公司 有价文件鉴别方法及其鉴别装置
CN104036290A (zh) * 2014-06-11 2014-09-10 新达通科技股份有限公司 一种纸币面值识别方法及装置
CN104658097A (zh) * 2015-03-11 2015-05-27 华中科技大学 一种基于图像的直方图匹配的人民币纸币面额识别方法
CN105447956A (zh) * 2015-11-06 2016-03-30 东方通信股份有限公司 一种拼接纸币的检测方法
CN105957238A (zh) * 2016-05-20 2016-09-21 聚龙股份有限公司 一种纸币管理方法及其系统
CN106548558A (zh) * 2016-11-07 2017-03-29 广州广电运通金融电子股份有限公司 一种票据一维信号的检测方法及装置

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3390146B2 (ja) * 1999-05-10 2003-03-24 冨士色素株式会社 偽札検知またはセキュリティ認証インキ及び検知認証方法
US20100144291A1 (en) * 2008-12-08 2010-06-10 Georgios Stylianou Vision assistance using mobile telephone
CN102324134A (zh) * 2011-09-19 2012-01-18 广州广电运通金融电子股份有限公司 有价文件鉴别方法及其鉴别装置
CN104036290A (zh) * 2014-06-11 2014-09-10 新达通科技股份有限公司 一种纸币面值识别方法及装置
CN104658097A (zh) * 2015-03-11 2015-05-27 华中科技大学 一种基于图像的直方图匹配的人民币纸币面额识别方法
CN105447956A (zh) * 2015-11-06 2016-03-30 东方通信股份有限公司 一种拼接纸币的检测方法
CN105957238A (zh) * 2016-05-20 2016-09-21 聚龙股份有限公司 一种纸币管理方法及其系统
CN106548558A (zh) * 2016-11-07 2017-03-29 广州广电运通金融电子股份有限公司 一种票据一维信号的检测方法及装置

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110146589A (zh) * 2019-06-17 2019-08-20 东北大学 一种管道漏磁数据的高清可视化方法
CN110146589B (zh) * 2019-06-17 2022-12-30 东北大学 一种管道漏磁数据的高清可视化方法
CN111832428A (zh) * 2020-06-23 2020-10-27 北京科技大学 一种应用于冷轧轧机断带故障诊断的数据增强方法
CN111832428B (zh) * 2020-06-23 2024-02-23 北京科技大学 一种应用于冷轧轧机断带故障诊断的数据增强方法

Also Published As

Publication number Publication date
CN106548558A (zh) 2017-03-29
CN106548558B (zh) 2019-07-23

Similar Documents

Publication Publication Date Title
WO2018082540A1 (zh) 一种票据一维信号的检测方法及装置
JP5616958B2 (ja) 紙幣検出器デバイスのための方法、および紙幣検出器デバイス
CN106056751B (zh) 冠字号码的识别方法及系统
EP2518684B1 (en) Fake finger determination device
CN106557764B (zh) 一种基于二进制编码字符水尺和图像处理的水位识别方法
CN107085883B (zh) 一种纸币识别的方法和装置
TW201604797A (zh) 用於檢測預定區域中特定標識圖像的方法及裝置
CN105374105A (zh) 一种用于移动终端鉴别纸币真伪的方法
CN105551133B (zh) 一种纸币拼接缝或折痕的识别方法及系统
WO2013170663A1 (zh) 一种纸类识别方法及相关装置
CN108320373B (zh) 一种纸币防伪标识的检测的方法及装置
CN106952393B (zh) 纸币识别方法和装置、电子设备和存储介质
CN110956737B (zh) 一种安全线识别方法和装置
CN104966349A (zh) 一种基于atm平台的纸币新旧检测方法
KR102007685B1 (ko) 하이브리드 위폐 감별 장치 및 시스템
CN105046808A (zh) 一种纸币多光谱高分辨率图像采集系统及采集方法
WO2019001191A1 (zh) 一种拼接票据的检测方法及拼接票据的检测装置
KR101001691B1 (ko) 웨이블렛 변환을 이용한 권종인식방법
CN109409158A (zh) 一种基于二维码边缘粗糙度的防伪方法
WO2018107574A1 (zh) 一种对印防伪特征的检测方法及装置
RU2688306C1 (ru) Способ и устройство обнаружения для наслоившихся друг на друга банкнот
WO2015085817A1 (zh) 一种多光谱券类质量检测方法和系统
CN106447908B (zh) 一种纸币鉴伪方法及装置
CN109767543B (zh) 一种纸币检测方法和系统
JPH09305768A (ja) 画像パターン識別装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17867838

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17867838

Country of ref document: EP

Kind code of ref document: A1