This technology relates to a bill discriminating apparatus that performs bill discrimination.
Prior Art
Bill discriminating apparatuses for performing bill discrimination are incorporated in cash automatic teller machines (ATM) installed at financial institutions, etc. and in automatic vending machines. A variety of discrimination technologies were proposed for this bill discriminating apparatus in the past. As one of these discrimination technologies, there is a technology that discriminates bills to which a foreign object such as tape is adhered.
For example, with the bill discriminating apparatus noted in Japanese Patent Laid-Open Gazette No. 7-6245, by comparing the detection voltage in relation to the thickness of a bill type, the tape detection reference voltage, and the negotiable security detection reference voltage, a detection is made of whether or not there is tape. Then, furthermore, when tape is adhered extending to both ends of the negotiable security, a text image of two locations on the negotiable security is taken, and by comparing these, discrimination is performed for whether this is authentic or not.
Also, with the bill determination apparatus noted in Japanese Patent Laid-Open Gazette No. 63-247895, a detection signal output from a thickness sensor while the bills are being conveyed is taken, data that shows the irregularity pattern of the bill is obtained, and by comparing this with a reference pattern, a determination is made of whether the bill is authentic or not. With this technology, it is also possible to perform two sheet overlap detection (hereafter called double feed detection) that detects when two bills or more are fed overlapping each other, or to perform detection of whether there is tape on the bill (hereafter called tape detection).
However, among bills to which tape is adhered (hereafter called taped notes), there could be authentic notes which are completely torn bills that are fixed using tape, or bills that look like they will tear that have been reinforced using tape, or counterfeit notes that have been altered by sticking together part of an authentic note and part of a counterfeit note (altered note). In the past, it was not possible to discriminate these types of bills with good precision. Also, with the prior art described above, discrimination of bill authenticity was performed by whether bill thickness was within a specified range, and the precision of authenticity discrimination was improved.
This technology was created to solve the problems described above, and its purpose is to improve bill discriminating precision.
To solve at least part of the problems described above, the following structure was used. Specifically, the first bill discriminating apparatus of this technology is a bill discriminating apparatus that performs bill discrimination, comprising a storage unit that stores in advance reference thickness distribution data that shows the distribution of the thickness of authentic notes, a thickness detection unit that detects the distribution of thickness of the bill that is subject to discrimination, and a discrimination unit that performs a specified discrimination based on the reference thickness distribution data and the detected detection thickness distribution data, wherein the discrimination unit performs the bill authenticity discrimination based on the reference thickness distribution data for an area for which the absolute value of the difference between the detected thickness distribution data and the reference thickness distribution data is a specified threshold value or less, and on the detected thickness distribution data.
The reference thickness distribution data and detected thickness distribution data described above is data that shows the spatial shape of a bill. Therefore, by comparing both of these, it is possible to improve the precision of bill discrimination to be better than prior art authenticity discrimination based on whether the bill thickness is within a specified scope. If the thickness distribution data is made to be three dimensional data that shows the in-plane distribution of a bill, it is possible to further improve the bill discrimination precision. Moreover, detection of bill thickness distribution may also be performed by scanning the bill by having it contact a roller or sensing pin, etc., or by performing this without contact using light or sound waves, etc. Ink irregularities due to intaglio printing are also included in bill thickness distribution.
Authenticity discrimination is also performed for when the absolute value of the difference between the detected thickness distribution data and the reference thickness distribution data is in a range of a specified threshold value or less. This is because for areas for which the absolute value of the difference between the detected thickness distribution data and the reference thickness distribution data is greater than a specified threshold value, this can be regarded as the thickness of the bill itself being thick, as multiple bills overlapping, as the thickness of the bill itself being thin, or as an adhered item such as tape being adhered to part of the bill or as being missing. By working in this way, even if there is adherence of tape, etc. or if there are abnormalities in the dimensions or thickness, it is possible to perform discrimination by effectively using other normal range data. For example, for a bill that is an authentic note which has been reinforced using tape, it is possible to determine this to be an “authentic note.”
Following, we will explain embodiments of the present technology based on embodiments in the following order.
The bill discriminating apparatus 100 comprises an image sensor 10, a thickness detection mechanism 20, and a control unit 30. The bill discriminating apparatus 100 also comprises a conveyance mechanism that is not illustrated such as a sensor for detecting the presence of the bill P, a conveyance roller for conveying the bill P, or a guide that guides the bill P in the conveyor.
The image sensor 10 takes an image of the bill P while the bill P is being conveyed. This image is used for detection of the bill P conveyance position or for discrimination.
The thickness detection mechanism 20 detects the thickness of the bill P during conveyance in the sub-scanning direction (bill conveyance direction) that is illustrated at multiple timings across the overall bill P by twelve of the sensors 24 which are placed in the main scanning direction. With this embodiment, detection of the thickness of the bill P is performed every 0.5 mm in the sub-scanning direction, and the average value of four detections was used as one thickness data. In other words, for the sub-scanning direction, the thickness data is data for every 2 mm. Furthermore, it is also possible to detect detailed thickness data. The detected multiple thickness data are arrayed in the thickness distribution data that shows the bill P thickness distribution, and used for discrimination of the bill P.
With this embodiment, we arranged twelve of the sensors 24 on the thickness detection mechanism 20, but it is also possible to arrange even more of the sensors 24, and for the main scanning direction, to detect the thickness at a large number of points. Moreover, for convenience of illustration, for the thickness detection mechanism 20, we did not depict parts other than the sensor 24 and the rotation axis 22a. We will give a detailed explanation of the thickness detection mechanism 20 later.
Also, with this embodiment, as shown in
The control unit 30 controls the operation of the conveyance mechanism, the image sensor 10, or the thickness detection mechanism 20, and also executes the discrimination process of the bill P which will be described later.
B. Thickness Detection Mechanism:
The thickness detection mechanism 20 comprises one reference roller 21, twelve detection rollers 22, twelve plate springs 23 that are paired with each detection roller 22, and twelve sensors 24.
The reference roller 21 is a roller for deciding the reference position in the thickness direction for the thickness detection of the bill P, and is fixed in relation to the thickness detection mechanism 20. The reference roller 21 is driven by a motor that is not illustrated, and also functions as a conveying roller for conveying the bill P.
The detection roller 22 is arranged so that its surface is in contact with the surface of the reference roller 21. This detection roller 22 is formed from a rotation axis 22a, an elastic body 22b that is provided around its periphery, and a cylindrical shaped roller unit 22c that is further provided around the periphery. The rotation axis 22a is common to twelve detection rollers 22, and is fixed to the thickness detection mechanism 20. By using this kind of structure, as shown in
The plate spring 23 is placed so as to be in contact with the surface of the detection roller 22. The sensor 24 is placed in a position that is separated from the plate spring 23 by a distance d=d1. With this embodiment, we used an electrostatic capacity type sensor for the sensor 24. As shown in
With this embodiment, we used an electrostatic capacity type sensor for the thickness detection mechanism 20, but instead of this, it is also possible to use another sensor that is capable of detecting changes in distance d, such as an eddy current type sensor or a piezoelectric type sensor, etc.
C. Control Unit:
The thickness distribution data generating unit 31 gets the bill P thickness data detected by the thickness detection mechanism 20, arrays these, and generates thickness distribution data. The image data acquisition unit 32 gets an image of the bill P that is taken by the image sensor 10.
The discrimination unit 33 comprises a denomination detection unit 31, a position detection unit 332, and a corrected distribution data generating unit 333, and performs discrimination of the bill P. By comparing the image data acquired by the image data acquisition unit 32 and the data that is stored inside the storage unit 34 to be described later, the denomination detection unit 331 detects the denomination of the bill P. The position detection unit 332 detects the conveyance position of the bill P based on the image data acquired by the image data acquisition unit 32. The conveyance position of the bill P is characterized by four parameters as will be described later, specifically, back and front, conveyance direction, shift volume, and skew angle. The corrected distribution data generating unit 333, of the parameters characterized by the position detection unit 332, based on the conveyance direction, shift volume, and skew angle, in order to compare with the reference thickness distribution data stored in the storage unit 34, corrects the thickness distribution data so as to correct the skew of both, and generates corrected distribution data. By working in this way, even if the position of the bill that is subject to discrimination is skewed from the reference, it is possible to perform discrimination processing with good precision.
The storage unit 34 stores various types of data related to authentic notes for the discrimination unit 33 to reference during discrimination of the bill P. The contents of these data are shown in typical form in
Moreover, with this embodiment, we had the storage unit 34 store reference thickness distribution data for each front and back for one denomination, but it is also possible to have it store one reference thickness distribution data that includes the front and back thickness distribution. The storage unit 34 may also be set to store reference thickness distribution data to match the detection mode of the thickness detection mechanism 20. The output unit 35 outputs the discrimination results of the discrimination unit 33 to the outside.
Moreover, the conveyance positions explained here are nothing more than examples, and this is not limited to these. It is also possible to define the conveyance position using even more parameters. It is also possible to omit part of the parameters described above. Also, with this embodiment, the conveyance position described above was characterized by the image processing of the bill P, but it is also possible to separately provide a sensor for detecting the conveyance position.
D. Discrimination:
The bill discriminating apparatus 100 of this embodiment may perform three types of discrimination: double feed detection, tape detection, and authenticity discrimination.
In this way, when there is an area for which the thickness is an integral multiple of the authentic note thickness, and the dimensions are greater than the dimensions of an authentic note, it is possible to detect that there is double feed. For example, since it is also possible to have a case with double feed of a bill (counterfeit note) for which the thickness is different from that of an authentic note, regardless of whether or not there is an area for which the thickness is an integral multiple of the thickness of an authentic note, when the dimensions are greater than the dimensions of an authentic note, it is also possible to have this case judged as double feed. Moreover, here, for convenience of explanation, we explained a summary of double feed detection based on the thickness data in the A-A cross section of
In this way, when there is an area for which the thickness is different from an integral multiple of the thickness of an authentic note, and there is an area for which the thickness is authentic note thickness t1, it is possible to judge that this is a taped note. Moreover, here, for convenience of explanation, we explained a summary of tape detection based on the thickness data for the cross section B-B of
When a tape T is not adhered to the bill P, as shown by the areas A and B shown in
With the authenticity discrimination of this embodiment, the area for which the difference between the detected data and the reference data is within a specified range (areas A and B) is stipulated as the area subject to authenticity discrimination. Then, for this area subject to authenticity discrimination, based on the reference data and the detected data, a detailed authenticity discrimination is performed for whether or not the difference between these data is within a specified reference range. The reference range can be set freely considering the bill wearing, etc. By performing authenticity discrimination in this way, regardless of whether or not the bill subject to discrimination is a taped note or not, it is possible to perform authenticity discrimination.
E. Discrimination Process:
E1. Discrimination Process Summary:
The bill discriminating apparatus 100 of this embodiment classifies bills subject to discrimination as authentic notes for which the dimensions and thickness distribution are correct, counterfeit notes for which the dimensions are abnormal, counterfeit notes for which the thickness distribution is abnormal, authentic notes to which tape is adhered, or counterfeit notes to which tape is adhered using the discrimination process shown hereafter.
Next, the bill size that corresponds to the denomination determined in step S110 is acquired from the storage unit 34 (step S120). This size includes the authentic note dimensions and thickness. Hereafter, these will be called the dimension reference value and the thickness reference value. Then, the difference between the maximum value of the bill thickness data and the thickness reference value, in other words, (maximum value of the bill thickness data)−(thickness reference value) is calculated (step S130), and a judgment is made of whether or not this value is greater than a specified value (step S140).
In step S140, when the difference between the maximum value of the bill thickness data and the thickness reference value is greater than the specified value, the double feed detection process (step S200) described later is executed, and after that, the authenticity discrimination process (step S300) is executed.
In step S140, when the difference between the maximum value of the bill thickness data and the thickness reference value is less than the specified value, a judgment is made of whether or not the bill dimensions are equivalent to the dimensions reference value (step S150). When the bill dimensions are equivalent to the dimensions reference value, the authenticity discrimination process (step S300) is executed. Meanwhile, when the bill dimensions and the dimension reference value are not equivalent, this is judged to be a counterfeit note (step S160), and the bill is returned (step S170). At this time, instead of returning, it is also possible to withdraw the bill that is judged to be a counterfeit note. This is the same for the process noted below as well.
E2. Double Feed Detection Process:
First, a judgment is made of whether or not the maximum value of the bill thickness data is an integral multiple of the thickness reference value (step S210). If the maximum value of the bill thickness data is an integral multiple of the thickness reference value, a judgment is made of whether or not the bill dimensions are greater than the dimension reference value (step S220). Then, if the bill dimensions are greater than the dimension reference value, this is judged as being double feed (step S230), and the bills are returned (step S250). Note that either of the processes of step S210 and step S220 can be omitted.
In step S220, if the bill dimensions are less than the dimension reference value, a judgment is made that this is a counterfeit note (step S240), and the bill is returned (step S250). Note that when a bill is returned, the process does not advance to the authenticity discrimination process (step S300) of
In step S210, when the maximum value of the bill thickness data is not an integral multiple of the thickness reference value, a judgment is made of whether or not the bill dimensions are equivalent to the dimension reference value (step S260). When the bill dimensions are equivalent to the dimension reference value, this is judged as not being double feed (step S270), and a return is done. When the bill dimensions and the dimension reference value are not equivalent, this is judged as a counterfeit note (step S240), and the bill is returned (step S250).
E3. Authenticity Discrimination Process:
First, based on the conveyance position that was judged with step S110 of
Next, for the entire area of the bill, the difference between the corrected distribution data and the reference thickness distribution data, in other words, (corrected distribution data)−(reference thickness distribution data) is calculated (step S330). Then, the same as shown in
Next, the difference between the corrected distribution data and the reference thickness distribution data sets the area that is inside the specified range as the area subject to authenticity discrimination (step S350). Then, a judgment is made of whether or not the area subject to authenticity discrimination exists at greater than a specified ratio (step S360). With this embodiment, it was set so that a judgment is made of whether or not the area subject to authenticity discrimination exists at 67% or greater.
In step S360, when the area subject to authenticity discrimination is not at greater than a specified ratio, this is judged as a counterfeit note (step S382), and the bill is returned (step S384). By working in this way, it is possible to ensure the precision of the authenticity discrimination. When the area subject to authenticity discrimination is at greater than the specified ratio, for the area subject to authenticity discrimination, a detailed analysis is done of the difference in the thickness distribution including ink irregularities due to intaglio printing (step S370), and a judgment is made of whether or not the analysis results are within a specified reference range. The reference range cane be set freely taking into consideration bill wearing, etc. Note that with this embodiment, the authenticity discrimination is performed based on the difference between the corrected distribution data and the reference distribution data, but it is also possible to perform authenticity discrimination based on the absolute value of the difference of both items.
In step S380, when the analysis results are within the reference range, this is judged as being an authentic note (step S386). When the analysis results are outside the reference range such as when the thickness of the bill itself is thick or thin, or the unevenness state of the bill surface is different from the unevenness state of the authentic note, etc., this is judged as being a counterfeit note (step S382), and the bill is returned (step S394).
Using the discrimination process described above, it is possible to classify the bills subject to discrimination into authentic notes for which the dimensions and thickness distribution is correct, counterfeit notes for which the dimensions are abnormal, counterfeit notes for which the thickness distribution is abnormal, authentic notes to which tape is adhered, and counterfeit to which tape is adhered.
With the bill discriminating apparatus 100 of this embodiment noted above, authenticity discrimination is performed based on reference thickness distribution data which is three dimensional data that shows the spatial shape of a bill and on corrected distribution data, so this allows for an improvement in the precision of bill discrimination compared to the prior art authenticity discrimination based on whether or not the bill thickness is within a specified range.
Also, the bill discriminating apparatus 100 of this embodiment performs authenticity discrimination for areas subject to authenticity discrimination for which the difference between the correction thickness distribution data and the reference thickness distribution data is within a specified range, so even in cases when there are abnormalities in the dimensions or size such as when there is tape adhered, it is possible to perform discrimination by effectively using other correct area data. Furthermore, when the thickness of the bill subject to discrimination is greater than the thickness of an authentic note, regardless of whether or not that bill is an authentic note, it is possible to perform double feed detection and tape detection. Therefore, it is possible to improve the precision of bill discrimination.
As noted above, this embodiment may be implemented in various forms that are in a range that do not stray from the gist of this embodiment. The following variation examples are possible, for example.
The thickness detection mechanism 20A comprises a detection roller 22d and an arm 23a of this detection roller 22d in place of the detection roller 22 and plate spring 23 that form the thickness detection mechanism 20. Then, the detection roller 22d is supported on the arm 23a by the axis 3c, and the arm 23a is supported on the thickness detection mechanism 20A by the axis 23b. The axis 23b is common to twelve arms 23a. The structure other than this is the same as for the thickness detection mechanism 20.
By detecting changes in the electrostatic capacity of the gap between the sensor 24 and the arm 23a, the sensor 24 is able to detect changes in the distance d between the sensor 24 and the plate spring 23.
With the thickness detection mechanism 20A explained above as well, like with the thickness detection mechanism 20 shown in
With the aforementioned embodiments and the variation example, bill thickness detection was performed by having the detection roller contact the bill and scanning, but the technology is not limited to this. For example, it is also possible to perform bill thickness detection by scanning a sensing pin. It is also possible to perform this without any contact using light (transmitted light, reflected light) or sound waves, etc.
With the aforementioned embodiment, the threshold value tht shown in
With the aforementioned embodiments, the corrected distribution data generating unit 333 was made to correct thickness distribution data based on the conveyance position, but the technology is not limited to this. It is also possible to have correction performed so as to compensate the displacement of both items for at least one of the thickness distribution data and the reference thickness distribution data.
This application claims the benefit of priority of Japanese Application No. 2003-414602 filed Dec. 12, 2003, the disclosure of which also is entirely incorporated herein by reference.
Number | Date | Country | Kind |
---|---|---|---|
2003-414602 | Dec 2003 | JP | national |
Number | Name | Date | Kind |
---|---|---|---|
5447240 | Makino | Sep 1995 | A |
5678678 | Brandt et al. | Oct 1997 | A |
Number | Date | Country |
---|---|---|
0 097 570 | Jan 1984 | EP |
0 798 670 | Oct 1997 | EP |
63-247695 | Oct 1988 | JP |
07-006245 | Jan 1995 | JP |
7-6245 | Jan 1995 | JP |
Number | Date | Country | |
---|---|---|---|
20050141759 A1 | Jun 2005 | US |