JP3381828B2 - Method and apparatus for determining damage to paper sheets - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for discriminating a paper sheet, which optically reads a paper sheet conveyed at a high speed and discriminates a damage of the paper sheet based on the read signal. In particular, the present invention relates to a method and apparatus for determining damage to a paper sheet that reduces the influence of uncertain factors such as vertical vibration of the paper sheet due to high-speed conveyance and folding of the paper sheet, and accurately determines the degree of damage to the paper sheet. .

[0002]

2. Description of the Related Art Paper sheets such as banknotes, checks, securities and certificates are not broken or chipped, but are discolored, dirty, worn, broken, wrinkled, scratched or peeled off by printing ink (hereinafter referred to as "damage"). Therefore, it is necessary to exclude such severely damaged paper sheets from the distribution process.

Therefore, conventionally, the light receiving sensor integrates the output signal of the transmitted light or the reflected light detected from the paper sheet to be discriminated and compares the integrated value with a predetermined threshold value to determine the paper sheet. Often to determine the presence or absence of damage.

For example, in Japanese Patent Laid-Open No. 59-57107, the reflected light or transmitted light from the paper to be judged is photoelectrically converted, and the difference between the average value of the photoelectric conversion signals and each photoelectric conversion signal is accumulated. There is disclosed a paper sheet damage determination method configured to determine the damage of a determination target paper sheet by comparing the sum with a predetermined determination value.

Further, in JP-A-60-146388, the amount of transmitted light or the amount of reflected light of a conveyed sheet is detected, and the detected amount of reflected light is divided into a printed portion and a plain portion of the sheet. There is disclosed a paper sheet discrimination device configured to perform integration and use each integration result to detect whether there is stain.

As described above, in the conventional paper sheet discrimination technology,
A light receiving sensor detects the amount of transmitted light or the amount of reflected light of the paper sheet conveyed at high speed, and the damage of the paper sheet is determined based on the detection output.

[0007]

However, in the banknote processing apparatus mounted on a cash dispenser or the like of a financial institution, the banknote itself fluctuates up and down due to high speed conveyance, so that the output of the light receiving sensor is not stable.

If the surface of the bill has folds or wrinkles, the angle of the reflected light greatly changes at the folds of the bill, and the diffusion distribution of the transmitted light also has a great influence at the folds. receive.

FIG. 30 is a diagram showing the influence of a fold of a bill on the diffusion distribution of transmitted light. As shown in FIG. 30A, in the case of a normal bill, the bill is transmitted in the vertical direction of the bill. Light is diffused, but if there is a fold in such a bill, the same figure (b)
Since the transmitted light diffuses in an oblique direction as shown by, the amount of light received by the light receiving element fluctuates under the influence of this fold.

From these things, even if the sheets are exactly the same, except for the sheets that have been extremely damaged, it may be determined that they are normal at some times and that they are damaged at other times. As a result, the accuracy of discrimination deteriorates.

Therefore, it is possible to reduce the influence of uncertain factors such as vertical vibration of the paper sheet and folding of the paper sheet due to high speed conveyance,
An important issue is how to accurately determine the degree of damage to paper sheets.

Therefore, the present invention solves the above problems by
A method and apparatus for determining damage to a paper sheet that can accurately determine the degree of damage to the paper sheet by reducing the influence of uncertain factors such as vertical vibration of the paper sheet and folding of the paper sheet due to high-speed conveyance. The purpose is to provide.

[0013]

In order to achieve the above-mentioned object, the invention of claim 1 optically reads a paper sheet conveyed at a high speed, and damages the paper sheet based on the read signal. In the damage determination method of the paper sheet to be determined, the paper sheet is irradiated with light when passing a predetermined position, and at least one of the transmitted light and the reflected light in the paper sheet is acquired as light reception data, It is characterized in that the length of the waveform trace formed by the acquired light reception data is calculated, and the degree of damage to the paper sheet is determined based on the calculated length.

According to a second aspect of the present invention, in a sheet damage determination device for optically reading a sheet conveyed at a high speed and determining the damage of the sheet based on the read signal, Irradiation means for irradiating the leaves with light, and when the paper leaves pass a predetermined position, at least one of the transmitted light and reflected light in the paper leaves of the light emitted by the irradiation means. Data acquisition means for acquiring as light reception data, length calculation means for calculating the length of the trajectory of the waveform formed by the light reception data acquired by the data acquisition means, and the length calculation means for calculating the length based on the length calculated by the length calculation means. It is characterized in that it is provided with a discriminating means for discriminating the degree of damage to the paper sheets.

The invention of claim 3 is characterized in that, in the invention of claim 2, the length calculating means calculates the length of the locus of the waveform from the value of each received light data acquired at a predetermined interval. And

According to a fourth aspect of the present invention, in the second aspect of the present invention, the discriminating means stores the length of the locus of the waveform formed by the light receiving data of the reference paper sheet as the reference length, and the storage means. A characteristic parameter calculation unit that calculates a characteristic parameter from the length of the trajectory of the waveform calculated by the length calculation unit and the reference length stored in the storage unit; and the paper sheet based on the characteristic parameter calculated by the characteristic parameter calculation unit. And a determining means for determining the degree of damage to the class.

According to a fifth aspect of the invention, in the second aspect of the invention, the data acquisition means acquires the light reception data divided into a predetermined number of blocks, and the length calculation means is the data acquisition means. It is characterized in that the lengths of the loci of the waveforms formed by the respective received light data are calculated, and the calculated lengths are added over all the blocks.

The invention according to claim 6 is the invention according to claim 2, further comprising low frequency component removing means for removing low frequency components from the received light data acquired by the data acquiring means, and the length calculating means. Is characterized in that the length of a locus of a waveform formed by the received light data acquired by the data acquisition means is calculated after the low frequency component of the received light data is removed by the low frequency component removing means.

According to a seventh aspect of the present invention, the recording used by the paper sheet damage discriminating apparatus for optically reading the paper sheet conveyed at high speed and discriminating the damage of the paper sheet based on the read signal. At least one of the transmitted light and the reflected light of the medium, which is the light emitted when the paper sheet passes through the predetermined position, is acquired as the light reception data, and the acquired light reception data is It is characterized in that a program for calculating the length of the locus of the formed waveform and discriminating the degree of damage to the paper sheet based on the calculated length is recorded.

[0020]

[0021]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. In the present embodiment,
A case where the present invention is applied to a bill discriminating apparatus will be described.

First, the concept of the bill discriminating process adopted in this embodiment will be described with reference to FIG.

FIG. 2 is a diagram showing an example of waveforms of output signals of reflected light of the reference bill and the bill to be discriminated.

A curve 21 shown in FIG. 5A is a waveform of an output signal of reflected light of a predetermined line of a certain reference bill, and a curve 22 shown in FIG. 4B is a waveform of the same line of the bill to be discriminated. 3 is a waveform of the output signal of the reflected light of.

That is, the curve 22 showing the output signal of the bill to be discriminated is not damaged even if it is a Japanese 1000-yen bill or a US dollar bill because the bill has damage such as folding and wrinkles. The fluctuation of the waveform becomes larger than that of the curve 21 showing the output signal of the reference bill.

Here, conventionally, since the degree of damage is discriminated by paying attention to the strength of the output signal, if the bill to be discriminated is conveyed at high speed and the bill itself vibrates in the vertical direction, the output signal of the output signal is irrelevant to the damage. The intensity changes, and as a result, the discrimination accuracy decreases.

Further, when paper-quality bills such as Japanese 1000-yen bills are used, the waveform of not only the reflected light but also the transmitted light becomes complicated due to the fatigue of the bills. In the case of using the bill, the waveform of the transmitted light becomes reverse and becomes gentle.

As described above, the degree of change in the waveform of the output signal differs depending on the paper quality of the bill to be discriminated and whether the reflected light or the transmitted light is used. However, the high-speed conveyance of banknotes deteriorates the discrimination accuracy, and the discrimination method cannot be widely applied to the damage discrimination of various paper sheets.

From these facts, in this embodiment, not the intensity itself of the transmitted light or the reflected light is focused, but the waveform of the transmitted light or the reflected light is focused, and the length of the waveform is used to damage the waveform. Banknotes are being identified.

That is, focusing on the intensity of transmitted light or reflected light itself, the intensity of transmitted light or reflected light changes when a bill is conveyed at high speed or when the bill has a crease. In the embodiment, the influence of such an uncertain factor is removed by focusing on the waveform itself.

In the present embodiment, the banknote is divided into a plurality of blocks, and the degree of damage of the banknote is determined by obtaining the later-described characteristic parameter γ in both the longitudinal direction and the lateral direction.

Next, the structure of the bill discriminating apparatus used in the present embodiment will be described.

FIG. 1 is a diagram showing the configuration of the bill discriminating apparatus 10 used in this embodiment.

As shown in the figure, the bill discriminating apparatus 10
Is a device that acquires image data using transmitted light and reflected light, and determines the damage of the banknote using the length of the waveform of each image data. The data collection unit 11 and the length calculation unit 12 And an image data storage unit 13, a determination unit 14, and a determination data storage unit 15.

The data collecting unit 11 uses the roller 11b to convey the bill 11a at high speed, and the transparent LED array 11c is used.
When the bill 11a reaches a predetermined position by irradiating light with the reflection LED array 11d, the transmitted light and the reflected light are detected by the light receiving element array 11e. Whether or not the bill 11a has reached a predetermined position is determined by a timing sensor (not shown) or the like, which is separately provided, or by monitoring the output of the light receiving element array 11e.

Then, the detection output of the light receiving element array 11e is amplified by the amplifier 11f, converted into analog / digital by the A / D converter 11g, and then output to the length calculator 12.

Specifically, the data collection unit 11
The photodiodes are arranged at a pitch of 1.6 mm and the signals at the pitch are converted into digital signals and output. The light receiving element array 11e is composed of 128 channels and is commonly used for transmitted light and reflected light. The data collection unit 11 temporarily stores the data of the transmitted light and the reflected light of the banknote in the image data storage unit 13 over the entire surface of the banknote.

For example, when four lights of reflected light green, reflected light infrared, transmitted light green, and transmitted light infrared are used, the reflecting LED array 11d and the transmitting L are arranged in this order.
The ED array 11c sequentially emits light and the light receiving element array 11e
Detects light in synchronization with each light emission. In addition, this Embodiment demonstrates the case where infrared rays of reflected light and infrared rays of transmitted light are used.

In this embodiment, the transmissive LED array 11c and the light receiving element array 11e are the bill 1
1a is arranged in the direction perpendicular to the conveying direction, and the reflection LED array 11d is arranged obliquely at the lower right part of the bill 11a, but the present invention is not limited to this.

Specifically, as shown in FIG. 3 (a), the light-receiving element array 11e can be obliquely arranged in the lower left portion in the same manner as the reflection LED array 11d, and as shown in FIG. 3 (b). As shown, the transmissive LED array 11c can also be arranged diagonally in the upper right portion of the banknote 11a.

Further, as shown in FIG. 3C, the light receiving element array 11e corresponding to the transmissive LED array 11c.
It is also possible to separately provide the light receiving element array 11e ′ corresponding to the reflection LED array 11d.

The length calculation unit 12 traces the waveform of the reflected light or the transmitted light received from the data collection unit 11 with a line segment having a length l to obtain the length (distance) of each curve.
Is a processing unit that approximates.

Specifically, the received light data of the transmitted light and the reflected light received from the data collection unit 11 is once stored as image data in the image data storage unit 13, and as shown in FIG. It is divided into 15 blocks up to B15, and the length of each curve is obtained in the longitudinal direction and the lateral direction for each block.

For example, one curve of a block is shown in FIG.
In case of the above, the length of the curve is calculated at a predetermined interval. The length L1 of the curve is L1 = l1 + l2 + l3
.... + ln will be obtained.

The image data storage unit 13 has four image memories and separately stores image data of four lights of reflected light green, reflected light infrared, transmitted light green, and transmitted light infrared. The image data is stored in the image memory.

The determination unit 14 adds the length L1 of the waveform curve of the bill 11a to be discriminated calculated by the length calculation unit 12 for each block and adds the sum of the lengths of the curves of the blocks in the longitudinal direction and the lateral direction. The characteristic length γ is calculated by dividing the sum of the calculated lengths by the length of the waveform curve of the reference bill stored in the discrimination data storage unit 15.

That is, the characteristic parameter γ is γ =
It is calculated as (sum of lengths in block) / (length of reference banknote) for each block in each of the longitudinal direction and the lateral direction.

Then, the calculated characteristic parameter γ is compared with the discrimination reference value stored in the discrimination data storage unit 15 to discriminate the damage degree in each of the longitudinal direction and the lateral direction in each block of the bill 11a.

Since the degree of damage is obtained for each of the image data of two types of light, that is, infrared light of reflected light and infrared light of transmitted light, these data are comprehensively judged and the bill 11a Determine the overall degree of damage.

Specifically, as shown in FIG. 6, the damage level of each block is 0 to 3 in consideration of both the characteristic parameter of the reflected light output and the characteristic parameter of the transmitted light output.
To determine which of the above applies.

The discrimination data storage unit 15 is a storage unit that stores in advance the length of the curve of the waveform of the reference bill and the discrimination reference value that is the comparison target of the characteristic parameters.
4 is accessed.

The configuration of the bill discriminating apparatus 10 shown in FIG. 1 has been described above.

Next, a processing procedure of the bill discriminating apparatus 10 shown in FIG. 1 will be described. However, the discrimination data storage unit 15
In, it is assumed that the length of the curve of the waveform of the reference bill and the discrimination reference value have already been stored.

FIG. 7 is a flowchart showing the processing procedure of the bill discriminating apparatus 10 shown in FIG. 1. Here, for convenience of explanation, only the case where infrared rays of reflected light are used will be explained.

As shown in the figure, the bill discriminating apparatus 10
Is a bill 1 in which the reflection LED array 11d is conveyed at high speed.
1a is irradiated with infrared light, and the reflected light is received by the light receiving element array 1
The image data is detected by 1e, and is stored in the image data storage unit 13 (step 701).

Here, as a preprocessing for length calculation, the length calculation unit 12 corrects the skew of the image in order to eliminate the case where the bill 11a is conveyed at an angle (step 702), and also performs high-speed conveyance. In the process, the bill 11a is often wavy in the vertical direction, so the low frequency component is removed using a filter (step 703).

After that, the length calculation unit 12 obtains the lengths of the curves in the longitudinal direction and the lateral direction for each block of the image data stored in the image data storage unit 13, and the determination unit 14 determines the length of each block. Is calculated (step 704). Here, the length of the curve can be simply calculated in the longitudinal direction, for example, from the data of the A / D conversion value of the adjacent element of each light receiving element.

Then, the judgment section 14 divides the sum of the block lengths by the length of the curve of the waveform of the reference bill stored in the judgment data storage section 15 to obtain the characteristic parameter γ (step 705), and the characteristic parameter γ is calculated. Parameter γ is the discrimination data storage unit 1
5 is compared with the discrimination reference value stored in step 5 (step 706),
The damaged block and the degree of damage are determined based on the comparison result (step 707). (For domestic banknotes, the largest one of the characteristic parameters γ of each block is compared with the discrimination reference value. For the transmitted light of US banknotes, the smallest one of the characteristic parameters γ is compared with the discrimination criterion as described later. The processing procedure of the bill discriminating apparatus 10 shown in FIG. 1 has been described above.

Next, an example of damage discrimination of 1000-yen bills using the bill discriminating apparatus 10 shown in FIG. 1 will be specifically described.

FIG. 8 is a diagram showing image data of a 1000-yen bill read by using the data collection unit 11 shown in FIG. 1. In FIG. 8, image data using green reflected light (hereinafter referred to as “reflection”) is shown. G image ”), image data using infrared reflected light (hereinafter referred to as“ reflected IR image ”), and image data using green transmitted light (hereinafter referred to as“ transmitted G image ”). .) And image data using infrared transmitted light (hereinafter referred to as “transmitted IR image”). The 1000-yen bill used here is a normal bill without damage (hereinafter referred to as "normal bill").

FIG. 9A is a diagram showing the variation of the amount of received light at a predetermined position of the reflected IR image of the normal bill shown in FIG.
2 and the variation mode on the line 91 is indicated by a curve 93.

Further, FIG. 8B is a diagram showing the variation of the received light amount at a predetermined position of the transmitted IR image of the normal bill shown in FIG. The curve 9 indicates the variation mode on the line 91, which is indicated by 94.
5 shows.

Referring to these figures, the reflection IR image and the transmission IR image have different variations on the same line, but since the influence of damage such as wrinkles is small, both are fine on the curve. The fluctuation is relatively small.

On the other hand, in FIGS. 10A and 10B, the amount of light received at a predetermined position in the reflected IR image and the transmitted IR image of a damaged banknote (hereinafter referred to as “damaged banknote”). The variation modes of the above are shown respectively.

Therefore, comparing the variation mode of the damaged banknote shown in the same figure with the variation mode of the normal banknote shown in FIG.
In both cases of 0 and 91, the damaged bill has more fine fluctuations on the curve than the normal bill.

For example, the curve 94 obtained by reading the line 90 of the transmitted IR image of the normal bill shown in FIG.
Comparing with the curve 104 obtained by reading the line 90 of the transmitted IR image of the damaged banknote shown in (b), the curve fluctuation is common in general, but the fluctuation of the curve 94 is comparatively gentle. The curve 104 changes sharply.

Therefore, the characteristic parameters of normal bills and damaged bills are different from each other, as will be described later.

FIG. 11 is a diagram showing characteristic parameters acquired from the transmission IR images of 40 damaged 1000-yen bills.

Here, the numbers on the horizontal axis in the figure indicate the numbers of the image data of the damaged banknotes used as samples, and the larger the degree of damage (stains and wrinkles) visually observed, the larger the number. ing. Also, the vertical axis is
The values of the characteristic parameters calculated by the determination unit 14 are shown.

Referring to the figure, it can be seen that the characteristic parameters calculated by the judgment unit 14 have a correlation with the degree of damage to the bill, and that the characteristic parameter increases as the degree of damage increases.

Further, each calculated value shown in the figure shows a variation when the same bill is conveyed ten times. In addition,
The number of tests is 10. As described above, the calculation result also varies for this characteristic parameter, but the variation is smaller than when the detection output is simply used.

FIG. 12 is a diagram showing the average value of the characteristic parameters of each sample shown in FIG. 11. In this figure, the discrimination reference value indicating whether or not the bill is a damaged bill is set to discriminate the damaged bill. You can

For example, if the cutoff value of the set characteristic parameter is set to 1.5, No. 1-No. 1
A sample of 0 is judged as a normal banknote and N
o. 11-No. The 40 samples will be judged as damaged banknotes.

The erroneous determination ratio ε1 in which an original normal bill is regarded as a damaged bill is 0%, and conversely, the erroneous determination ratio ε2 in which a damaged bill is mistakenly regarded as a normal bill is 5%.

This erroneous determination ratio ε1 is defined as ε1 = n / m. Note that n is the number of test times of the samples whose average value is smaller than the set value among the samples whose average value is smaller than the set value, and m is 10 times the number of banknote samples whose average value is smaller than the set value. is there.

The erroneous determination ratio ε2 is defined as ε2 = nn / mm. Note that this nn is the number of test times of samples whose average value is larger than the set value and whose feature value is smaller than the set value, and mm is 10 times the number of banknote samples whose average value is larger than the set value. is there.

The banknotes whose misjudgment ratio is suppressed to 5% or less do not have such a large degree of damage corresponding to the damage level 1 and the damage level 2 shown in FIG.

FIG. 13 is a diagram showing the characteristic parameters obtained from the reflection IR images of 40 damaged 1000-yen bills, and FIG. 14 is a diagram showing the average value of the characteristic parameters shown in FIG. The horizontal axis of FIG. 13 shows the samples arranged in the order of the degree of damage (wrinkles, folds, lines) visually observed.

As shown in FIG. 13, in this case, if the characteristic parameter is 1.3 or less, the damage level is 0 (Deg).
ree0), a characteristic parameter of 1.3 to 1.5 has a damage level 1 (Degree 1), and a characteristic parameter of 1.5 to 1.7 has a damage level of 2 (De
green2), and a characteristic parameter of 1.7 or more is a damage level 3 (Degree3).

Referring to FIGS. 13 and 14, even when the reflection IR image is used, the degree of damage such as wrinkles and folds and the value of the characteristic parameter are closely related to each other as in the case of using the transmission IR image. You can see that

FIG. 15 shows the transmitted IR image and the reflected IR image.
It is a figure which shows the damage discrimination | determination result using the characteristic parameter each acquired from the image.

As shown in the figure, here the vertical axis is the transmission I
By taking the value of the characteristic parameter acquired from the R image and the value of the characteristic parameter acquired from the reflection IR image on the horizontal axis, each sample is plotted at the corresponding position.

That is, when determining the degree of damage, the characteristic parameters based on the transmission IR image and the reflection IR image are calculated and plotted at a position corresponding to the calculation result of the dispersion diagram shown in FIG. The damage level can be determined.

Heretofore, an example of the damage determination of the 1000-yen bill has been described.

Next, one example of damage discrimination of US 1-dollar bills using the bill discriminating apparatus 10 shown in FIG. 1 will be specifically described.

FIG. 16 is a diagram showing image data of a US dollar bill read by using the data collection unit 11 shown in FIG. 1. In FIG. 16, a reflection G image, a reflection IR image and a transmission G image are shown. An image and a transmission IR image are shown. The US 1-dollar bill used here is a normal bill.

FIG. 17A is a diagram showing the variation of the amount of received light at a predetermined position of the reflected IR image of the normal bill shown in FIG. 16, and specifically, the variation mode on the line 170 is indicated by a curve 172. The curve 173 indicates the variation mode in the line 171.

FIG. 16B is a diagram showing the variation of the received light amount at a predetermined position of the transmitted IR image of the normal bill shown in FIG. 174, and the variation pattern on the line 171 is shown by a curve 175.

On the other hand, FIGS. 18 (a) and 18 (b) show variations of the amount of received light at predetermined positions in the reflected IR image and the transmitted IR image of the damaged banknote, respectively.

Therefore, comparing the variation mode of the damaged banknote shown in FIG. 17 with the variation mode of the normal banknote shown in FIG.
In the case of the R image, in both of the cases of the lines 170 and 171, the damaged banknote has more fine fluctuations on the curve than the normal banknote.

For example, a curve 172 obtained by reading the line 170 of the reflected IR image of the normal bill shown in FIG.
Line 1 of the reflected IR image of the damaged banknote shown in FIG.
Comparing the curve 182 obtained by reading 70 with the curve 182 generally shows that the curve fluctuations are common, but the curve 172 changes comparatively gently, while the curve 182 changes sharply.

On the other hand, in the case of the line 171 of the transmitted IR image, the damaged bill has less fine fluctuations on the curve than the normal bill. The reason for this is that paper quality is different from that of Japanese banknotes, and in the case of US $ 1 banknotes, wrinkles and the like decrease on the contrary as the user fatigues.

FIG. 19 is a diagram showing characteristic parameters obtained from transmission IR images of 30 damaged US $ 1 bills, and FIG. 20 is a diagram showing average values of the characteristic parameters.

Here, the numbers shown on the horizontal axis of the figure indicate the numbers of the image data of the damaged banknotes used as samples, and the larger the degree of damage visually observed, the larger the number. The vertical axis represents the value of the characteristic parameter calculated by the determination unit 14.

Referring to the figure, the characteristic parameter calculated by the judging unit 14 has a correlation with the degree of damage of the bill, but unlike the case of the 1000-yen bill, the characteristic parameter decreases as the degree of damage increases. Inversely proportional.

That is, in the case of US dollar bills, the printed pattern becomes unclear as it circulates, and the difference in color between the pattern and the base becomes smaller, so that the fluctuation of the output signal obtained from the bill becomes smoother. .

Therefore, in the case of a US dollar bill, the length of the curve becomes short, and the characteristic parameter of the transmitted IR image also becomes smaller as the bill becomes more damaged.

FIG. 21 is a diagram showing characteristic parameters obtained from reflected IR images of 30 damaged US dollar bills, and FIG. 22 is a diagram showing average values of the characteristic parameters shown in FIG. .

As can be seen from FIG. 21, unlike the case of the 1000-yen bill, the characteristic parameter does not show a clear fluctuation tendency. 7, No. 11 and No. The 14 characteristic parameters have larger values than the other samples, because the three bills have large wrinkles.

FIG. 23 shows the transmission IR image and the reflection IR.
It is a figure which shows the damage discrimination | determination result using the characteristic parameter each acquired from the image.

As shown in the figure, here the vertical axis is the transmission I
By taking the value of the characteristic parameter acquired from the R image and the value of the characteristic parameter acquired from the reflection IR image on the horizontal axis, each sample is plotted at the corresponding position.

That is, when determining the degree of damage, the characteristic parameters based on the transmission IR image and the reflection IR image are calculated and plotted at a position corresponding to the calculation result of the dispersion chart shown in FIG. The damage level can be determined.

FIG. 24 and FIG. 25 are diagrams showing the characteristic parameters obtained from the transmission G image, and as can be seen from these figures, the data variation tendency is almost the same as in the case of the transmission IR image also in this case. . Note that series 1 to series 6 indicate the number of tests.

Heretofore, an example of damage determination of a US dollar bill has been described.

Next, the result of comparison with a conventional method for discriminating dirty banknotes called the G / IR method will be described.

FIG. 26 is a diagram showing the result of calculation by the parameter G / IR method using the watermark portion of the 1000-yen bill as the judgment area. That is, in this conventional technique, the ratio of the sum of the amount of transmitted light of green (G) and the sum of the amount of infrared light (IR) is used as a parameter.

As shown in the figure, in this case, the damage level cannot be determined because the G / IR is located near 0.9, regardless of the degree of stain damage on the bill.

On the other hand, in the present embodiment, the characteristic parameters of the watermark portion of the 1000-yen bill are obtained as shown in FIG. 27, and further normalized as shown in FIG. Note that the banknotes for each sample are G / IR
The same method is used as the method.

Referring to FIG. 27 and FIG. 28, since the degree of damage of the bill is found based on the characteristic parameter, the above G
It can be seen that the discrimination technique using this characteristic parameter is more effective than the case of using the / IR method.

FIG. 29 is a diagram showing a result of applying the G / IR method to a US dollar bill (entire surface).
It can be seen that the / IR method is not suitable for dirt damage determination even for US dollar bills.

By the way, the damage discriminating technique based on the characteristic parameters used in the present embodiment does not require any special hardware, and the existing hardware (for example, shared with the denomination identifying sensor) is used. Can be realized.

When existing hardware is used, the program corresponding to the length calculation unit 12 and the determination unit 14 shown in FIG. 1 and the length of the curve of the waveform of the reference bill stored in the determination data storage unit 15 are used. A storage medium storing the data and the determination reference data is prepared, and the program and data are read from the storage medium and executed.

As described above, in the present embodiment,
Rather than focusing on the intensity itself of the transmitted or reflected light, long
Calculating section 12 obtains the length of the waveform of the transmitted or reflected light is,
Since the determination unit 14 is configured to determine the damaged banknote based on the characteristic parameter calculated based on this length,
The following effects can be obtained.

1) The influence of uncertain factors such as vertical vibration of bills due to high speed conveyance and folding of bills can be reduced, and the degree of damage to bills can be accurately determined.

2) It is possible to discriminate the damage of bills at low cost without adding special hardware.

3) A wide range of damages such as whole stains, plain stains, printed stains, discoloration, fatigue, folds or wrinkles can be set as the discrimination target.

4) It can be used to discriminate a wide range of banknotes such as Japanese 1000 yen bills and US dollar bills.

In the above embodiment, the case where γ = (sum of lengths in block) / (reference bill length) is used as the characteristic parameter is shown, but the present invention is not limited to this. Instead of, for example, γ '= |
It is also possible to use other characteristic parameters such as (sum of lengths in block) − (length of reference bill) |.

Further, in the above embodiment, the case where the image data is divided into 15 blocks has been shown, but the present invention is not limited to this, and the blocks can be set to a desired size and made variable. It is also possible to do so. In this case, if the block size is increased, stains, wrinkles, fatigue, etc. on the rough portion will be detected, and if the block size is decreased, fine stains, wrinkles, fatigue, etc. will be detected. In that case, the reference value is also set individually beforehand according to the number of blocks.

Further, in the above example, the length of the waveform is obtained based on the length between two points of the value of the received light data for each cell of the light receiving element array, but the present invention is not limited to this, and the value at a coarser pitch is used. The length of the waveform may be calculated based on On the contrary, a finer pitch sensor array can be used.

[0121]

As described in detail above, the present invention is
When the paper sheet conveyed at high speed passes through a predetermined position, light reception data is acquired from the paper sheet, the waveform of the predetermined light reception data is traced to calculate the length of the waveform, and the calculated length is calculated. Since the degree of damage to the paper sheet is determined based on the above, the following effects can be obtained.

1) The influence of uncertain factors such as vertical vibration of paper sheets and folding of paper sheets due to high-speed conveyance can be reduced, and the degree of damage to paper sheets can be accurately determined.

2) It is possible to discriminate the damage of paper sheets at low cost without adding special hardware.

3) It is possible to determine a wide range of damages such as whole stains, plain stains, print stains, discoloration, fatigue, folds or wrinkles.

4) It can be used to discriminate a wide range of paper sheets such as Japanese 1000 yen bills and US dollar bills.

Further, according to the present invention, since the length of the waveform is simply calculated by the length between the respective data, it is possible to quickly perform the discrimination processing.

Further, according to the present invention, the reference length of the waveform formed by the predetermined received light data of the reference paper sheet is stored, and the characteristic parameter is calculated from the calculated waveform length and the reference length,
Since the degree of damage to the paper sheet is determined based on the calculated characteristic parameter, the determination parameter can be used as a measure for damage determination.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration of a bill discriminating apparatus used in the present embodiment.

FIG. 2 is a diagram showing an example of waveforms of output signals of reflected light of a reference bill and a bill to be discriminated.

FIG. 3 is a diagram showing another example of the data collection unit shown in FIG. 1.

4 is a diagram showing an example of block division of image data stored in an image data storage unit shown in FIG.

5 is a diagram showing a processing concept of a length calculation unit shown in FIG.

FIG. 6 is a diagram showing an example of a discrimination diagram used by the determination unit shown in FIG. 1.

7 is a flowchart showing a processing procedure of the bill discriminating apparatus shown in FIG.

FIG. 8: 1 read using the data collection unit shown in FIG.
The photograph which shows the halftone image which displayed the image data of a 000 yen bill on the display.

FIG. 9 is a photograph showing a halftone image in which a variation of the amount of received light of a normal bill at a predetermined position is displayed on a display.

FIG. 10 is a photograph showing a halftone image in which the variation of the received light amount at a predetermined position of a damaged bill is displayed on the display.

FIG. 11: Transmission IR of 40 damaged 1000-yen bills
The figure which shows the characteristic parameter acquired from the image.

FIG. 12 is a diagram showing average values of characteristic parameters of each sample shown in FIG. 11.

FIG. 13: Reflection IR of 40 damaged 1000 yen bills
The figure which shows the characteristic parameter acquired from the image.

FIG. 14 is a diagram showing an average value of characteristic parameters of each sample shown in FIG. 13.

FIG. 15 is a diagram showing a damage determination result using feature parameters respectively acquired from a transmission IR image and a reflection IR image.

FIG. 16 is a photograph showing a halftone image in which image data of a US dollar bill read using the data collection unit shown in FIG. 1 is displayed on a display.

FIG. 17 is a photograph showing a halftone image in which the variation of the amount of received light at a predetermined position of the reflected IR image of a normal bill is displayed on the display.

FIG. 18 is a photograph showing a halftone image in which the variation of the received light amount at a predetermined position of the reflected IR image and the transmitted IR image of a damaged bill is displayed on the display.

FIG. 19: Transmission I of 30 damaged US $ 1 banknotes
The figure which shows the characteristic parameter acquired from R image.

20 is a diagram showing the average value of the characteristic parameters of each sample shown in FIG.

FIG. 21: Reflection I of 30 US dollar bills with damage I
The figure which shows the characteristic parameter acquired from R image.

22 is a diagram showing an average value of characteristic parameters of each sample shown in FIG. 21. FIG.

FIG. 23 is a diagram showing a damage determination result using feature parameters respectively acquired from a transmission IR image and a reflection IR image.

FIG. 24 is a diagram showing characteristic parameters acquired from a transmission G image.

FIG. 25 is a diagram showing average values of characteristic parameters acquired from a transmission G image.

FIG. 26 is a diagram showing a result of calculation by the parameter G / IR method using a watermark portion of a 1000-yen bill as a determination area.

FIG. 27 is a diagram in which characteristic parameters of the present embodiment are obtained for a watermark portion of a 1000-yen bill.

28 is a diagram in which the characteristic parameters shown in FIG. 27 are normalized.

FIG. 29 is a diagram showing a result of applying the G / IR method to a US dollar bill.

FIG. 30 is a diagram showing an influence of a fold of a bill on a diffusion distribution of transmitted light.

[Explanation of symbols]

10 bill discriminating apparatus 11 data collecting unit 11a bill 11b roller 11c transmissive LED array 11d reflective LED array 11e light receiving element array 11f amplifier 11g A / D converter 12 length calculating unit 13 image data storage unit 14 discriminating unit 15 discrimination data Memory

Front page continuation (58) Fields surveyed (Int.Cl. ⁷ , DB name) G07D ^7/ 00-7/20 G01B 11/30 G01N 21/89

Claims (7)

(57) [Claims] 1. A method for determining a damage of a paper sheet, which optically reads a paper sheet conveyed at a high speed and judges the damage of the paper sheet based on the read signal, wherein the paper sheet has a predetermined position. When passing, light is emitted, at least one of the transmitted light and the reflected light of the paper sheet is acquired as light reception data, and the length of the locus of the waveform formed by the acquired light reception data is calculated. A method for determining damage to a paper sheet, which comprises determining the degree of damage to the paper sheet based on the length. 2. A paper sheet damage discriminating apparatus for optically reading a paper sheet conveyed at high speed and discriminating the damage of the paper sheet based on the read signal, and irradiating the paper sheet with light. Irradiation means and data acquisition for acquiring at least one of transmitted light and reflected light of the light emitted by the irradiation means as light reception data when the paper leaves pass through a predetermined position. Means, a length calculation means for calculating the length of the locus of the waveform formed by the light reception data acquired by the data acquisition means, and a damage degree of the paper sheet based on the length calculated by the length calculation means. An apparatus for discriminating damage to paper sheets, comprising: a discriminating means for discriminating. 3. The paper sheet damage determination device according to claim 2, wherein the length calculation means calculates the length of the locus of the waveform from the value of each received light data acquired at a predetermined interval. . 4. The discriminating means stores, as a reference length, the length of a locus of a waveform formed by the received light data of the reference paper sheet, the length of the locus of the waveform calculated by the length calculating means, and the A feature parameter calculating unit that calculates a feature parameter from the reference length stored in the storage unit, and a determining unit that determines the degree of damage to the paper sheet based on the feature parameter calculated by the feature parameter calculating unit. The paper sheet damage determination device according to claim 2. 5. The data acquisition unit acquires light reception data divided into a predetermined number of blocks, and the length calculation unit calculates a length of a locus of a waveform formed by each light reception data acquired by the data acquisition unit. 3. The paper sheet damage determination device according to claim 2, wherein each of the calculated lengths is added to all the blocks. 6. The low-frequency component removing unit for removing low-frequency components from the received light data acquired by the data acquisition unit, wherein the length calculation unit is a waveform formed by the received light data acquired by the data acquisition unit. 3. The paper sheet damage discriminating apparatus according to claim 2, wherein the length of the locus is calculated after the low frequency component of the received light data is removed by the low frequency component removing means. 7. A recording medium used by a paper sheet damage determination device that optically reads a paper sheet conveyed at high speed and determines the damage of the paper sheet based on the read signal. When at least one of the transmitted light and the reflected light of the light emitted when the paper passes through the predetermined position is received as the received light data, the length of the locus of the waveform formed by the acquired received light data And a program for determining the degree of damage of the paper sheet based on the calculated length.