Bill handling device and correction method

Abstract

In a bill handling device (1), an image sensor module (10) has an upper line sensor (11S) and a lower line sensor (12S) for capturing a bill image, which is an image of a bill, and a control board (20) corrects bill pixel values, which are pixel values of pixels forming the bill image, by using a correction coefficient calculated from a reference value and medium pixel values which are pixel values of pixels forming a medium image which is an image at a time when a prescribed medium is captured by the upper line sensor (11S) and the lower line sensor (12S).

Description

Paper currency handling device and correction method

[0001] The present disclosure relates to a paper currency handling device and a correction method.

[0002] A paper currency handling device used in an automated teller machine (ATM: Automatic Teller Machine) or the like has a paper currency discrimination unit for discriminating paper currency. The paper currency discrimination unit has, for example, a conveyance mechanism for conveying paper currency and a plurality of types of sensor modules used for discriminating the paper currency conveyed by the conveyance mechanism. As one of the sensor modules used for discriminating paper currency, for example, an image sensor module for taking an image of the paper currency can be mentioned.

[0003] Japanese Patent Application Laid-Open No. 2011-066712, Japanese Patent Application Laid-Open No. 2009-181397

[0004] Since there are individual variations in the sensitivity of the line sensors included in the image sensor module, different pixel values may be output from each line sensor for the same paper currency. Also, due to the influence of the assembly accuracy of the paper currency discrimination unit or the like, the distance between the conveyed paper currency and the line sensor may differ among a plurality of paper currency discrimination units. Therefore, variations in discrimination accuracy may occur among a plurality of paper currency discrimination units, and the discrimination results for the same paper currency may differ.

[0005] Therefore, the present disclosure proposes a technique capable of improving the discrimination accuracy of paper currency.

[0006] The paper currency handling device of the present disclosure has an image sensor module and a control board. The image sensor module has a line sensor for taking a paper currency image that is an image of paper currency. The control board corrects the paper currency pixel value, which is the pixel value of the pixel forming the paper currency image, using a correction coefficient calculated from a medium pixel value, which is the pixel value of the pixel forming a medium image that is an image taken by the line sensor of a predetermined medium, and a reference value.

[0007] According to the present disclosure, the discrimination accuracy of paper currency can be improved.

[0008] Figures showing an example configuration of the banknote handling device of Embodiment 1 of this disclosure. Flowchart showing an example of the processing procedure in the manufacturing phase of the banknote handling device of Embodiment 1 of this disclosure. Flowchart showing an example of the processing procedure in the operation phase of the banknote handling device of Embodiment 1 of this disclosure. Figures showing an example of operation2 of this disclosure. Figures showing an example of operation of the banknote handling device of Embodiment 2 of this disclosure. Figures showing an example of operation of the banknote handling device of Embodiment 2 of this disclosure.

[0009] The embodiments of this disclosure will be described below with reference to the drawings.

[0010] [Example 1] <Configuration of the banknote handling device> Figure 1 is a diagram showing an example configuration of the banknote handling device according to Example 1 of the present disclosure. In Figure 1, the banknote handling device 1 has an image sensor module 10 and a control board 20. The image sensor module 10 and the control board 20 are mounted on a banknote authentication unit (not shown) of the banknote handling device 1.

[0011] The image sensor module 10 includes an upper line sensor 11S, an upper memory 11M, a lower line sensor 12S, and a lower memory 12M. The upper line sensor 11S captures an image of the top surface of a banknote being transported between the upper line sensor 11S and the lower line sensor 12S (hereinafter sometimes referred to as the "banknote top surface image"). The lower line sensor 12S captures an image of the bottom surface of a banknote being transported between the upper line sensor 11S and the lower line sensor 12S (hereinafter sometimes referred to as the "banknote bottom surface image"). The upper line sensor 11S and the lower line sensor 12S are connected to the control board 20. The upper memory 11M is connected to the upper line sensor 11S, and the lower memory 12M is connected to the lower line sensor 12S. Hereinafter, the banknote top surface image and the banknote bottom surface image may be collectively referred to as the "banknote image". In the following, the upper line sensor 11S and the lower line sensor 12S may be collectively referred to as "line sensor S". Also, in the following, the upper memory 11M and the lower memory 12M may be collectively referred to as "sensor memory M". An example of line sensor S is a CIS (Contact Image Sensor). An example of sensor memory M is a non-volatile memory such as ROM (Read Only Memory) or flash memory.

[0012] The control board 20 includes a processor 21, a memory 22, and an FPGA (Field-Programmable Gate Array) 23. Examples of the processor 21 include a CPU (Central Processing Unit), a DSP (Digital Signal Processor), and an LSI (Large Scale Integrated Circuit). Examples of the memory 22 include RAM (Random Access Memory), ROM (Read Only Memory), and flash memory.

[0013] <Processing Procedures in the Banknote Handling Device> The processing procedures in the banknote handling device 1 will be explained below, divided into the processing procedures in the manufacturing phase of the banknote handling device 1 and the processing procedures in the operation phase of the banknote handling device 1.

[0014] <Processing Procedure in the Manufacturing Phase of a Banknote Handling Device> Figure 2 is a flowchart showing an example of the processing procedure in the manufacturing phase of a banknote handling device according to Embodiment 1 of this disclosure.

[0015] In Figure 2, in step S100, the control board 20 is connected to the image sensor module 10 by the operator.

[0016] Once the control board 20 is connected to the image sensor module 10, in step S105, the processor 21 resets the counter n to "0".

[0017] Next, in step S110, the processor 21 increments counter n by "1".

[0018] Next, in step S115, as a predetermined medium is transported between the upper line sensor 11S and the lower line sensor 12S, an image of the upper surface of the predetermined medium (hereinafter sometimes referred to as the "medium upper surface image") is captured by the upper line sensor 11S, and an image of the lower surface of the predetermined medium (hereinafter sometimes referred to as the "medium lower surface image") is captured by the lower line sensor 12S. An example of the predetermined medium transported between the upper line sensor 11S and the lower line sensor 12S in step S115 is a dedicated white medium used for adjusting the line sensor S (hereinafter sometimes referred to as the "dedicated white medium"). Hereinafter, the medium upper surface image and the medium lower surface image may be collectively referred to as the "medium image".

[0019] The upper line sensor 11S outputs the pixel values ​​of each of the multiple pixels that form the image on the top surface of the medium (hereinafter sometimes referred to as "upper surface pixel values") to the control board 20, and the lower line sensor 12S outputs the pixel values ​​of each of the multiple pixels that form the image on the bottom surface of the medium (hereinafter sometimes referred to as "bottom surface pixel values") to the control board 20. Hereinafter, the upper surface pixel values ​​and bottom surface pixel values ​​may be collectively referred to as "medium pixel values".

[0020] Next, in step S120, the processor 21 calculates the sum of multiple media top pixel values ​​in the media top image (hereinafter sometimes referred to as the "top pixel sum") and the sum of multiple media bottom pixel values ​​in the media bottom image (hereinafter sometimes referred to as the "bottom pixel sum"). Hereinafter, the top pixel sum and the bottom pixel sum may be collectively referred to as the "media pixel sum". Multiple pixels in the media image that are subject to the calculation of the media pixel sum (hereinafter sometimes referred to as the "pixels subject to sum") are predetermined. As pixels subject to sum, for example, all pixels that make up the media image, pixels in a predetermined range in the media image, pixels at a specific position among the pixels that make up the media image, etc. can be used.

[0021] Next, in step S125, the processor 21 updates the media pixel sum value. For example, the processor 21 sequentially updates the media pixel sum value by calculating the average value obtained by dividing the sum of the n media pixel sum values ​​calculated n times by n.

[0022] Next, in step S130, the processor 21 determines whether the value of counter n has reached a predetermined value N. If the value of counter n has reached the predetermined value N (step S130: Yes), the process proceeds to step S135; if the value of counter n has not reached the predetermined value N (step S130: No), the process returns to step S110.

[0023] In step S135, the processor 21 stores the top pixel sum value after N updates (hereinafter sometimes referred to as the "final top sum value") in the upper memory 11M, and stores the bottom pixel sum value after N updates (hereinafter sometimes referred to as the "final bottom sum value") in the lower memory 12M. Hereinafter, the final top sum value and the final bottom sum value may be collectively referred to as the "final sum value".

[0024] Next, in step S140, the processor 21 stores a predetermined reference value RV in the upper memory 11M and the lower memory 12M, respectively. After the processing in step S140, the processing procedure shown in Figure 2 is completed.

[0025] The above describes an example of the processing procedure in the manufacturing phase of the banknote handling device 1.

[0026] <Processing Procedure in the Operation Phase of the Banknote Handling Device> Figure 3 is a flowchart showing an example of the processing procedure in the operation phase of the banknote handling device according to Embodiment 1 of this disclosure. In the processing procedure shown in Figure 3, after the processing of steps S200 to S215, which is the processing procedure when the banknote handling device 1 is started up, the processing of steps S220 to S230, which is the processing procedure when banknotes are authenticated in the banknote handling device 1, is executed. Furthermore, the processing of step S220 is executed only once at the start of authentication in each transaction, while the processing of steps S225 to S230 is executed repeatedly each time an authentication is performed on each banknote that is sequentially transported within the banknote handling device 1 in each transaction.

[0027] In Figure 3, in step S200, the processor 21 reads the final top-level sum from the upper memory 11M and the final bottom-level sum from the lower memory 12M.

[0028] Next, in step S205, the processor 21 reads a reference value RV from the upper memory 11M and the lower memory 12M, respectively.

[0029] Next, in step S210, the processor 21 calculates an upper surface correction coefficient according to formula (1) based on the final upper surface summation value and the reference value RV, and calculates a lower surface correction coefficient according to formula (2) based on the final lower surface summation value and the reference value RV. Hereinafter, the upper surface correction coefficient and the lower surface correction coefficient may be collectively referred to as "correction coefficients". • Upper surface correction coefficient = Reference value RV / Final upper surface summation value …(1) • Lower surface correction coefficient = Reference value RV / Final lower surface summation value …(2)

[0030] Next, in step S215, the processor 21 stores the top surface correction coefficient and the bottom surface correction coefficient in the memory 22.

[0031] After the startup processing procedure for the banknote handling device 1 is completed, in step S220, the processor 21 reads the top surface correction coefficient and the bottom surface correction coefficient from the memory 22 and sets the read top surface correction coefficient and bottom surface correction coefficient into the registers of the FPGA 23.

[0032] Next, in step S225, as the banknote is transported between the upper line sensor 11S and the lower line sensor 12S, the upper line sensor 11S captures an image of the top surface of the banknote, and the lower line sensor 12S captures an image of the bottom surface of the banknote. The upper line sensor 11S also outputs the pixel values ​​of each of the multiple pixels that form the top surface image of the banknote (hereinafter sometimes referred to as "top surface pixel values ​​of the banknote") to the control board 20, and the lower line sensor 12S also outputs the pixel values ​​of each of the multiple pixels that form the bottom surface image of the banknote (hereinafter sometimes referred to as "bottom surface pixel values ​​of the banknote") to the control board 20.

[0033] Next, in step S230, FPGA 23 corrects the top pixel values ​​of the banknote by multiplying each of the top pixel values ​​by a top correction coefficient, and corrects the bottom pixel values ​​of the banknote by multiplying each of the bottom pixel values ​​by a bottom correction coefficient. Hereinafter, the top pixel values ​​and bottom pixel values ​​of the banknote may be collectively referred to as "banknote pixel values". FPGA 23 outputs the corrected banknote pixel values ​​to processor 21.

[0034] Next, in step S235, the processor 21 performs identification of the top surface of the banknote using the corrected top surface pixel values, and also performs identification of the bottom surface of the banknote using the corrected bottom surface pixel values.

[0035] The above describes an example of the processing procedure during the operation phase of the banknote handling device 1.

[0036] Furthermore, if the line sensor S has multiple wavelengths, it is preferable that the processes in steps S105 to S140 shown in Figure 2 and the processes in steps S200 to S230 shown in Figure 3 be performed for each of the multiple wavelengths of the line sensor S.

[0037] <Operation of the Banknote Handling Device> Figures 4, 5, 6, and 7 show examples of the operation of the banknote handling device of Embodiment 1 of this disclosure. The operation of the banknote handling device 1 will be described below with reference to Operation Example 1 and Operation Example 2. Operation Example 1 and Operation Example 2 are two different operation examples of banknote handling devices 1, each having a different image sensor module 10.

[0038] Furthermore, the following explanation will take the case where the pixels subject to addition are all pixels that form the media image as an example. Note that, as described above, the same processing is performed on the top and bottom surfaces of a given medium, and the same processing is performed on the top and bottom surfaces of a banknote, so the following explanation applies to both the top and bottom surfaces.

[0039] <Operation Example 1 (Figures 4 and 5)> In Figures 4 and 5, when the first final summation value PA1 in the first dedicated white medium MA is calculated to be "9500", and the first reference value RVA is "10000", the first correction coefficient CF1 is calculated to be "1.05".

[0040] Here, the first reference value RVA is set to the midpoint between the first lower threshold THLA and the first upper threshold THUA. Therefore, here, the first reference value RVA is set to "10000", which is the midpoint between the first lower threshold THLA, which is "9000", and the first upper threshold THUA, which is "11000". The range from the first lower threshold THLA, which is "9000", to the first upper threshold THUA, which is "11000", corresponds to the acceptable range of the final added value when the final added value is calculated using the first dedicated white medium MA.

[0041] Since the first correction coefficient CF1 is calculated to be "1.05", when authenticating banknotes, all banknote pixel values ​​are corrected by multiplying each of the banknote pixel values ​​by the first correction coefficient CF1 of "1.05".

[0042] <Operation Example 2 (Figures 6 and 7)> In Figures 6 and 7, when the second final summation value PA2 in the first dedicated white medium MA is calculated to be "10500", and the first reference value RVA is "10000", the second correction coefficient CF2 is calculated to be "0.95".

[0043] Since the second correction coefficient CF2 is calculated to be "0.95", when authenticating banknotes, all banknote pixel values ​​are corrected by multiplying each of the banknote pixel values ​​by the second correction coefficient CF2 of "0.95".

[0044] The first example has been described above.

[0045] [Embodiment 2] <Operation of the Banknote Handling Device> It is realistically difficult to manufacture dedicated white media with exactly the same color tone, density, and characteristics such as transmittance and reflectance. Also, it takes time and cost to strictly select dedicated white media with exactly the same color tone and density from a large number of dedicated white media. On the other hand, if they are multiple dedicated white media of the same lot, there is a high possibility that their color tone, density, and characteristics are almost the same. Therefore, the banknote handling device 1 of Embodiment 2 operates as follows.

[0046] FIGS. 8, 9, and 10 are diagrams showing operation examples of the banknote handling device of Embodiment 2 of the present disclosure.

[0047] As shown in FIG. 8, the processor 21 calculates the third final addition value PA3 as "10000" from the second dedicated white medium MB. Also, as shown in FIG. 10, the range from the second lower threshold value THLB of "9000" to the second upper threshold value THUB of "11000" corresponds to the pass range of the final addition value when the final addition value is calculated using the second dedicated white medium MB. Therefore, the second reference value RVB shown in FIG. 10 is set to "10000", which is the intermediate value between the second lower threshold value THLB and the second upper threshold value THUB.

[0048] On the other hand, when the dedicated white medium from which the media pixel value is acquired is changed from the second dedicated white medium MB of the first lot L1 to the third dedicated white medium MC of the second lot L2, the processor 21, as shown in FIG. 9, calculates the fourth final addition value PA4 as "11000" from the third dedicated white medium MC. Note that the change of the lot of the dedicated white medium is instructed to the banknote handling device 1 from an operator using, for example, a touch panel (not shown) that the banknote handling device 1 has.

[0049] Next, the processor 21 calculates, according to Equation (3), the difference between the final addition values between lots (hereinafter sometimes referred to as "lot difference") caused by the difference in the lots of the dedicated white media. Therefore, here, the lot difference is calculated as "+1000". ・Lot difference = Fourth final addition value PA4 - Third final addition value PA3... (3)

[0050] Therefore, as shown in Figure 10, the processor 21 determines the acceptable range for the final added value when the final added value is calculated using the third dedicated white medium MC as the acceptable range for the final added value when the final added value is calculated using the second dedicated white medium MB, which is shifted by +1000 from the acceptable range for the final added value when the final added value is calculated using the second dedicated white medium MB. Thus, as shown in Figure 10, the range from the third lower threshold THLC of "10000" to the third upper threshold THUC of "12000" is determined as the acceptable range for the final added value when the final added value is calculated using the third dedicated white medium MC. Therefore, the third reference value RVC shown in Figure 10 is set to "11000", which is the midpoint between the third lower threshold THLC and the third upper threshold THUC. In this way, when the dedicated white medium from which the media pixel values ​​are obtained is changed from the second dedicated white medium MB of the first lot L1 to the third dedicated white medium MC of the second lot L2, the processor 21 changes the reference value RV from the second reference value RVB corresponding to the second dedicated white medium MB to the third reference value RVC corresponding to the third dedicated white medium MC.

[0051] The above describes Example 2.

[0052] [Example 3] <Operation of the banknote handling device> If the processor 21 detects that there is a media pixel value among the multiple media pixel values ​​output from the line sensor S that is below the lower threshold LV, it may determine that there is dirt on the dedicated white media and, without calculating the media pixel sum value, output a warning message prompting the operator to replace the dedicated white media with a new dedicated white media.

[0053] Furthermore, if the processor 21 detects that there is a media pixel value among the multiple media pixel values ​​output from the line sensor S that is greater than the upper limit value UV, it may determine that electrostatic noise is occurring in the dedicated white medium and, without calculating the media pixel sum value, output a warning message prompting the operator to replace the dedicated white medium with a new dedicated white medium.

[0054] Furthermore, if the final added value does not fall within the acceptable range set by the upper and lower thresholds, the processor 21 may determine that there is an abnormality in the line sensor S and, instead of using the final added value outside the acceptable range, output a warning message prompting the operator to replace the line sensor S with a new one.

[0055] For example, the processor 21 may cause the banknote handling device 1 to output the above warning message to a touch panel (not shown).

[0056] The above describes Example 3.

[0057] As described above, the banknote handling device of this disclosure (banknote handling device 1 of the embodiment) comprises an image sensor module (image sensor module 10 of the embodiment) and a control board (control board 20 of the embodiment). The image sensor module has line sensors (upper line sensor 11S and lower line sensor 12S of the embodiment) that capture a banknote image, which is an image of a banknote. The control board corrects the banknote pixel values, which are the pixel values ​​of the pixels that form the banknote image, using a correction coefficient calculated from a reference value and a medium pixel value, which is the pixel value of the pixels that form the medium image, which is an image of a predetermined medium when captured by the line sensors.

[0058] This reduces the impact of variations in line sensor sensitivity and the assembly accuracy of the banknote authentication unit on which the image sensor module is mounted, thereby reducing variations in authentication accuracy among multiple banknote authentication units and improving banknote authentication accuracy.

[0059] 1. Banknote handling device 10. Image sensor module 11S. Upper line sensor 11M. Upper memory 12S. Lower line sensor 12M. Lower memory 20. Control board 21. Processor 22. Memory 23. FPGA

Claims

1. A banknote handling device comprising: an image sensor module having a line sensor for capturing a banknote image, which is an image of a banknote; and a control board, wherein the control board corrects the banknote pixel values, which are the pixel values ​​of pixels forming the banknote image, using a correction coefficient calculated from a reference value and a medium pixel value, which is the pixel value of pixels forming the banknote image, which is an image of a medium when a predetermined medium is captured by the line sensor.

2. The banknote handling device according to claim 1, wherein the control board has a processor, and the processor calculates the correction coefficient from the sum of the multiple media pixel values ​​and the reference value.

3. The banknote handling device according to claim 2, wherein the image sensor module further has a first memory for storing the added value and the reference value, the control board further has a second memory, and the processor reads the added value and the reference value from the first memory when the banknote handling device is started, calculates the correction coefficient from the added value and the reference value read from the first memory, and stores the calculated correction coefficient in the second memory.

4. The control board further comprises an FPGA, and when the banknote handling device authenticates the banknote, the processor reads the correction coefficient from the second memory, sets the correction coefficient read from the second memory in the FPGA, and the FPGA corrects the banknote pixel value using the correction coefficient set by the processor, as described in claim 3.

5. The banknote handling device according to claim 2, wherein the processor calculates the correction coefficient for each of the multiple wavelengths when the line sensor has multiple wavelengths.

6. The banknote handling device according to claim 1, wherein the control board calculates the correction coefficient in the manufacturing phase of the banknote handling device using a dedicated white medium as the predetermined medium.

7. The banknote handling device according to claim 1, wherein the control board has a processor, the predetermined medium includes a first medium and a second medium of different lots, and the processor changes the reference value from a first reference value which corresponds to the first medium to a second reference value which corresponds to the second medium when the source of acquiring the medium pixel value is changed from the first medium to the second medium.

8. The banknote handling device according to claim 2, wherein the processor outputs a warning when there is a media pixel value among the plurality of media pixel values ​​that is less than a lower limit.

9. The banknote handling device according to claim 2, wherein the processor outputs a warning when there is a media pixel value among the plurality of media pixel values ​​that is greater than an upper limit value.

10. The banknote handling device according to claim 2, wherein the processor outputs a warning if the added value does not fall between the upper threshold and the lower threshold.

11. A banknote handling device comprising: an image sensor module having a line sensor for capturing a banknote image, which is an image of a banknote; and a control board, wherein the control board corrects the banknote pixel values, which are the pixel values ​​of pixels forming the banknote image, using a correction coefficient calculated from a reference value and a medium pixel value, which is the pixel value of pixels forming the banknote image, when a predetermined medium is captured by the line sensor.

Images