EP3951699A1

EP3951699A1 - Paper sheet sorting device, white reference data adjustment method, and program

Info

Publication number: EP3951699A1
Application number: EP20784402.8A
Authority: EP
Inventors: Toru Seki; Kenta YUJI
Original assignee: Japan Cash Machine Co Ltd
Current assignee: Japan Cash Machine Co Ltd
Priority date: 2019-03-29
Filing date: 2020-03-03
Publication date: 2022-02-09
Also published as: EP3951699A4; WO2020202985A1

Abstract

To appropriately perform shading correction of an image of a paper sheet. The present invention includes a plurality of light receiving units that are placed at locations facing the other surface of a paper sheet being moved to receive detection light having passed through the paper sheet and that are arranged along a width direction orthogonal to a moving direction of the paper sheet, an image data generating unit that generates image data of a tone corresponding to the detection light having entered the light receiving units, and a white-reference-data storing unit that stores therein white reference data obtained by irradiating a white reference member to be used for shading correction of the image data with the detection light. The above configuration includes a movement restricting unit that is located at end parts in a width direction of a transport path and that forms side walls of the transport path, and an adjusting unit that adjusts the white reference data for shading correction of image data of the light receiving units to a predetermined value on a condition that locations in the width direction of the light receiving units are in a predetermined specific region.

Description

Field

The present invention relates to a paper sheet sorting device, a white-reference-data adjustment method, and a program.

Background

Paper sheet sorting devices that receive a paper sheet inserted from an insertion port into a transport path and perform sorting processing by the authenticity, the denomination, or the like of the paper sheet are conventionally known. In these paper sheet sorting devices, an image of a paper sheet transported on the transport path is taken by an image sensor to generate image data indicating the paper sheet. This image data is used for authentication of the paper sheet and the like.
For example, Patent Literature 1 discloses a paper sheet sorting device that includes an image sensor including a light emitting unit that irradiates one surface of a paper sheet on a transport path with detection light, and a light receiving unit on which the detection light having passed through the paper sheet to the other surface side is incident. In such a paper sheet sorting device, a technique of performing shading correction of an image generated by the image sensor is adopted in some cases.

Citation List

Patent Literature

Patent Literature 1: Japanese Patent Application Laid-open No. 2009-295125

Summary

Technical Problem

White reference data used for shading correction is generated by a predetermined method. However, it is difficult to appropriately generate white reference data for correcting image data of pixels located in a predetermined specific region (for example, near side walls of the transport path) depending on methods for generating white reference data (for example, depending on shapes of a white reference member). In view of these circumstances, the present invention has an object to enable adjustment of white reference data of pixels located in the specific region described above.

Solution to Problem

In order to achieve the above object, a paper sheet sorting device according to the present invention includes a transport unit that moves a paper sheet along a transport path, a light emitting unit that irradiates one surface of the paper sheet being moved with detection light, a plurality of light receiving units that are placed at locations facing the other surface of the paper sheet being moved to receive the detection light having passed through the paper sheet and that are arranged along a width direction orthogonal to a moving direction of the paper sheet, an image data generating unit that generates image data of tones corresponding to the detection light having entered the light receiving units, a white-reference-data storing unit that stores therein white reference data obtained by irradiating a white reference member to be used for shading correction of the image data with the detection light, a movement restricting unit that is located at end parts in a width direction of the transport path and that forms side walls of the transport path, and an adjusting unit that adjusts the white reference data for shading correction of image data of the light receiving units to a predetermined value on a condition that locations in the width direction of the light receiving units are in a predetermined specific region.

Advantageous Effects of Invention

According to the present invention, shading correction of an image of a paper sheet is appropriately performed.

Brief Description of Drawings

[FIGS. 1] FIGS. 1(a) and (b) are respectively an external perspective view and a sectional view of a banknote sorting device according to a first embodiment.
[FIGS. 2] FIGS. 2(a) and (b) are respectively an exploded perspective view and an enlarged sectional view of the banknote sorting device according to the first embodiment.
[FIGS. 3] FIGS. 3(a) to (c) are explanatory diagrams of a white reference member according to the first embodiment.
[FIG. 4] FIG. 4 is a functional block diagram of the banknote sorting device according to the first embodiment.
[FIGS. 5] FIGS. 5(a) to (c) are explanatory diagrams of unadjusted white reference data according to the first embodiment.
[FIGS. 6] FIGS. 6(a) to (c) are explanatory diagrams of an adjustment method of white reference data according to the first embodiment.
[FIG. 7] FIG. 7 is a flowchart of pre-shipment adjustment processing according to the first embodiment.
[FIGS. 8] FIGS. 8(a) to (d) are explanatory diagrams of an adjustment method of white reference data according to a second embodiment.
[FIGS. 9] FIGS. 9(a) to (c) are explanatory diagrams of white reference data according to a third embodiment.

Description of Embodiments

FIG. 1(a) is an external perspective view of a banknote sorting device H according to a first embodiment. FIG. 1(b) is a sectional view along a line A-A in FIG. 1(a). This banknote sorting device H includes a control device that includes a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), a flash memory, and the like.
The banknote sorting device H is configured to include a lower unit 1 and an upper unit 2 as illustrated in FIG. 1(a), and a paper sheet (for example, a banknote) is inserted from an insertion port 3a to a transport path 4 formed between an upper surface of the lower unit 1 and a lower surface of the upper unit 2. The banknote inserted into the insertion port 3a moves in the transport path 4 in a moving direction indicated by arrows in FIG. 1(b).
As illustrated in FIG. 1(b), a transport device 5 is provided in the banknote sorting device H. The transport device 5 is configured to include a plurality of lower transport members 5a including transport drums (rollers), a transport belt, and the like placed on a lower side of the transport path 4, and upper transport members 5b including transport drums and the like placed on an upper side of the transport path 4 and placed to face the lower transport members 5a, respectively. The lower transport members 5a according to the present example are transport drums that are rotationally driven by a motor while being in contact with a lower surface of a banknote at the outer peripheries, transport belts, and the like. The upper transport members 5b rotate together with an associated lower transport members 5a with the banknote interposed therebetween while being in contact with the upper surface of the banknote at the outer peripheries. The banknote inserted to the insertion port 3a is positioned between the lower transport members 5a and the associated upper transport members 5b to be nipped on the both surfaces, and moves on the transport path 4 at a constant speed along with rotation of the lower transport members 5a. The transport device 5 for transporting a banknote is not limited to the example described above.
As illustrated in FIG. 1(b), the banknote sorting device H includes sensors 6 (6a, 6b, and 6c). The control device of the banknote sorting device H performs authentication and the like of a banknote inserted to the insertion port 3a according to information acquired from the sensors 6. Specifically, the sensors 6 are configured to include two image sensors 6 (6a and 6b) and one magnetic sensor 6c. However, a configuration in which a banknote is sorted using components other than these sensors may be adopted.
The first image sensor 6a is a CIS (Contact Image Sensor). The second image sensor 6b is a CIS similarly to the first image sensor 6a. Each of these image sensors includes a light emitting unit (for example, light emitting elements Ga described later) that irradiates a banknote on the transport path 4 with detection light, and a light receiving unit (for example, pixels Gb described later) on which the detection light is incident. The light emitting unit of each of the image sensors includes two kinds of light sources, that is, a reflective light source and a transmissive light source. Each of the light sources emits plural types of detection light having different wavelengths. The "detection light" in the present invention can include invisible light (such as infrared rays) as well as visible light (wavelength = 360 to 760 nanometers).
For example, the detection light emitted from a reflective light source of the first image sensor 6a is reflected from one surface of the banknote on the transport path 4 and enters the light receiving unit of the first image sensor 6a. The detection light emitted from a reflective light source of the second image sensor 6b is reflected from the other surface of the banknote on the transport path 4 and enters the light receiving unit of the second image sensor 6b.
The detection light emitted from a transmissive light source of the first image sensor 6a passes through the banknote on the transport path 4 from a one surface side to the other surface side and enters the light receiving unit of the second image sensor 6b. The detection light emitted from a transmissive light source of the second image sensor 6b passes through the banknote on the transport path 4 from the other surface side to the one surface side and enters the light receiving unit of the first image sensor 6a.
The light receiving unit of each of the image sensors (6a and 6b) is configured by a plurality of pixels. A detection signal corresponding to the detection light (transmitted light or reflected light) having entered each of these pixels is input to the control device described above. The CPU of the control device executes a program stored in the ROM to generate image data indicating the banknote based on the detection signals.
Black correction processing to remove dark current components from the image data is performed. Shading correction is performed to the image data after the black correction processing is performed, which is explained in detail later. The shading correction is a method to reduce luminance unevenness included in the image data and is, for example, a method of dividing the tone of each pixel in image data by the tone of an associated pixel of corresponding white reference data to correct the tone of each pixel in the image data. The white reference data used in the shading correction is, for example, generated in a process of manufacturing the banknote sorting device H and is stored in the flash memory of the control device.
The magnetic sensor 6c detects iron included in ink on a print surface of the banknote. The CPU sorts the banknote according to the detection signals from these sensors 6. For example, when a banknote is determined to be illegitimate, the banknote is ejected by the transport device 5 from the insertion port 3a. On the other hand, when a banknote is determined to be legitimate, the banknote is moved to a loading port 3b. While detailed explanations are omitted, the banknote sorting device H is attached to, for example, a paper sheet processing device described in Japanese Patent Application Laid-open No. 2009-295125 . The banknote having moved to the loading port 3b of the banknote sorting device H is subsequently accommodated in an accommodating unit of the paper sheet processing device.
FIG. 2(a) is an exploded perspective view of the banknote sorting device H. FIG. 2(b) is a sectional view along a line B-B in FIG. 1(a) described above. As illustrated in FIG. 2(a), the banknote sorting device H includes a lower cover 7, an upper tray 8, and reflecting white reference pieces 9 (9a and 9b) in addition to the lower unit 1, the upper unit 2, the first image sensor 6a, and the second image sensor 6b described above. Components of the banknote sorting device H other than the first image sensor 6a, the second image sensor 6b, the lower cover 7, the upper tray 8, and the reflecting white reference pieces 9 (9a and 9b) are omitted in FIG. 2(b).
As illustrated in FIG. 2(a), the upper tray 8 is formed in a substantially box shape having a bottomed concave portion 8a on an upper surface thereof and the first image sensor 6a is housed in the concave portion 8a of the upper tray 8 as illustrated in FIG. 2(b). A lower surface 8b of the upper tray 8 forms an upper surface (a ceiling surface) of the transport path 4 as illustrated in FIG. 2(b). The upper tray 8 described above is formed of a light transmissive material, for example, transparent resin.
FIG. 2(b) illustrates a plurality of light emitting elements Ga that are the transmissive light sources of the first image sensor 6a. In FIG. 2(b), some of all the light emitting elements Ga are selectively illustrated. These light emitting elements Ga are arranged to extend along an X-axis direction (width direction) orthogonal to a Y-axis direction in the horizontal direction assuming that a direction in which a banknote on the transport path 4 moves is the Y-axis direction (transport direction). In the present embodiment, 720 light emitting elements Ga are arrayed in the X-axis direction.
As illustrated in FIG. 2(b), the first image sensor 6a is housed in the concave portion 8a of the upper tray 8 so as to locate the light emitting elements Ga on the side of the lower surface 8b described above. As described above, the upper tray 8 is light transmissive. Therefore, the detection light emitted from the light emitting elements Ga can pass through the upper tray 8 and enter the transport path 4.
The second image sensor 6b in the present embodiment includes a plurality of pixels Gb on which the detection light emitted from the first image sensor 6a is incident as illustrated in FIG. 2(b). In FIG. 2(b), some of all the pixels Gb are selectively illustrated. As many (720) pixels Gb as the light emitting elements Ga of the first image sensor 6a are provided in the second image sensor 6b described above. The pixels Gb are arrayed in the X-axis direction similarly to the light emitting elements Ga, and one light emitting element Ga and one pixel Gb face in such a manner that optical axes L thereof (see FIG. 2(b)) align with each other.
As illustrated in FIG. 2(a), the lower cover 7 includes a right wall surface 7a, a left wall surface 7b, recesses 7c that are formed by cutting out a back end edge in the transport direction of the lower cover and that penetrate in an upper-lower direction through the lower cover, and an upper surface 7d located at a place other than the recesses 7c. The lower cover 7 is provided to cover the upper surface (on the side of the pixels Gb) of the second image sensor 6b as illustrated in FIG. 2(b). The upper surface 7d of the lower cover 7 constitutes the bottom surface of the transport path 4. A height (a distance h illustrated in FIG. 2(b)) from the bottom surface (the lower cover 7) of the transport path 4 to the ceiling surface (the upper tray 8) in the present embodiment is about 2 millimeters.
When the lower cover 7 is installed on the lower unit 1, the lower transport members 5a such as the transport belt of the transport device 5 protrude through the recesses 7c toward the transport path 4. A banknote on the transport path 4 moves in the transport direction with rotation of the transport belt.
The lower cover 7 is formed of a light transmissive material such as transparent resin similarly to the upper tray 8 described above. Therefore, the detection light emitted from the light emitting elements Ga of the first image sensor 6a and having passed through a banknote on the transport path 4 can pass through the lower cover 7 (the upper surface 7d) to be incident on the pixels Gb of the second image sensor 6b.
As illustrated in FIG. 2(b), the right wall surface 7a and the left wall surface 7b of the lower cover 7 form side walls of the transport path 4. Specifically, the right wall surface 7a forms a side wall on the right side as viewed from a banknote moving on the transport path 4. The left wall surface 7b forms a side wall on the left side as viewed from the banknote moving on the transport path 4. In the present embodiment, a region from the right wall surface 7a to the left wall surface 7b in a region on the X axis (a main scanning axis) is referred to as "region R" for the sake of explanations. This region R is a region where an image for sorting a banknote is taken.
In the present embodiment, the length (the distance from the right wall surface 7a to the left wall surface 7b) of the region R is about 4 millimeters longer than the width in the X-axis direction of a banknote moving on the transport path 4. With this configuration, the width of the transport path 4 is shorter than, for example, in a configuration in which the length of the region R is equal to or more than 4 millimeters longer than the width of a banknote. Therefore, the present embodiment has an advantage in that the banknote sorting device H is easily downscaled.
As described above, the banknote sorting device H according to the present embodiment can perform shading correction of an image. Specifically, the banknote sorting device H can perform both shading correction of an image obtained by reflective light sources and shading correction of an image obtained by transmissive light sources.
The banknote sorting device H generates white reference data to be used for shading correction of an image obtained by reflective light sources, using a reflective light source and a light receiving unit that are located on an outer side of the region R described above. Meanwhile, white reference data to be used for shading correction of an image obtained by transmissive light sources (the light emitting elements Ga) is generated using a transmissive light source and a light receiving unit (the pixels Gb) that are located on an inner side of the region R described above.
The banknote sorting device H includes the reflecting white reference pieces (9a and 9b) to generate white reference data to be used for shading correction of an image obtained by the reflective light sources. As illustrated in FIG. 2(a), two white reflectors T are provided on each of the reflecting white reference pieces 9. The reflecting white reference pieces 9 are provided in such a manner that the relevant reflectors T are located in regions r on outer sides of the region R as illustrated in FIG. 2(b). Respective ends of the reflectors T abut on the lower cover 7 as illustrated in FIG. 2(b). In the present embodiment, the pixels Gb are arrayed in the entire region of the region R and the regions r.
One reflector T out of the two reflectors T is irradiated with light emitted from a reflective light source located in the associated region r among the reflective light sources of the first image sensor 6a. Light emitted from this reflective light source and reflected from the reflector T enters a reflected-light receiving element (a pixel) of the first image sensor 6a. Similarly, the other reflector T out of the two reflectors T is irradiated with light emitted from a reflective light source located in the associated region r among the reflective light sources of the second image sensor 6b. Light emitted from this reflective light source and reflected from the reflector T enters a reflected-light receiving element (a pixel) of the second image sensor 6b. The CPU of the control device generates white reference data for correcting an image obtained by the reflective light sources according to the tone of the reflected light having entered the light receiving element from the reflectors T.
As described above, the white reference data for an image obtained by the reflective light sources is generated using the reflecting white reference pieces 9 (the reflectors T). Meanwhile, the white reference data for an image obtained by the transmissive light sources (the light emitting elements Ga) is generated using a white reference member 10 (see FIGS. 3) described later. A period in which the white reference data for an image obtained by the reflective light sources is generated can be appropriately set. For example, a configuration in which the white reference data is generated immediately after a banknote is inserted into the insertion port 3a (before the banknote is detected by the image sensors) may be adopted.
FIG. 3(a) is a perspective view of the white reference member 10 described above. In the present embodiment, white reference data is generated using the white reference member 10 in a process of manufacturing the banknote sorting device H. However, a configuration in which white reference data is generated at a different moment may be used. When white reference data is to be generated, the white reference member 10 is placed on the transport path 4 (the upper surface 7d of the lower cover 7) (see FIG. 3(b) described later). The white reference member 10 is removed by a worker after the white reference data is generated. That is, the white reference member 10 is not placed on the transport path 4 when the banknote sorting device H is shipped as a finished product.
As illustrated in FIG. 3(a), the white reference member 10 is configured to include a white reference sheet 10a, and a guard member 10b in a rectangular ring shape holding the outer peripheral edge of the white reference sheet 10a. The white reference sheet 10a transmits light (of a type and an intensity) similar to a white part of banknotes when irradiated with detection light of the light emitting elements Ga. The white reference sheet 10a is formed, for example, with a substantially identical thickness to that of a banknote.
The guard member 10b is a substantially plate-like ring member (a frame body) having an outer edge in a substantially rectangular shape, and a penetrated opening part K is provided as illustrated in FIG. 3(a). The white reference sheet 10a is attached (sticked) to the guard member 10b so as to close the opening part K. The guard member 10b is formed of a material that is higher in the strength and larger in the specific gravity (heavier) than the white reference sheet 10a. For example, the guard member 10b is formed of an alloy such as stainless steel.
The white reference member 10 including the guard member 10b has an advantage that an inconvenience of the white reference sheet 10a being bent at the time of placement of the white reference member 10 on the transport path 4 is suppressed as compared to, for example, the white reference member 10 configured only of the white reference sheet 10a. There is also an advantage that an inconvenience of displacement of the white reference member 10 (the white reference sheet 10a) due to wind or the like is suppressed.
FIG. 3(b) is an explanatory diagram of the white reference member 10 placed on the transport path 4. When white reference data is to be generated, a worker temporarily places the white reference member 10 on the transport path 4. Specifically, the white reference member 10 is placed at a position that is between the first image sensor 6a and the second image sensor 6b and that is irradiated with detection light emitted from the light emitting elements Ga of the first image sensor 6a.
As illustrated in FIG. 3(b), the length of the long side of the white reference member 10 is approximate (substantially equal) to the length of the region R. In this case, the short sides of the white reference member 10 (the guard member 10b) are located in the vicinity of (substantially abut on) the side walls (7a and 7b) of the transport path 4. The configuration of the white reference member 10 is not limited to the example described above.
At the time of generating the white reference data, the detection light is emitted from the light emitting elements Ga in a state where the white reference member 10 is placed in the region R of the transport path 4. In this case, the detection light applied to the white reference sheet 10a of the white reference member 10 out of the detection light from the light emitting elements Ga passes through the white reference sheet 10a to be incident on the pixels Gb. Appropriate white reference data is generated from the detection light having entered the pixels Gb.
Meanwhile, the guard member 10b of the white reference member 10 is lower in the optical transmittance than the white part (the white reference sheet 10a) of a banknote. For example, in a case where the guard member 10b is metallic, the guard member 10b does not transmit the detection light. Therefore, when the detection light is emitted from the light emitting elements Ga of the first image sensor 6a in a state where the white reference member 10 is installed in the region R, the detection light applied to the guard member 10b (long-side end parts 10b') does not pass toward the second image sensor 6b and therefore is not received by the pixels Gb. In this case, the white reference data for performing shading correction of image data generated by the relevant pixels Gb is not appropriately generated (see FIG. 5(b) described later).
As is understood from the above explanations, white reference data for correcting an image of the pixels Gb facing the long-side end parts 10b' of the guard member 10b is not appropriately generated when the white reference data is generated using the white reference member 10 in the present embodiment. That is, the white reference data of the pixels Gb located near the side walls of the transport path 4 is not appropriately generated. If the white reference data of the pixels Gb located near the side walls of the transport path 4 is not appropriately generated, a problem arises, for example, in a case where a banknote B moves near the side walls of the transport path 4 as illustrated in FIG. 3(c).
For example, in a case where banknotes are illegally drawn (stolen) from an ATM (Automatic Teller Machine) or the like, a technique of spraying robbery ink to the banknotes is conventionally used (for example, see Japanese Patent Application Laid-open No. 2000-322625 ). From circumstances described above, a configuration that enables to judge whether robbery ink has adhered to an inserted banknote is preferable in the banknote sorting device H. However, there are many cases where robbery ink adheres to a site near an outer edge of a banknote in the width direction. Therefore, if shading correction of an image taken near the side walls of the transport path 4 is not appropriately performed, whether robbery ink has adhered to the banknote cannot be correctly judged in some cases.
In view of these circumstances, the banknote sorting device H of the present embodiment includes a configuration (such as a white-reference-data generating unit 107 illustrated in FIG. 4 described later) in which shading correction of an image near the side walls of the transport path 4 is appropriately performed. According to this banknote sorting device H, an advantage is achieved that robbery ink having adhered to the vicinity of an outer edge of a banknote can be appropriately found while the width of the transport path 4 is relatively shortened.
Specifically, the banknote sorting device H according to the present embodiment adopts a configuration in which the white reference data for correcting an image of pixels Gb facing the guard member 10b is adjustable after generation of the white reference data using the white reference member 10. With this configuration, shading correction of the image of pixels Gb near the side walls of the transport path 4 is appropriately performed even in a case where the banknote B moves near the side walls of the transport path 4. Therefore, the banknote B can be sorted with a clear image.
A method of providing a region where the long-side end parts 10b' of the guard member 10b of the white reference member 10 can be placed on an outer side of the region R (on outer sides of the side walls) and causing the entire region R to face the white reference sheet 10a is also conceivable as another method of appropriately generating the white reference data of pixels Gb near the side walls of the transport path 4. However, this method is likely to have an inconvenience in being difficult to downscale the banknote sorting device H. This inconvenience is suppressed in the present embodiment.
FIG. 4 is a functional block diagram of a banknote sorting device 100 (H) according to the first embodiment. The CPU of the banknote sorting device 100 executes programs to realize various functions. However, the banknote sorting device 100 may be connected to be communicable with an external computer in the manufacturing process of the banknote sorting device 100 and the external computer may realize some of the functions illustrated in FIG. 4 (for example, the white-reference-data generating unit 107 described later). In the case described above, a combination of the banknote sorting device 100 and the external computer can correspond to the "paper sheet sorting device" of the present invention.
As illustrated in FIG. 4, the banknote sorting device 100 includes a transport control unit 101, a sensor unit 102, a sensor control unit 103, an image data generating unit 104, a white-reference-data storage unit 105, a correcting unit 106, and the white-reference-data generating unit 107. The transport control unit 101 controls the transport device 5 described above. For example, when a banknote is inserted into the insertion port 3a, the transport control unit 101 outputs a drive signal for rotating the transport units 5a (such as the transport belt) of the transport device 5 so as to cause the banknote to move in the transport direction.
The sensor unit 102 is configured to include the first image sensor 6a and the second image sensor 6b described above. The sensor control unit 103 controls turning-on of the light emitting elements Ga of the first image sensor 6a. For example, the sensor control unit 103 turns on the light emitting elements Ga in a period (hereinafter, "image taking period") in which a banknote passes the region R described above. The sensor control unit 103 inputs a signal indicating the tone of the detection light having entered each of the pixels Gb (the detection light having passed through the banknote) at each timing in the image taking period, to the image data generating unit 104 described later.
The sensor control unit 103 of the present embodiment turns on the light emitting elements Ga in a period (hereinafter, "generation period") in which the white reference data is generated in the manufacturing process of the banknote sorting device H. The sensor control unit 103 inputs a signal indicating the tone of the detection light having entered each of the pixels Gb (the detection light having passed through the white reference sheet 10a) at each timing in the generation period, to the white-reference-data generating unit 107 described later.
The image data generating unit 104 generates image data indicating an image of the banknote inserted into the insertion port 3a. Specifically, a signal indicating the tone of the detection light having entered each of the pixels Gb (the detection light having passed through the banknote) at each timing in the image taking period when the banknote passes the region R described above is input to the image data generating unit 104. The image data generating unit 104 generates the image data indicating the banknote from these signals.
The white-reference-data storage unit 105 stores therein white reference data to be used for shading correction of the image data generated by the image data generating unit 104. This white reference data is generated by the white-reference-data generating unit 107 described later.
The correcting unit 106 performs shading correction of the image data generated by the image data generating unit 104 using the white reference data described above. Black correction processing of the image data is performed in the banknote sorting device 100 prior to the shading correction. Specifically, the banknote sorting device 100 stores therein dark current components acquired from the pixels Gb in a period in which the light emitting elements Ga of the image sensors (6a and 6b) are off. When the image data is generated, the banknote sorting device 100 subtracts the dark current components described above from the generated image data.
The white-reference-data generating unit 107 is configured to include an adjustment position determining unit 108 and an adjusting unit 109, and generates the white reference data to be stored in the white-reference-data storage unit 105. As described above, the signal indicating the tone of the detection light having entered each of the pixels Gb (the detection light having passed through the white reference sheet 10a) in the generation period in which the white reference data is generated in the manufacturing process of the banknote sorting device 100 is input to the white-reference-data generating unit 107 from the sensor unit 102 described above. The white-reference-data generating unit 107 generates unadjusted white reference data (see FIG. 5(b) described later) from these signals.
The unadjusted white reference data is data that enables to specify the tone of the detection light incident in the generation period for each pixel Gb. The adjustment position determining unit 108 of the white-reference-data generating unit 107 determines a pixel Gb of which the unadjusted white reference data is to be adjusted among the pixels Gb arrayed in the scanning axis direction. While details are described later, the adjustment position determining unit 108 can determine a pixel Gb facing the guard member 10b of the white reference member 10.
The adjusting unit 109 of the white-reference-data generating unit 107 adjusts the unadjusted white reference data of the pixel Gb determined by the adjustment position determining unit 108 to a predetermined value ("average tone Iave" described later). The white-reference-data generating unit 107 stores the white reference data adjusted by the adjusting unit 109 in the white-reference-data storing unit 105.
FIGS. 5(a) and 5(b) are explanatory diagrams of the unadjusted white reference data. FIG. 5(a) is a sectional view along the line B-B illustrated in FIG. 1(a) similarly to FIG. 2(b) described above. However, the first image sensor 6a, the second image sensor 6b, the lower cover 7, and the upper tray 8 among the components of the banknote sorting device 100 are selectively illustrated in FIG. 5(a). Optical paths of the detection light emitted from the light emitting elements Ga are indicated by broken arrows in FIG. 5(a).
In a specific example of FIG. 5(a), a case in which the white reference member 10 (the white reference sheet 10a and the guard member 10b) is placed in the region R of the transport path 4 is assumed. As described above, when the white reference member 10 is placed on the transport path 4, the long-side end parts 10b' of the guard member 10b of the white reference member 10 are located near the side walls (7a and 7b) of the transport path 4.
As illustrated in FIG. 5(a), the detection light applied to the long-side end parts 10b' of the white reference sheet 10a in the detection light emitted from the first image sensor 6a (the light emitting elements Ga) passes through the white reference sheet 10a and enters the second image sensor 6b (the pixels Gb). Meanwhile, the detection light applied to the long-side end parts 10b' of the guard member 10b in the detection light emitted from the first image sensor 6a does not pass through toward the side of the second image sensor 6b. Therefore, the detection light applied to the long-side end parts 10b' of the guard member 10b does not reach the second image sensor 6b.
FIG. 5(b) is an explanatory diagram of the unadjusted white reference data. FIG. 5(b) illustrates the tone I (brightness) of the detection light (transmitted light) having entered the second image sensor 6b at each position on the X axis in the region R (a region between the side walls of the transport path 4). In the present embodiment, the position of the right wall surface 7a of the transport path 4 (a right end of the region R) on the X axis is described as "position Pea" for the sake of explanations. Similarly, the position of the left wall surface 7b of the transport path 4 (a left end of the region R) on the X axis is described as "position Peb".
A region on the X axis where the long-side end part 10b' of the guard member 10b located near the right side wall 7a and the second image sensor 6b face is described as "region Rx". Similarly, a region on the X axis where the guard member 10b located near the left wall surface 7b and the second image sensor 6b face is described as "region Ry".
In a specific example of FIG. 5(b), a case where the guard member 10b abuts on the right wall surface 7a and the left wall surface 7b is assumed. In this specific example, the coordinates of the right end of the region Rx are the position Pea (common to the right wall surface 7a) as illustrated in FIG. 5(b). Similarly, the coordinates of the left end of the region Ry are the position Peb (common to the left wall surface 7b). In the specific example of FIG. 5(b), the coordinates of the left end of the region Rx are a position Px and the coordinates of the left end of the region Ry are a position Py.
A configuration in which the white reference data is generated by placing the white reference sheet 10a in the entire region R (including the region Rx and the region Ry) and irradiating the white reference sheet 10a with the detection light is originally suitable. FIG. 5(c) illustrates white reference data generated by placing the white reference sheet 10a in the entire region R.
However, as is understood from FIGS. 5(b) and 5(c), the tones I in the region Rx and the region Ry where the guard member 10b is located are different from the tones I in a case where the white reference sheet 10a is supposed to be in the region Rx and the region Ry. There may be an inconvenience that appropriate shading correction of image data cannot be performed with the unadjusted white reference data described above.
In view of these circumstances, the tones I in an adjustment region Ra including the region Rx in the region R can be adjusted in the present embodiment. The tones I in an adjustment region Rb including the region Ry in the region R can also be adjusted. A region other than the adjustment region Ra and the adjustment region Rb in the region R is referred to also as "region Rc" in some cases for the sake of explanations. The white reference sheet 10a is located in the entire region Rc (the guard member 10b is not located).
FIGS. 6(a) to 6(c) are explanatory diagrams of a configuration for adjusting the unadjusted white reference data. FIG. 6(a) illustrates the unadjusted white reference data similarly to FIG. 5(b) described above. However, a part of the unadjusted white reference data including the region Rx (near the side wall) and the adjustment region Ra described above is illustrated in the specific example of FIG. 6(a).
As described above, the adjustment region Ra is a region including the region Rx where the guard member 10b is located. Specifically, a region from the position Pea (the right wall surface 7a) to a search start position Psa on the X axis is set as the adjustment region Ra as illustrated in FIG. 6(a). The search start position Psa described above is located on the side of the left end (the position Px) of the region Rx (the guard member 10b) toward the region Rc (the white reference sheet 10a). Similarly, a region from the position Peb (the left wall surface 7b) to a search start position Psb on the X axis is set as the adjustment region Rb. The search start position Psb described above is located on the side of the left end (the position Py) of the region Ry toward the region Rc (the white reference sheet 10a).
As described above, the adjustment region Ra where the unadjusted white reference data is to be adjusted is defined by the search start position Psa. The search start position Psa is determined in advance, for example, in consideration of the thickness of the long-side end parts 10b' in the X-axis direction. The adjustment region Rb where the unadjusted white reference data is to be adjusted is defined by the search start position Psb. The search start position Psb is determined in advance, for example, in consideration of the thickness of the guard member 10b in the X-axis direction. The flash memory of the banknote sorting device 100 stores therein the search start position Psa and the search start position Psb described above.
FIG. 6(a) illustrates the average tone Iave. Among the pixels Gb located in the region Rc, tones I of a pixel Gb located at the search start position Psa and four pixels Gb successively arrayed from this pixel Gb to the side of the region Rc (a total of five pixels Gb) are sampled in the present embodiment. The average value of these five tones I is calculated and the calculation result is stored as the average tone Iave. The calculation method of the average tone Iave is not limited to the example described above. For example, a configuration in which tones I of pixels other than the above five pixels Gb are sampled and the average tone Iave is calculated by these tones I may be adopted.
FIG. 6(a) also illustrates a range (Iave-W≤I≤Iave+W) within deviation thresholds W from the average tone Iave. This range of the tones I is hereinafter referred to also as "normal range" in some cases. The deviation threshold W in the present embodiment is set in advance in such a manner that the tones I of the pixels Gb facing the white reference sheet 10a among the pixels Gb in the adjustment region Ra are in the normal range and the tones of the pixels Gb in the region Rx are out of the normal range.
The banknote sorting device 100 according to the present embodiment specifies positions (pixels Gb) on the X axis where the tones I are out of the normal range. Specifically, whether an absolute value (hereinafter, "deviation dimension") of a difference between the tone I of each pixel and the average tone Iave is larger than the deviation threshold W is sequentially determined in the direction of a white arrow illustrated in FIG. 6(a) (X-axis direction) from the tone I at the search start position Psa. In the specific example of FIG. 6(a), it is first determined that the deviation dimension exceeds the deviation threshold W at the position Px.
FIG. 6(b) is an explanatory diagram of a part of adjusted white reference data. In a specific example of FIG. 6(b), a part of the white reference data near the region Rx is illustrated. Further, a case where the deviation dimension is determined to exceed the deviation threshold W at the position Px is assumed in the specific example of FIG. 6(b) similarly to the specific example of FIG. 6(a) described above. In this case, the tones I of a region (the region Rx) on the side of the position Px toward the position Pea (the right wall surface 7a) are adjusted to the average tone Iave.
FIG. 6(c) is an explanatory diagram of the entire adjusted white reference data. In a specific example of FIG. 6(c), a case in which the deviation dimension is determined to exceed the deviation threshold W at the position Px is assumed similarly to the specific example of FIG. 6(b) described above. In this case, the tones I in a region on the side of the position Px toward the position Pea are adjusted to the average tone Iave. The average tone Iave to which the tones I in the adjustment region Ra are adjusted is hereinafter described as "average tone Iave1".
The banknote sorting device 100 adjusts the tones I in the adjustment region Rb in an identical manner to that in the adjustment region Ra described above. Specifically, the banknote sorting device 100 samples tones I of a pixel Gb located at the search start position Psb and four pixels Gb successively arrayed from this pixel Gb to the side of the region Rc (a total of five pixels Gb). The average value of these five tones I is calculated and the calculation result is stored as an average tone Iave2. A configuration in which the average tone Iave2 is calculated from tones I of pixels other than the above five pixels Gb may be used.
The banknote sorting device 100 subsequently determines whether the absolute value (deviation dimension) of a difference between the tone I of each position (each pixel Gb) and the average tone Iave2 is larger than the deviation threshold W, starting from the search start position Psb toward the position Peb. In the specific example of FIG. 6(c), a case in which it is first determined that the deviation dimension exceeds the deviation threshold W at the position Py is assumed. In this case, the banknote sorting device 100 adjusts the tones I from the position Py to the position Peb in the unadjusted white reference data to the average tone Iave2.
After adjusting the tones I of the adjustment region Ra and the adjustment region Rb, the banknote sorting device 100 stores the white reference data in the white-reference-data storing unit 105. With this white reference data, shading correction of an image of a banknote taken near (the adjustment region Ra and the adjustment region Rb) the side walls (7a and 7b) of the transport path 4 can be appropriately performed.
FIG. 7 is a flowchart of pre-shipment adjustment processing performed by the banknote sorting device 100. This pre-shipment adjustment processing is performed in the manufacturing process of the banknote sorting device 100 (before shipment). The banknote sorting device 100 adjusts the unadjusted white reference data in the pre-shipment adjustment processing and stores the adjusted white reference data in the white-reference-data storing unit 105.
As described above, the banknote sorting device 100 adjusts the tones I of the unadjusted white reference data in both the adjustment region Ra (near the right wall surface 7a) and the adjustment region Rb (near the left wall surface 7b) in the region R through which a banknote passes. The banknote sorting device 100 performs the pre-shipment adjustment processing in which the tones I of the adjustment region Ra out of the adjustment region Ra and the adjustment region Rb are adjusted, and subsequently performs the pre-shipment adjustment processing in which the tones I of the adjustment region Rb are adjusted. In a specific example of FIG. 7, a case in which the tones I of the adjustment region Ra are adjusted in the pre-shipment adjustment processing is assumed.
As illustrated in FIG. 7, when the pre-shipment adjustment processing is started, the banknote sorting device 100 acquires the tones I of five pixels Gb from the unadjusted white reference data (S1). As described above, for example, when the tones I of the pixels Gb in the adjustment region Ra are to be adjusted, the tones I of five pixels Gb arrayed in the left direction (the opposite direction of the X axis) from the search start position Psa that is the end of the left side (the opposite side to the right wall) of the adjustment region Ra (that is, pixels Gb arrayed in the region Rc of FIG. 6(a)) are acquired. The detection light having passed through the white reference sheet 10a is incident on these pixels Gb.
After the tones I of the five pixels Gb are acquired from the unadjusted white reference data, the banknote sorting device 100 calculates the average tone Iave from the tones I of these pixels Gb (S2). The banknote sorting device 100 subsequently designates a target pixel Gb from the pixels Gb (S3). At first Step S3, the pixel Gb located at the search start position Psa described above is designated as the target pixel Gb. At next Step S3, a pixel immediately to the right (located one pixel to the side of the right wall surface 7a) of the pixel Gb designated at first Step S3 is designated as the target pixel Gb.
After designating the target pixel Gb, the banknote sorting device 100 calculates the deviation dimension of the tone I of the target pixel Gb. Specifically, the banknote sorting device 100 subtracts the average tone Iave calculated at Step S2 described above from the tone I of the target pixel Gb and stores the absolute value of the subtraction result as the deviation dimension of the target pixel Gb.
After calculating the deviation dimension of the target pixel Gb, the banknote sorting device 100 determines whether the deviation dimension is larger than the deviation threshold W (S5). When determining that the deviation dimension is smaller than the deviation threshold W (NO: S5), the banknote sorting device 100 returns the processing to Step S3 described above. The banknote sorting device 100 subsequently repeatedly performs Step S4 and Step S5 while changing the target pixel Gb at Step S3. In the above configuration, whether the deviation dimension of each pixel Gb is larger than the deviation threshold W is repeatedly determined while the target pixel Gb is shifted one by one toward the side wall of the transport path 4.
When having determined that the deviation dimension of the target pixel Gb is larger than the deviation threshold W (YES: S5), the banknote sorting device 100 adjusts the white reference data of pixels from this target pixel Gb to the side wall of the transport path 4 to the average tone Iave (S6). For example, when the white reference data of the adjustment region Ra near the right wall surface 7a is to be adjusted, the white reference data of pixels Gb from the target pixel Gb to the right wall surface 7a is adjusted to the average tone Iave. After adjusting the white reference data, the banknote sorting device H ends the pre-shipment adjustment processing.

A second embodiment of the present invention is explained below. In the following embodiments exemplified below, reference signs referred to in the descriptions of the first embodiment are used also for elements with actions and functions identical to those of the first embodiment, and detailed explanations of the respective elements are appropriately omitted.
FIG. 8(a) is a sectional view of the banknote sorting device H according to the second embodiment. FIG. 8(a) corresponds to the sectional view along the line B-B illustrated in FIG. 1(a) in the first embodiment described above. However, FIG. 8(a) selectively illustrates the first image sensor 6a, the second image sensor 6b, the lower cover 7, and the upper tray 8 among the components of the banknote sorting device 100. Optical paths of detection light emitted from the light emitting elements Ga are indicated by broken arrows in FIG. 8(a).
As illustrated in FIG. 8(a), a guide part 8L and a guide part 8R are provided on the upper tray 8. The guide part 8R is substantially plate-like and is erected along the right side wall (7a) of the transport path 4. As illustrated in FIG. 8(a), the lower end of the guide part 8R is located on the side of an abutment face (indicated by S in FIG. 8(a)) of the upper tray 8 abutting on the lower cover 7 toward the bottom surface of the transport path 4. This lower end of the guide part 8R abuts on the vicinity of the right end of a banknote placed on the bottom surface of the transport path 4 and restricts movement of the banknote in an upper direction (Z-axis direction).
The guide part 8L is substantially plate-like similarly to the guide part 8R and is erected along the left side wall (7b) of the transport path 4. As illustrated in FIG. 8(a), the lower end of the guide part 8L is located on the side of the abutment face S of the upper tray 8 on the lower cover 7 toward the bottom surface of the transport path 4 and abuts on the vicinity of the left end of a banknote placed on the bottom surface of the transport path 4 to restrict movement of the banknote in the upper direction.
A contrast example in which the guide parts 8 (R and L) are not provided is assumed. In this contrast example, there is a case in which the left end or the right end of a banknote moves (uplifts) to the abutment face S described above. This case has an inconvenience that the possibility of an end of a banknote entering into a space between the upper tray 8 and the lower cover 7, which causes the banknote to be immovable on the transport path 4 and leads to a jam is not completely eliminated. The guide parts 8 (R and L) in the second embodiment have an advantage that uplift of the right and left ends of a banknote is suppressed as described above and therefore the aforementioned inconvenience is suppressed.
The guide parts 8 (R and L) in the second embodiment are provided in the region R (see FIG. 5(a) described above) where an image to be used for sorting of a banknote is taken. This configuration has an advantage that the banknote sorting device H can be more easily downscaled, for example, as compared to a configuration in which the guide parts 8 (R and L) are provided outside the region R.
However, members located between the light emitting elements Ga of the first image sensor 6a and the pixels Gb of the second image sensor 6b need to transmit light. In view of these circumstances, the guide parts 8 (R and L) are formed of a light transmissive member (for example, transparent resin) similarly to the upper tray 8.
However, in the case where the transparent guide parts 8 (R and L) are provided in the region R, another problem that the detection light is refracted due to the guide parts 8 (R and L) may occur. That is, a problem that the optical path of the detection light is deviated from the original optical path due to the guide parts 8 (R and L) is likely to occur. If the above problem occurs, an inconvenience that the white reference data to be used for shading correction of image data is not appropriately generated may occur.
For example, the detection light emitted from a light emitting element Ga located immediately above the guide part 8R is originally to be incident on a pixel Gb located immediately below the guide part 8R. However, for example, when the detection light emitted from the light emitting element Ga is refracted by the guide part 8R as illustrated in FIG. 8(a), the detection light may be incident on a different pixel Gb (for example, an adjacent pixel Gb) from the pixel Gb on which the detection light is originally to be incident. In this case, the tone I of white reference data (unadjusted white reference data) of the pixel Gb on which the detection light is originally to be incident becomes smaller than the original tone I (in a case where the guide part 8R is not provided). Meanwhile, the tone I of white reference data of the different pixel Gb becomes larger than the original tone I.
FIG. 8(b) is an explanatory diagram of unadjusted white reference data in the second embodiment. FIG. 8(b) illustrates the tone I of detection light incident on the second image sensor 6b at each position on the X axis in the region R (a region between the side walls of the transport path 4).
In the first embodiment described above, the white reference data is generated by irradiating the white reference member 10 (see FIGS. 3) with the detection light. In the second embodiment, the white reference member 10 is not placed on the transport path 4 (the region R) at the time of generating the white reference data. In this second embodiment, the white reference data can be periodically generated in a period after the manufacturing process of the banknote sorting device H (a period in which the banknote sorting device H operates in the market) because the white reference member 10 is not required to be placed on the transport path 4 at the time of generating the white reference data. For example, the white reference data can be generated each time the banknote sorting device H is powered on.
The tone I of detection light emitted from a light emitting element Ga in a case where the detection light passes through white reference paper (for example, the white reference sheet 10a described above) is about one-tenth of that in a case where the detection light does not pass through the white reference paper. In consideration of these circumstances, white reference data where the tone of each pixel Gb is reduced to one-tenth is stored in the white-reference-data storing unit 105 in the second embodiment.
In the second embodiment, as illustrated in FIG. 8(b), the position of the right wall surface 7a of the transport path 4 (the right end of the region R) on the X axis is described as "position Pea" similarly to the first embodiment described above. The position of the left wall surface 7b of the transport path 4 (the left end of the region R) on the X axis is described as "position Peb".
In the second embodiment, uncorrected white reference data of the pixels Gb in the adjustment region Ra near the right side wall of the transport path 4 and the adjustment region Rb near the left side wall thereof is adjusted in an identical manner to that in the first embodiment described above. Specifically, with respect to a pixel Gb where the absolute value (deviation dimension) of a difference between the tone I and the average tone Iave is larger than the deviation threshold W, the tone I of this pixel Gb is adjusted to the average tone Iave.
In a specific example of FIG. 8(b), the deviation dimension of the pixel Gb at the position Px is first larger than the deviation threshold W as viewed in a direction from the center of the region R toward the position Pea (the right side wall). In this specific example, the unadjusted white reference data from the position Px to the position Pea is adjusted to the average tone Iave. Similarly, the deviation dimension of the pixel Gb at the position Py is first larger than the deviation threshold W as viewed in a direction from the center of the region R toward the position Peb (the left side wall). In this specific example, the unadjusted white reference data from the position Py to the position Peb is adjusted to the average tone Iave.
FIG. 8(c) is an explanatory diagram of the adjustment region Ra in the second embodiment. FIG. 8(c) illustrates the unadjusted white reference data similarly to FIG. 8(b) described above. However, in a specific example of FIG. 8(c), a part of the unadjusted white reference data including the adjustment region Ra is illustrated.
As illustrated in FIG. 8(c), the adjustment region Ra is a predetermined region from the position Psa to the position Pea (the right side wall) similarly to the first embodiment described above. The position Psa is previously set in consideration of the shape of the guide part 8R and the refractive index (a range affected by refraction of the detection light) thereof. Similarly, the adjustment region Rb is a predetermined region from the position Psb to the position Peb (the left side wall) as in the first embodiment described above. The position Psb is previously set in consideration of the shape of the guide part 8L and the refractive index thereof.
FIG. 8(d) is a diagram illustrating white reference data where the tones I of the adjustment region Ra are adjusted in the specific example of FIG. 8(c) described above. In a specific example of FIG. 8(d), a case in which the tones I from the position Px to the position Pea are adjusted to the average tone Iave is assumed. The tones I of the adjustment region Rb are adjusted by an identical method to that in the adjustment region Ra. The banknote sorting device H multiplies the tones I of the entire region of the unadjusted white reference data where the tones I of the adjustment region Ra and the adjustment region Rb have been adjusted by one-tenth to be stored as white reference data.
Also in this second embodiment, an identical effect to that in the first embodiment described above is achieved. The locations where the guide parts (8R and 8L) are provided can be appropriately changed. For example, a configuration in which guide parts are provided on the upper surface 7d of the lower cover 7 may be adopted. While provided near the side walls of the transport path 4 in the second embodiment, the guide parts may be provided near the center of the transport path 4. In this case, a region including the guide parts provided near the center is set as an adjustment region and white reference data of pixels Gb located in this adjustment region is adjusted.

In the first embodiment described above, white reference data to be actually used in shading correction is determined from unadjusted white reference data obtained in a state where the white reference member 10 (see FIG. 3(a)) is placed in the region R of the transport path 4. In the second embodiment, white reference data is determined from unadjusted white reference data obtained in a state where the white reference member 10 is not placed in the region R of the transport path 4 (obtained by directly irradiating the detection light from the light emitting elements Ga to the pixels Gb).
The unadjusted white reference data obtained in the state where the white reference member 10 is placed is hereinafter described also as "first unadjusted data" in some cases for the sake of explanations. The unadjusted white reference data obtained in the state where the white reference member 10 is not placed is described also as "second unadjusted data" in some cases. In a third embodiment, white reference data D (see FIG. 9(c) described later) is determined using both the first unadjusted data and the second unadjusted data while details are described later.
FIGS. 9(a) to 9(c) are explanatory diagrams of the third embodiment. The banknote sorting device 100 in the third embodiment includes the region R (including the regions Rx and Ry) where a banknote passes similarly to the first embodiment described above. As described above, the region Rx in the region R is a region near the right wall surface 7a. The region Ry in the region R is a region near the left wall surface 7b. The region R other than the region Rx and the region Ry is hereinafter described also as "region Rz" for the sake of explanations. The region Rz is a region including the center of the region R.
The banknote sorting device 100 in the third embodiment includes N pixels G. Specifically, the N pixels G are provided in a line in the order of a pixel G1, a pixel G2 ··· and a pixel GN from the left wall surface 7b to the right wall surface 7a. For example, n pixels G from the pixel G1 to a pixel Gn among the N pixels G are provided in the region Ry. N-2n pixels G from a pixel G(n+1) to a pixel G(N-n) are provided in the region Rz and n pixels G from a pixel G(N-n+1) to the pixel GN are provided in the region Rx. In the third embodiment, an ith pixel G as viewed from the left wall surface 7b is described also as "pixel Gi" (1≤i≤N) in some cases.
FIG. 9(a) is an explanatory diagram of the first unadjusted data. The first unadjusted data is substantially same as the unadjusted white reference data illustrated in FIG. 5(b) in the first embodiment described above.
Specifically, pixels from the pixel G(n+1) to the pixel G(N-n) in the region Rx of the region R face the white reference sheet 10a of the white reference member 10. Therefore, the tones (hereinafter, also "tones Ia") of the first unadjusted data of the pixels G in the region Rz can be adopted as the white reference data. Meanwhile, pixels from the pixel G1 to the pixel Gn in the region Ry of the region R face the guard member 10b of the white reference member 10. Therefore, the tones Ia in the region Ry are not appropriate as the white reference data. Also the tones Ia in the region Rx of the region R are not appropriate as the white reference data similarly to the tones Ia in the region Ry.
FIG. 9(b) is an explanatory diagram of the second unadjusted data. The second unadjusted data is substantially same as the unadjusted white reference data illustrated in FIG. 8(b) in the second embodiment described above. However, even when a pixel G is irradiated with detection light that is more intense than a predetermined tone (hereinafter, "upper limit tone Im"), the second unadjusted data of the pixel G is the upper limit tone Im in the third embodiment.
As illustrated in FIG. 9(b), the second unadjusted data in the region Rz of the region R is the upper limit tone Im. Meanwhile, the detection light is refracted in the region Rx and the region Ry of the region R due to the guide parts 8 as described above. Therefore, the tone (hereinafter, also "tone Ib") of the second unadjusted data of each pixel G in the region Rx and the region Ry is not sometimes the upper limit tone Im in some cases as illustrated in FIG. 9(b).
With this configuration, the second unadjusted data in the region Rz is different from the actual tone of the detection light when the actual tone of the detection light is more intense than the upper limit tone Im. Meanwhile, the first unadjusted data in the region Rz indicates the actual tone of the detection light. Therefore, it is desirable to correct the tone of each pixel G in the region Rz with the white reference data obtained from the first unadjusted data than with the white reference data obtained from the second unadjusted data.
As described above, the first unadjusted data in the region Rx and the region Ry is not appropriate as the white reference data. Therefore, it is desirable to correct the tone of each pixel G in the region Rx and the region Ry with the white reference data obtained from the second unadjusted data than with the white reference data obtained from the first unadjusted data. With correction of the tone of each pixel G in the region Rx and the region Ry with the white reference data obtained from the second unadjusted data, the tones of pixels G that do not face a banknote are corrected to the upper limit tone Im and the tones of pixels G that do not face a banknote are corrected to a value lower than the upper limit tone Im. Therefore, there is an advantage that the outer edge of the banknote is easily detected.
In consideration of these circumstances, a configuration in which the white reference data in the third embodiment is determined in such a manner that the white reference data for correcting the tone of each pixel G in the region Rz is determined from the first unadjusted data and the white reference data for correcting the tone of each pixel G in the region Rx and the region Ry is determined from the second unadjusted data is adopted.
FIG. 9(c) is a conceptual diagram of the white reference data D in the third embodiment. The white reference data D is configured to include N correction values d. The correction values d respectively correspond to any of the pixels G. Hereinafter, the correction value d corresponding to a pixel Gi is described as "correction value di". Assuming the tone of a pixel Gi before correction is an actually-measured tone Ivi (1≤i≤N), the tone (hereinafter, "corrected tone Ici") of the pixel Gi after shading correction is obtained by multiplying the actually-measured tone Ivi by the correction value di (Ici=Ivi×di).
These correction values d include correction values d determined from the first unadjusted data (see FIG. 9(a)) and correction values d determined from the second unadjusted data (see FIG. 9(b)). This configuration is restated as that the white reference data D is configured to include correction values d (a first correction value) obtained by irradiating the white reference member 10 with the detection light and correction values d (a second correction value) obtained by irradiating the pixels G (a light receiving unit) with the detection light not having passed through the white reference member 10.
For example, a correction value d1 is determined from a tone Ia1 of the pixel G1 in the region Ry of the second unadjusted data. Similarly, a correction value d2 to a correction value dn are determined from tones Ia (Ia2 to Ian) of the pixel G2 to the pixel Gn in the region Ry of the second unadjusted data. A correction value d(N-n+1) is determined from a tone Ia(N-n+1) of the pixel G(N-n+1) in the region Rx of the second unadjusted data. Similarly, a correction value d(N-n+2) to a correction value dN are determined from tones Ia (Ia(N-n+2) to IaN) of the pixel G(N-n+2) to the pixel GN in the region Rx of the second unadjusted data.
As is understood from the above explanations, the correction value d corresponding to each pixel G in the region Rx and the region Ry is determined from the second unadjusted data. Specifically, the correction value di corresponding to each pixel G in the region Rx and the region Ry (when 1≤i≤n, N-n≤i≤N) is obtained by dividing the upper limit tone Im by a tone Ibi of the second unadjusted data (di=Im/Ibi).
Meanwhile, the correction value d corresponding to each pixel G in the region Rz is determined from the first unadjusted data. Specifically, a correction value d(n+1) is determined from a tone Ia(n+1) of the pixel G(n+1) in the region Rz of the first unadjusted data. Similarly, a correction value d(n+2) to a correction value d(N-n) are determined from tones Ia (Ia(n+2) to Ia(N-n)) of the pixel G(n+2) to the pixel G(N-n) in the region Rz of the first unadjusted data. Specifically, assuming the tone I of white as a white tone Iw, the correction value di corresponding to each pixel G in the region Rz (when n+1≤i≤N-n-1) is obtained by dividing the white tone Iw by the tone Iai of the first unadjusted data (di=Iw/Iai).
The pixels G for which the correction value d is determined from the first unadjusted data and the pixels G for which the correction value d is determined from the second unadjusted data can be appropriately changed. For example, a configuration in which the correction values for pixels G in a region wider than the region Rz are determined from the first unadjusted data may be adopted.
A paper sheet sorting device of the present invention is, for example, a paper sheet sorting device described below.
A paper sheet sorting device (100) of the present invention includes a transport unit (the transport control unit 101) that moves a paper sheet along a transport path (4), a light emitting unit (the light emitting elements Ga) that irradiates one surface of the paper sheet being moved with detection light, a plurality of light receiving units (the pixels Gb) that are placed at locations facing the other surface of the paper sheet being moved to receive the detection light having passed through the paper sheet and that are arranged along a width direction (X-axis direction, scanning axis direction) orthogonal to a moving direction (Y-axis direction) of the paper sheet, an image data generating unit (the image data generating unit 104) that generates image data of tones corresponding to the detection light having entered the light receiving units, a white-reference-data storing unit (the white-reference-data storing unit 105) that stores therein white reference data obtained by irradiating a white reference member (10) to be used for shading correction of the image data with the detection light, a movement restricting unit (the right wall surface 7a and the left wall surface 7b) that is located at end parts in a width direction of the transport path and that forms side walls of the transport path, and an adjusting unit (the adjusting unit 109) that adjusts the white reference data for shading correction of image data of the light receiving units to a predetermined value (the average tone Iave) on a condition that locations in the width direction of the light receiving units are in a predetermined specific region (the adjustment region Ra and the adjustment region Rb). With the above configuration, shading correction of an image of a paper sheet is appropriately performed.
According to a preferred aspect of the present invention, the specific region is a region of which a distance in the width direction to the movement restricting unit is equal to or less than a predetermined distance (a distance from Psa to Pea or a distance from Psb to Peb in FIGS. 5). According to another preferred aspect of the present invention, an abutment unit (the guide parts) that is provided between the light emitting unit and the light receiving units to abut on one surface or the other surface of the paper sheet and that is formed of a member transmitting light is included, and the location in the width direction of the abutment unit is in a specific region.

Reference Signs List

1 lower unit, 10 white reference member, 10a white reference sheet, 10b guard member, 2 upper unit, 3a insertion port, 3b loading port, 4 transport path, 5 transport device, 5a lower transport member, 5a transport unit, 5b upper transport member, 6a first image sensor, 6b second image sensor, 6c magnetic sensor, 7 lower cover, 7a right wall surface, 7a one to right wall surface, 7b left wall surface, 7c recess, 7d upper surface, 8 guide part, 8 upper tray, 8L guide part, 8R guide part, 8a concave portion, 8b lower surface, 9 reflecting white reference piece, 100 banknote sorting device, 101 transport control unit, 102 sensor unit, 103 sensor control unit, 104 image data generating unit, 105 white-reference-data storage unit, 106 correcting unit, 107 white-reference-data generating unit, 108 adjustment position determining unit, 109 adjusting unit, 10b' long-side end part.

Claims

A paper sheet sorting device comprising:
a transport unit that moves a paper sheet along a transport path;

a light emitting unit that irradiates one surface of the paper sheet being moved with detection light;

a plurality of light receiving units that are placed at locations facing the other surface of the paper sheet being moved to receive the detection light having passed through the paper sheet and that are arranged along a width direction orthogonal to a moving direction of the paper sheet;

an image data generating unit that generates image data of tones corresponding to the detection light having entered the light receiving units;

a white-reference-data storing unit that stores therein white reference data obtained by irradiating a white reference member to be used for shading correction of the image data with the detection light;

a movement restricting unit that is located at end parts in a width direction of the transport path and that forms side walls of the transport path; and

an adjusting unit that adjusts the white reference data for shading correction of image data of the light receiving units to a predetermined value on a condition that locations in the width direction of the light receiving units are in a predetermined specific region.
The paper sheet sorting device according to claim 1, wherein the specific region is a region of which a distance in the width direction to the movement restricting unit is equal to or less than a predetermined distance.
The paper sheet sorting device according to claim 1 or 2, comprising an abutment unit that is provided between the light emitting unit and the light receiving units to abut on one surface or the other surface of the paper sheet and that is formed of a member transmitting light, wherein
a location in the width direction of the abutment unit is in the specific region.
An adjustment method of white reference data that is used in a paper sheet sorting device including: a transport unit that moves a paper sheet along a transport path; a light emitting unit that irradiates one surface of the paper sheet being moved with detection light; a plurality of light receiving units that are placed at locations facing the other surface of the paper sheet being moved to receive the detection light having passed through the paper sheet and that are arranged along a width direction orthogonal to a moving direction of the paper sheet; an image data generating unit that generates image data of tones corresponding to the detection light having entered the light receiving units; and a movement restricting unit that is located at end parts in a width direction of the transport path and that forms side walls of the transport path, to perform shading correction of the image data, the method comprising
a step of adjusting the white reference data for shading correction of image data of the light receiving units to a predetermined value on a condition that locations in the width direction of the light receiving units are in a predetermined specific region.
A program that causes a computer to adjust white reference data to be used in a paper sheet sorting device including: a transport unit that moves a paper sheet along a transport path; a light emitting unit that irradiates one surface of the paper sheet being moved with detection light; a plurality of light receiving units that are placed at locations facing the other surface of the paper sheet being moved to receive the detection light having passed through the paper sheet and that are arranged along a width direction orthogonal to a moving direction of the paper sheet; an image data generating unit that generates image data of tones corresponding to the detection light having entered the light receiving units; and a movement restricting unit that is located at end parts in a width direction of the transport path and that forms side walls of the transport path, to perform shading correction of the image data, the program causing the computer to function as
an adjusting unit that adjusts the white reference data for shading correction of image data of the light receiving units to a predetermined value on a condition that locations in the width direction of the light receiving units are in a predetermined specific region.
A paper sheet sorting device comprising:
a transport unit that moves a paper sheet along a transport path;

a light emitting unit that irradiates one surface of the paper sheet being moved with detection light;

a plurality of light receiving units that are placed at locations facing the other surface of the paper sheet being moved to receive the detection light having passed through the paper sheet and that are arranged along a width direction orthogonal to a moving direction of the paper sheet;

an image data generating unit that generates image data of tones corresponding to the detection light having entered the light receiving units; and

a white-reference-data storing unit that stores therein white reference data to be used for shading correction of the image data, wherein

the white reference data is configured to include a first correction value obtained by irradiating a white reference member with the detection light and a second correction value obtained by irradiating the light receiving units with the detection light not having passed through the white reference member.