JP2641079B2 - Bill validator - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a banknote discriminating apparatus for discriminating the denomination and authenticity of banknotes in an automatic cash dispensing and dispensing apparatus installed in a financial institution or the like after detecting the direction of banknote transport. Related to the device.

[0002]

2. Description of the Related Art In a cash dispensing and dispensing apparatus, a bill discriminating apparatus is provided for discriminating a bill deposited by a customer, its denomination, authenticity and the like, and storing the bill in a safe. Also, a bill discriminating device is provided in order to check in advance the denomination and authenticity of the bill even when dispensing the bill to the customer from the automatic cash dispensing and dispensing device. This kind of bill discriminating apparatus is provided with a sensor unit for detecting the intensity of reflected light or transmitted light of a printing pattern of a bill facing a bill to be discriminated. In addition, a sensor unit including a magnetic sensor is provided for detecting the magnetic characteristics of the magnetic ink used for printing the bill to be discriminated.

FIG. 2 is an explanatory diagram of a conventional general bill reading state when the above-described sensor unit is used. In the conventional bill discriminating apparatus, when the bill 1 to be discriminated is conveyed in the direction of arrow 2, a sensor unit (not shown) disposed opposite thereto scans the bill surface in the direction of arrow 3, and the bill 1 to be discriminated. Of the specific area 4 is read. For example, in the case of detecting reflected light, an electric signal that changes in strength is obtained from the sensor unit according to the print pattern of the banknote 1 to be determined. This signal takes a characteristic value corresponding to the specific area 4 determined by the denomination and its transport direction.

FIG. 3 shows a block diagram of a bill discriminating apparatus for processing such a signal. The device shown in FIG.
And an analog / digital conversion circuit (A / D) 11 for receiving the output, a buffer memory 12, a comparison circuit 1
3, an arithmetic processing circuit 14, a discrimination circuit 15, a sample timing pulse generator 16, an address counter 17, and a standard pattern memory 18.

[0005] The output of the sensor unit 10 is a continuous analog signal that changes in strength as the bill to be discriminated is conveyed. This is converted into a digital signal in the analog / digital conversion circuit 11 according to the output timing of the sample timing pulse generator 16. as a result,
Regarding one bill to be discriminated, a large number of digital gradation data (discriminated signals) corresponding to the read electric signal are
The data is stored in the buffer memory 12. This is called a discriminated data signal. The standard pattern memory 18 stores a standard pattern signal to be compared with the discriminated data signal.

FIG. 4 is an explanatory view of a bill discriminating method using a conventional bill discriminating apparatus. This graph shows the signal level on the vertical axis and the sampling points on the horizontal axis. The solid line is the standard pattern signal, and the signal above is the upper limit of 19
A, the lower signal indicates the lower limit 19B. Assuming that this is a standard pattern signal for a specific direction of a ten-yen note, for example, when reading a ten-yen note of the to-be-determined bill in the same direction, a to-be-identified data signal 5 indicated by a broken line in FIG. Can be

The address counter 17 shown in FIG. 3 sequentially calls the discriminated data signal stored in the buffer memory 12 and the standard pattern signal stored in the standard pattern memory 18 and outputs them to the comparison circuit 13. Comparison circuit 1
In step 3, whether the data signal to be determined is between the upper limit value and the lower limit value of the standard pattern signal is compared, and the comparison result is output to the arithmetic processing circuit 14. The arithmetic processing circuit 14 performs a constant calculation based on the comparison results performed for a large number of sampling points, and sends data for determining the denomination, transport direction, authenticity, and the like of the discriminated bill to the discrimination circuit 15. Output to The determination circuit 15 outputs the determination result to an external circuit (not shown).

The standard pattern signal to be stored in the standard pattern memory 18 as described above is usually generated as follows. First, the specific area 4 (refer to) of a plurality of genuine bills is read by the sensor unit 10 shown in FIG. 3, and signals corresponding to the print patterns are collected. Then, the maximum value and the minimum value of the signal read at each sampling point are obtained. If this is plotted in the manner shown in FIG. 4, the upper limit 19A and the lower limit 19B of the standard pattern signal can be specified.

As can be seen from the graph of FIG. 4, the difference between the upper limit value and the lower limit value varies at each sampling point depending on the contents of the bill printing pattern. This is because the range of variation is different.

[0010] Even in the case of the same bill, the specific area 4 is different between the case where the right side of the front surface is confronted with the sensor and the case where the left side of the front side is conveyed toward the sensor. The standard pattern signal is different. Needless to say, the same applies to the case of the back side. Therefore, there are three types of Banknotes of Japan: "10,000 yen", "5,000 yen", and "1,000 yen". Therefore, the front and back sides and the transport direction (hereinafter referred to as denomination direction) are considered. Then, four types of standard pattern signals are set for each denomination.

Further, since the bill transport area is set larger than the external dimensions of the bill, a travel standard pattern signal for each travel area of each bill is set in consideration of the travel state of the bill.

[0012]

However, as described above, in order to select the optimum specific area 4 (see FIG. 2) for each bill, a pattern or the like must be set in advance in the direction of the denomination of each bill and in each running area. Since it is necessary to collect and analyze an enormous amount of data, it takes a lot of trouble and time to select the optimal specific area 4 when the pattern of the banknote changes, and also, the capacity is large. A data storage memory is required, and the device becomes very expensive. Also, if there is a markedly soiled portion or a missing portion in a specific area of the bill to be discriminated, or if there is a detection error or the like in the apparatus itself, there is a possibility that a genuine bill is erroneously distinguished from a fake bill or vice versa.

Furthermore, if the specific area 4 is set using only genuine bills with little dirt, the criterion for discrimination becomes strict, and genuine bills are discriminated as counterfeit bills and the number of rejected bills increases.

The present invention has been made in view of the above points, and a bill discriminating apparatus capable of discriminating the denomination and authenticity of a bill to be discriminated with high accuracy without collecting and analyzing a huge amount of data. The purpose is to provide.

[0015]

According to a first aspect of the present invention, there is provided a sensor unit for reading all print patterns of a bill to be discriminated, and a data storage for storing discriminated data including bill tone data obtained from the sensor unit. In the bill discriminating apparatus including the memory, the bill gradation data is searched from the discriminated data, and the bill gradation data selecting means to be taken out, and the data dividing means for dividing the selected bill gradation data into a plurality of block areas,
Division gray scale data of each block for each region by the data processing each <br/> Histogram de - Histogram de generates the data - and data generating means, each histogram data for each block area,
Preset standard histogram data and ratio for each banknote
And a bill determining means for performing a comparison operation to determine a bill.

According to a second aspect of the present invention, there is provided a bill discriminating apparatus comprising: a sensor unit for reading all printed patterns of a bill to be discriminated; and a data storage memory for storing discriminated data including bill tone data obtained from the sensor unit. In the apparatus, the banknote gradation data is searched for from the discriminated data, and the banknote gradation data selection unit is extracted. The data division unit divides the selected banknote gradation data into a plurality of block regions. de the scale data integration process - a data calculating means, the product
The operation data of each block for each region was separated treated, preset the weighted data of each bill - and determining denomination direction determination unit denominations and the reading direction of the bill by weighting processing based on the data, the block region Histogram data generating means for processing each divided gradation data to generate each histogram data, and denomination and reading of each histogram data
The present invention is characterized in that there is provided a bill authenticity judging means for judging the authenticity of the bill by performing a comparison operation with standard histogram data set in advance for the bill whose direction has been judged.

[0017]

According to the first aspect, the discriminated data read by the sensor unit and stored in the data storage memory includes banknote gradation data based on all print patterns on the front and back surfaces of the discriminated banknote and reading of non-banknotes. Contains data . The banknote gradation data selecting means searches and extracts only the banknote gradation data from the discriminated data. The data dividing means divides the banknote gradation data into a plurality of block areas. The histogram data generation means performs data processing on the divided gradation data for each block area to generate predetermined histogram data. The bill determining means each histogram data, and comparison operations with each standard histogram data preset for each bill, of the determination banknote denomination to determine authenticity.
As a result, the denomination and authenticity can be accurately determined based on the entire printing pattern of the banknote to be determined, and since the histogram data having a small data amount is used, the memory capacity can be reduced.

Further, in the second invention, the data calculation means integrates the divided gradation data for each block area.
And create integral data. The denomination direction determining means performs a weighting process on the integral data for each block area based on the weighting data, and determines the denomination of the banknote and the reading direction thereof. Bill authenticity judgment means, and the standard histogram data denominations and the reading direction of the determined object discriminated bill, and comparison operation and the histogram data generated by the histogram data generating means, determine the authenticity of the discrimination banknote I do. Therefore, after specifying the denomination and the reading direction,
Can be set strictly to some extent,
Significantly improved.

[0019]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 5 is a block diagram of the bill discriminating apparatus according to the present invention. In FIG. 5, reference numeral 20 denotes a sensor unit that reads a one-dimensional image. Perform a read operation. An analog (gradation) signal output from the sensor section 20 is amplified to an optimum value by an amplifier section 22, and an analog / digital converter (A / D) 23
Is converted into a digital signal (data to be determined). The discriminated data is temporarily stored in the register 24 and further sent to the data select circuit 39. The discriminated data sent to the data select circuit 39 is stored in the data storage memory 26 at the address position designated by the address select circuit 29. The data storage memory 26 is S
-It is composed of RAM. Address select circuit 2
9 is connected to a sensor select circuit 25. The sensor select circuit 25 outputs an address corresponding to the sensor selected by the sensor section 20 to the address select circuit 29. The sensor select circuit 25 and the address select circuit 29 are connected to the CPU 28 via a data bus 30 and an address bus 31, respectively. CPU28
Is connected to a control data storage memory 32 via these buses 30 and 31. This control data storage memory 3
Reference numeral 2 denotes a ROM, which stores a program for performing a control operation of the CPU 28 and an arithmetic processing described later.

FIG. 6 is a configuration diagram of the sensor section 20 described above. That is, the sensor unit 20 includes an LED 33 serving as a light emitting element and a light receiving transistor 34 serving as a light receiving element disposed on both sides of a carry-in entrance of the transport path of the bill P to be determined.
And each pair of LED 33 and light receiving transistor 34
Constitutes transmission type position sensors 46a and 46b, respectively, and outputs a reading start signal to the CPU 28 (see FIG. 5).
The sensor unit 20 has an LED array 35 as a light source. The emitted light is reflected by the bill P to be discriminated and is incident on the cell hook lens array 36. The incident light received by the cell hook lens array 36 is received by the one-dimensional image sensor 37. The amplifier 22 is connected to the output side of the image sensor 37 as shown in FIG.

A transport roller 38 for transporting the banknote P to be discriminated is arranged below the sensor unit 20. The peripheral surface of the transport roller 38 is formed black (system) in order to minimize the reflectance of the light emitted from the LED array 35.

FIG. 1 is a functional block diagram of the CPU 28. In FIG. 1, reference numeral 40 denotes bill tone data selection means. This selection means 40 searches only the banknote gradation data from the discriminated data obtained from the sensor unit 20,
Take out. That is, the determination target data read by the sensor unit 20 includes unnecessary data read by the reflected light from the transport roller 38 and banknote gradation data read from the print pattern of the determination target banknote P. Therefore, the banknote gradation data selecting unit 40 searches the data storage memory 26 based on the digital slice value as described later, and selects the external address where the banknote gradation data is stored.

Reference numeral 41 denotes data dividing means. The data dividing unit 41 performs an arithmetic process based on the external address selected by the banknote gradation data selecting unit 40, as described later, to evenly divide the banknote gradation data into a plurality of block areas. That is, in the present embodiment, one side of the banknote P to be discriminated is equally divided into eight block areas, and the boundary of each block area is calculated as a boundary address.

Reference numeral 42 denotes a histogram data generating means. The data generating means 42 extracts the divided grayscale data for each block area based on the boundary address, performs data processing on the divided grayscale data of each block area as described later, and generates histogram data. The histogram data of each block area is stored in the internal memory 45.

Further, reference numeral 43 denotes a bill judgment means.
The bill determination means 43 reads out the standard histogram data in the corresponding block area for each bill from the standard histogram data storage section 44 of the control data storage memory 32, and the standard histogram data and the histogram data generation means 42 for each block area. The difference from each of the generated histogram data is obtained, and the absolute value is added. Then, it is determined that the bill with the smallest added value is the denomination of the bill P to be determined. However, when the minimum added value is equal to or higher than the set level, it is determined to be a counterfeit note.

Next, the operation of the bill discriminating apparatus according to the present invention will be described with reference to the operation flowchart of FIG. First, FIG.
When the banknote P to be discriminated is carried into the transport path and the light reception of one light receiving transistor 34 is interrupted, a detection signal is sent from the light receiving transistor 34 to the CPU 28. When the CPU 28 receives this detection signal, the sensor unit 2
0 and the transport roller 38 are operated. Accordingly, the one-sided print pattern is read by the one-dimensional image sensor 37 along with the transport of the bill P to be discriminated.
From 0, an analog discriminated signal is output (step 1).

The signal to be discriminated is amplified to an optimum value by the amplifier 22, and is converted into a digital (data to be discriminated) signal by the A / D converter 23. This data to be determined is stored in the register 24 as it is.

On the other hand, the sensor select circuit 25
Each time a pixel of the image sensor 37 of the sensor section 20 is selected under the control of 8, the corresponding address is output to the address select circuit 29. Thus, the address select circuit 29 directly accesses the data storage memory 26 and specifies a storage address. Therefore, since the CPU 28 does not access the data storage memory 26, the data to be determined of the register 24 can be stored in the data storage memory 26 via the data select circuit 39 in a short time.

If the number of pixels in the main scanning direction of the image sensor 37 of the sensor section 20 is 256, the data to be determined is stored in the data storage memory 26 at addresses XX00 to XXFF as shown in FIG. . Assuming that the number of one scan of the banknote P to be determined in the sub-scanning direction is 256, the data to be determined is stored in the data storage memory 26 at the address 00.
XX to FFXX. That is, the discriminated data on one side of the discriminated bill P is stored in the data storage memory 26 at addresses 0000 to FFFF. In this case, if the gradation data per pixel is 1 byte, the number of data to be determined in the entire printing pattern on one side of the banknote P to be determined and the pattern on the peripheral surface of the conveying roller 38 is 64 Kbytes (see FIG. 8).

When all the data to be discriminated are stored in the data storage memory 26, the banknote gradation data selecting means 4 of the CPU 28
0 accesses the control data storage memory 32 and reads the selection control program. Thereby, this selecting means 40
Selects the external address of the banknote gradation data included in the discriminated data based on the digital slice value. In this embodiment, the digital slice value is set to 40H (hexadecimal hexadecimal). That is, the end portion (edge portion) of the banknote P to be discriminated is not printed and the reflected light is large, whereas the peripheral surface of the transport roller 38 is black (system), so the reflected light is small. Accordingly, when the digital slice value is set to 40H, if the digital slice value is larger than this value, it can be determined that the bill P is the end of the bill P.

Now, the banknote gradation data selecting means 40 sets the left light receiving transistor 34 in step 2 of FIG.
Output the detection signal first, it is determined that the banknote P to be discriminated is skewed and conveyed so that the left side comes first, and the address select circuit 29 is controlled in step 3 to control the addresses FF00, FE00,. 0000, FF01, FE01... Are sequentially output to the data storage memory 26. As a result, each pixel data is sent from the data storage memory 26 via the data select circuit 39 and the data bus 30 in the order of these read addresses. The data selecting means 40 compares each of the pixel data with the digital slice value 40H, and detects an address of the pixel data exceeding the slice value. In the present embodiment, as shown in FIG. 8, the address of FC01 (pixel data indicated by CA) is detected, and this corresponds to the left end of the bill P to be discriminated.

Next, the banknote gradation data selecting means 40 sequentially outputs the addresses 00FF, 01FF, 02FF,... FFFF, 00FE, 01FE to the data storage memory 26, and converts the pixel data sent in the order of these addresses into digital slices. Compare with value 40H. In this case, the address of the pixel data exceeding the slice value is 04FD (pixel data indicated by C8), which corresponds to the right end of the bill P to be discriminated.

Next, the banknote gradation data selecting means 40 outputs addresses 0000, 0001, 0002,... 00FF, 0100, 0101... To the data storage memory 26 in order, and converts the pixel data sent in order of these addresses into digital slice values. Compare with 40H. In this case, the address of the pixel data exceeding the slice value is 0302 (pixel data indicated by BA), which corresponds to the leading end of the bill P to be discriminated.

Finally, the banknote gradation data selecting means 40 sequentially outputs the addresses FFFF, FFFE, FFFD,... FF00, FEFF, FEFE. Compare with value 40H. The address of the pixel data exceeding the slice value in this case is FDFC (pixel data indicated by BF), which corresponds to the trailing end of the discriminated bill P.

Based on the external addresses of FC01, 04FD, 0302, and FDFC, the storage address of the banknote gradation data corresponding to the print pattern of the banknote P to be determined can be determined. These external addresses are supplied to the data dividing means 41. The data division means 41 reads the division control program from the control data storage memory 32 in step 4.
That is, for example, the data dividing means 41 defines the address 0302 as “l”, the 04FD as “m”, the FC01 as “p”, and the FDFC as “q”, and determines the banknote gradation data as two-dimensional x and y. I do. FIG. 9 shows an example in which the banknote gradation data is represented in two dimensions of x and y based on the external address.

Next, since each address "l", "m", "p", "q" is represented by 2 bytes, the data dividing means 41 divides these addresses into x-axis and y-axis, The x-axis corresponds to the sub-scanning direction, and the y-axis corresponds to the main scanning direction. For example, since the address “l” is 0302, l (H) = 03 = y
Axis, l (l) = 02 = x axis. Therefore, the address "m" is 04
Since FD, m (h) = 04 = y-axis, m (l) = FD = x-axis, and the address “p” is FC01, so p (h) = FC = y-axis, p (l) = 01 = x-axis Since the address "q" is FDFC, q (h) = FD = y-axis and q (l) = FC = x
Axis.

Subsequently, the data dividing means 41 calculates a boundary address of each area by the following equation in order to divide the banknote gradation data into block areas. That is, the boundary address Axy when dividing the address “l” and “m” into four is: Ax = {X (m (l) −l (l) ) / 4} + l (l) (1) Ay = { Y (m (h) −l (h) ) / 4｝ + l (h) (2)

The boundary address Cxy when dividing the addresses "p" and "q" into four is: Cx = {X (q (l) -p (l) ) / 4} + p (l) (3) Cy = {Y (q (h) -p (h) ) / 4} + p (h) (4)

Further, addresses "l", "p" and "m",
The boundary address Bxy when "q" is divided into two is: Bx = (Ax-Cx) / 2 + Cx = ( Ax + Cx ) / 2 (5) By = (Cy-Ay) / 2 + Ay = ( Ay + Cy ) / 2 … (6)

When the banknote gradation data is divided into eight by the above equations (1) to (6), it can be divided into block areas Z1 to Z8 as shown in FIG. Therefore, the discriminated bill P
Is divided into 16 block areas because it has both front and back sides.

As described above, when the banknote gradation data is divided into eight block areas, the histogram data generation means 42
Performs data processing on the divided gradation data of each of the block areas Z1 to Z8 to generate histogram data (step 5). That is, the histogram data generating means 42 reads the data processing program from the control data storage memory 32, first sends out all the addresses included in the block area Z1 to the address selection circuit 29, and outputs the pixels corresponding to the addresses of the data storage memory 26. Data, that is, divided gradation data is taken in.

The number of samples, that is, the number of pixels is added to each of the same-level divided gray-scale data using the level of each of the divided gray-scale data in the block area Z1 as a population. Is generated as histogram data. In this embodiment, the divided gradation data has a one-byte structure, and can be represented as a population of 256 (number of data). Therefore, the number of pixels corresponding to all addresses of the block area Z1 is distributed for each of the 256 divided grayscale data levels, and constitutes each histogram data. That is, each histogram data of the block area Z1 is represented by H1, and the level of the divided gradation data is represented by U (1, 2, 2,.
3,... 256), the entire histogram data Z1 (U) of the block area Z1 is expressed as follows. Zn (U) = ΣHn (U: 1, 2, 3,... 256) (7)

FIG. 10 is a waveform diagram showing the frequency distribution of the histogram data of the block area Z1. In FIG. 10, the horizontal axis represents the level (0 to 256) of the divided gradation data, and the vertical axis represents the number of pixels. Hereinafter, similarly, the histogram data generating means 42 generates each histogram data Hn (U) of the other block areas Z2 to Z8. All the histogram data Z1 of each of the obtained block areas Z1 to Z8
(U) to Z8 (U) are stored in the internal memory 45.

When the generation of the histogram data for each of the block areas Z1 to Z8 is completed, the bill judging means 43 takes in the standard histogram data from the standard histogram data storage section 44 of the control data storage memory 32. This standard histogram data includes standard multiple "10,000 yen vouchers"
It is obtained by dividing the front and back of “5,000 yen ticket” and “1,000 yen ticket” into 8 block areas in the same manner as described above, and generating the histogram data for each of two loading directions, It is stored in the standard histogram data storage unit 44 in advance. Therefore, assuming that the pattern of the standard histogram data is Skn, k = 12 depending on the front and back sides of the three denominations and the two loading directions.
And n = 1 to 8 because n is divided into eight block areas Z1 to Z8.

In order to execute step 6, the banknote judgment means 43 next executes each corresponding pattern in the first pattern S1n of standard histogram data (for example, a pattern obtained by loading a table of ten thousand yen from one direction). The histogram data S11 (U) to S18 (U) of the block area and the internal memory 4
5, each block area Z1 to Z8 of the discriminated bill P stored in
Histogram data Z1 (U) to Z8 (U)
Is calculated after being converted into binary data for each histogram data, and the absolute value of the difference is obtained. That is, all the generated histogram data Z1 (U) to Z8
Each of the histogram data of (U) is used as binarized data and D
1 to D256, and each histogram data T11 of all the histogram data S11 (U) to S18 (U) of the pattern S1n.
Absolute value | T1 -D1 |, | T2 -D of difference from 1 to T256
2 | ... | S256-D256 |

Next, the banknote judgment means 43 determines the absolute value of the difference | T1-D1 |, | T2 for each of the block areas Z1 to Z8.
-D2 | ... | S256-D256 | is added. That is, assuming that the total added value of each block area is Zk (n) and the absolute value of the difference is R (U), Zk (n) = Σ (U = 1, 2, 3,... 256) | R (U ) | (8)

FIG. 11 shows the calculation result of the absolute value. In the block area Z1 for the pattern S1n, the total added value Z1 of the absolute value of the difference between the histogram data D1 to D256 is obtained.
(1) is 598H, total added value Z1 of the same Z2 (2) is 14C
H, Z3 total addition value Z1 (3) is 94DH, Z4 total addition value Z1 (4) is 29FH, Z5 total addition value Z1
(5) is 356 H, the total sum Z1 of Z6 is (6) is BDOH,
The total addition value Z1 (7) of Z7 is D5H, and the total addition value Z1 (8) of Z8 is 49AH.

Hereinafter, similarly, the banknote determination means 43 determines whether each of the block areas Z in the other patterns S2n, S3n,.
Histogram data D1 to D25 for each of 1 to Z8
The absolute value of the difference with respect to 6 is obtained, and the total added value Zk (n) for each block area is calculated according to equation (8) for each pattern.

Subsequently, the bill judging means 43 checks each block area Z1 to Z12 in each of the twelve patterns S1n to S12n.
The total addition value ZK (1) to ZK (8) of Z8 is added. That is, assuming that the judgment value is DK, DK = Σ (n = 1, 2,... 8) Zk (n) (9) Therefore, the judgment value of the pattern S1n is 2765H and the pattern S2n
The judgment value of 2FDOH, ... the judgment value of pattern S12n is 26E2H
Becomes

Finally, the bill judgment means 43 determines whether each of the patterns S
The respective judgment values Dk of 1n to S12n are compared, and the banknote having the pattern of the smallest judgment value Dk is judged as the banknote P to be discriminated (step 7). That is, in FIG. 11, the S3n pattern indicated by D3 = 27AH is the smallest judgment value, so that the banknote indicated by the S3n pattern is judged to be the corresponding denomination.
In the above embodiment, when the smallest determination value Dk is equal to or larger than the set range, the bill determining means 43 determines that the bill P to be determined is a fake bill.

Next, a bill discriminating apparatus according to the second invention will be described. FIGS. 12 and 13 are a functional block diagram and a block diagram of the overall configuration of the CPU of the bill discriminating apparatus of the second invention. First, in FIG. 13, a buffer memory 27 is connected to the output side of the register 24 instead of the data select circuit 39 (see FIG. 5). The buffer memory 27 is used for storing the data to be determined stored in the register 24, and sequentially outputs the data to be determined to a predetermined address position of the data storage memory 26 as described above.
In FIG. 13, since the other configuration is the same as that of the apparatus of FIG. 5, the same reference numerals are given and the description is omitted.

Next, in FIG. 12, the data calculation means 49 extracts the divided gradation data for each block area based on the boundary address, and calculates the integral value of the divided gradation data for each block area. The integrated value of each block area is stored in the internal memory 45. The denomination direction judging means 50 reads out the weighting data set in advance for each bill from the weighting data storage section 47 of the control data storage memory 32, integrates this weighting data with the above-mentioned integrated value calculated for each block area, Add the integrated value for each area.

By the way, the weighting data is obtained by dividing the front and back of a standard plurality of "10,000 yen", "5,000 yen", and "1,000 yen" into block areas, respectively. In particular, it has a data configuration in which divided gradation data of a large level is extracted, and a large numerical value is set in the extracted portion.
Then, the denomination direction determining means 50 selects the largest added value of the integrated value, and thereby determines the denomination of the banknote P to be discriminated, its front and back, and the transport direction (the denomination direction).

The bill authenticity judging means 51 performs predetermined arithmetic processing on the histogram data of each block area generated by the histogram data generating means 42 to compress the data, and the compressed histogram data and the standard compressed histogram data storage 44 ', And the standard compressed histogram data of the banknote for which the denomination direction has been determined is processed as described later.
If the result of this calculation is a numerical value in the reference data in the reference data storage unit 48, the discrimination target bill P is determined to be genuine.

Next, the operation of the bill discriminating apparatus of the second invention will be described with reference to the operation flowchart of FIG. First, in the same manner as described above, the print pattern of the
0 (step 1), the analog signal to be determined is digitally converted by the A / D converter 23, and the data is stored in the data storage memory 2 via the register 24 and the buffer memory 27.
6 stores the data to be determined. Next, the banknote tone data selecting means 40 determines the outer shape address based on the digital slice value in steps 2 and 3 in the same manner as described above. That is, as shown in FIG.
D, 0302, FDFC is determined. Next, the data dividing means 4
1 determines the banknote gradation data as two-dimensional X and Y based on the outer shape address in the same manner as described above, and determines the outer shape address as “l”, “m”, “p”, “p” as shown in FIG. q "
Is defined.

Subsequently, the data dividing means 41 calculates the boundary address of each area by the following equation in order to divide the banknote gradation data into block areas. That is, in this embodiment, as shown in FIG. 15, the address “l” and “m” are divided into eight, and the boundary address Axy is Ax = ｛X (m (l) −l (l) ) / 8 ｝ + L (l) (10) Ay = {Y (m (h) -l (h) ) / 8} + l (h) (11)

The boundary address Cxy for dividing the addresses “p” and “q” into eight is: Cx = {X (q (l) −p (l) ) / 8} + p (l) (12) Cy = {Y (q (h) -p (h) ) / 8} + p (h) (13)

Further, addresses "l", "p" and "m",
The boundary address Bxy when “q” is divided into two is: Bx = (Ax−Cx) / 2 + Cx = ( Ax + Cx ) / 2 (14) By = (Cy−Ay) / 2 + Ay = ( Ay + Cy ) / 2 … (15)

When the banknote gradation data is divided into 16 by the above equations (10) to (15), Z1 to Z16 are obtained as shown in FIG.
(Step 4). Accordingly, the bill P to be discriminated has both front and back sides, and is therefore divided into 32 block areas.

As described above, when the banknote gradation data is divided into 16 block areas, the data calculating means 49 calculates the integral value of the divided gradation data of each of the block areas Z1 to Z16 (step 5). That is, the data operation means 49 reads the operation program from the control data storage memory 32, first sends out all the addresses included in the block area Z1 to the address selection circuit 29, and
, Ie, the divided gradation data.

Then, the integral value S1 of each divided gradation data in the block area Z1 is calculated. That is, assuming that the number of pixels in the block area Z1 is t and the number of gradation data of each pixel is e, it can be calculated by Sn = Σe (x) (16). Hereinafter, similarly, the data calculating means 49 calculates the integrated values S2 of the other block areas Z2 to Z16.
S16 are calculated based on the above equation (16). The obtained integral values S1 to S16 of the respective block areas Z1 to Z16 are stored in the internal memory 45.

When the calculation of the integral value is completed, the denomination direction judging means 50 takes in the weight data from the weight data storage section 47 of the control data storage memory 32. If the pattern is Skn, k = 12 depending on the front and back sides of the three denominations and two loading directions, and n is 16
Is divided into the block areas Z1 to Z16 of
= 1 to 16. Next, the denomination direction determining means 50 determines the weighting data W11 to W116 of each block area in the first pattern S1n of the weighting data (for example, a pattern obtained by loading a table of ten thousand yen coins from one direction). Each block area Z1 to Z of the discriminated bill P stored in the memory 45
The product of the integrals S1 to S16 at 16 is calculated after being converted into binary data, and these integrated values are added. That is,
Assuming that the weighting data is Wkn, the added value Gk of the integrated value with the integrated value Sn is as follows: Gk = Σ (n = 1, 2, 16) Sn × Wkn (17)

FIG. 16 shows the calculation result of the integrated value and the added value. The integrated value in the pattern S1n is 16 when n = 1.
AH, 2D5H when n = 2, 3C5H when n = 3, 140 when n = 4
H... 278H when n = 15, FD when n = 16, and the sum G1 of these integrated values is 2AD5H. Hereinafter, similarly, the denomination direction judging means 50 sets the other patterns S2n, S3n.
The integrated value and the added value in S12n are calculated by equation (17). Subsequently, the denomination direction judging means 50 checks each pattern S1n
The banknotes having the pattern of the largest addition value are determined to be the banknotes P to be discriminated. That is, FIG.
In S6, since the S3n pattern represented by G3 = 52A3H is the largest addition value, it is determined that the bill indicated by the S3n pattern is the corresponding denomination and direction (step 7).

As described above, when the denomination direction is determined, the histogram data generating means 42 processes the divided gradation data of each of the block areas Z1 to Z16 as described above, and generates the histogram data H (U = U All histogram data Z1 (U) to Z16 (U) consisting of (1, 2,..., 256) are generated (step 8). These histogram data Z1 (U)
ＺZ16 (U) are stored in the internal memory 45. When the generation of the histogram data is completed, the bill authenticity determination unit 51
Reads all the histogram data Z1 (U) to Z16 (U) of each block area Z1 to Z16 from the internal memory 45,
The following arithmetic processing is performed (step 9). That is, assuming that each histogram data is Hn (U), the minimum value of the histogram data is Hn (min), and the maximum value of the histogram data is Hn (max), the normalized operation data En (U) is as follows. . En (U) = {Hn (U) −Hn (min)} × {Hn (max) −Hn (min)} (18)

The bill validity judging means 51 checks each block area Z
When the normalized operation data En (U) of 1 to Z16 is calculated,
Subsequently, data compression of the operation data En (U) is performed as in the following equation. That is, assuming that normalized operation data obtained by compressing En (U) is Ln (hereinafter, referred to as compressed histogram data), Ln (X) = M / 4 (19). In this equation (19), M indicates the maximum value from which the bill tone data can be read by the device, and the A / D converter 23
Is determined by the conversion function. In this embodiment, since the A / D converter 23 of 8-bit processing is used, M = 256.
Thus, the normalized operation data En (U) is data-compressed to 1/4.

The solid line in FIG. 17 indicates the normalized operation data En (U
= 1, 2, 3,... 256), and the broken line indicates the compressed histogram data L which has been compressed to 1/4.
n (X = 1, 2, 3,... 64).

When the data compression is completed, the bill authenticity judging means 51 uses the standard compressed histogram data storage section 44 ′ of the control data storage memory 32 to output the standard value of the bill judged by the denomination direction judging means 50. The compressed histogram data is taken in (step 10). The standard compressed histogram data is obtained by converting the front and back of a plurality of standard “10,000 yen notes”, “5,000 yen notes”, and “1,000 yen notes” in the same manner as described above.
Divided into two block areas, and the above
This is calculated by equations (18) and (19). Therefore, if the pattern of the standard compressed histogram data is Skn, similarly, the denomination direction k = 12 and the number n of block areas are n =
It becomes 16. Next, the banknote authenticity determination unit 51 outputs the standard compressed histogram data P31 of the block area in the banknote pattern S3n determined in the denomination direction (for example, a pattern obtained by carrying the back of a ten thousand yen note from another direction). (X) to P316 (X)
And each block area Z of the bill P to be discriminated in the internal memory 45
Compressed histogram data L1 (X) to L16 in 1 to Z16
Find the absolute value of the difference from (X). That is, the absolute value of the difference is
P31 (X) -L1 (X) |, | P32 (X) -L2 (X) |, ... | P31
6 (X) -L16 (X) |

Next, the bill validity judging means 51 adds these absolute values for each block area. That is, when the sum of the absolute values is R3n, it can be expressed by the following equation. R3n = Σ (X = 1,2,... 64) | P3n (X) −Ln (X) | (20) FIG. 18 shows the sum R3n of the absolute values for each of the block areas Z1 to Z16. In Z1, R31 = 138H, Z2
R32 = 194H... R316 of the same Z16 = 157H. After that, the bill authenticity determining means 51 stores the control data storage memory 3
The reference data T3n corresponding to the banknote pattern S3n is read from the second reference data storage unit 48 (step 1).
1). Then, these reference data and the added value R3n are compared for each of the block areas Z1 to Z16, and if it is determined that all the added values R3n are smaller than the reference data T3n, it is determined that the bill P to be determined is a genuine bill. . Note that the denomination is previously determined by the denomination direction determining means 50 as described above. On the other hand, when at least one of the added values R3n is larger than the reference data T3n, the bill validity determining means 51 determines that the bill P to be determined is a fake bill.

In the above embodiment, the data is compressed to 1/4, but the data compression ratio may be set arbitrarily.
Further, it is not always necessary to perform data compression, but in this case, the memory capacity is increased accordingly. In the above two embodiments, only one side of the banknote P to be discriminated is divided into block areas, and the denomination and authenticity thereof are discriminated. You may make it use for discrimination of a kind etc. In this case, the sensor units 20 are provided above and below, respectively. The block area is 8 or 1
In addition to the division into six, it is of course possible to arbitrarily divide it into 16 or more or 16 or less. Further, although the sensor section 20 is constituted by the reflection type image sensor 37, the same effect can be obtained by using a transmission type image sensor or a magnetic sensor.

[0070]

As described above, according to the present invention,
The banknote tone data is searched and extracted from the discriminated data read by the sensor unit, divided into a plurality of block areas, and histogram data is generated from the divided tone data of each area, and this data is similarly obtained in advance. By comparing with the standard histogram data, the denomination and authenticity of the banknote to be determined can be determined. According to another aspect of the present invention, the denomination and the reading direction of the banknote are determined in advance by performing a weighting process on the integral data obtained by calculating each of the divided grayscale data, and the generated histogram data is Standard histogram data and
By performing the comparison operation, the authenticity of the bill to be determined can be determined. Therefore, the denomination and authenticity of the discriminated bill can be discriminated with high accuracy without collecting and analyzing a huge amount of data on the bill. Therefore, it is possible to provide a bill discriminating apparatus that quickly and accurately discriminates the denomination and the like even when the pattern of the bill is changed or a new bill is issued, and that requires a small memory capacity.

[Brief description of the drawings]

FIG. 1 is a functional block diagram of a CPU according to the present invention.

FIG. 2 is an explanatory view of a conventional state of reading a bill.

FIG. 3 is a block diagram of a conventional bill discriminating apparatus.

FIG. 4 is a diagram illustrating a conventional bill discrimination method.

FIG. 5 is a block diagram of a bill discriminating apparatus according to the present invention.

FIG. 6 is a configuration diagram of a sensor unit according to the present invention.

FIG. 7 is an operation flowchart according to the present invention.

FIG. 8 is a diagram showing a storage area of data to be determined in a data storage memory according to the present invention, together with an address.

FIG. 9 is a diagram showing a block area of divided gradation data.

FIG. 10 is a diagram showing a frequency distribution of histogram data according to the present invention.

FIG. 11 is a diagram showing a calculation result of a block area according to the present invention.

FIG. 12 is a functional block diagram of a CPU according to the second invention.

FIG. 13 is a block diagram of a bill discriminating apparatus of a second invention.

FIG. 14 is an operation flowchart according to the second invention.

FIG. 15 is a diagram showing a block area according to the second invention.

FIG. 16 is a diagram showing a calculation result of a judgment of a denomination direction according to the second invention.

FIG. 17 is a diagram showing a frequency distribution of data according to the second invention.

FIG. 18 is a diagram showing a calculation result of a true / false judgment according to the second invention.

[Explanation of symbols]

 Reference Signs List 20 sensor unit 26 data storage memory 28 CPU 29 address select circuit 38 transport roller 40 banknote gradation data selecting unit 41 data dividing unit 42 data calculating unit 43 banknote determining unit 44 standard data histogram storage unit 44 'standard compressed histogram data storage unit 45 Internal memory 47 Weighting data storage unit 48 Reference data storage unit 49 Data calculation unit 50 Denomination direction determination unit 51 Banknote authenticity determination unit

Claims (3)

(57) [Claims] 1. Reading all printed patterns of a bill to be discriminated
Sensor section to take, and the banknote gradation data obtained from the sensor section
Data storage memory for storing the data to be determined
In the paper currency discriminating apparatus, the banknote gradation data is searched for from the discriminated data, and is retrieved.
Means for selecting banknote gradation data to be output, and dividing the selected banknote gradation data into a plurality of block areas
Data dividing means for performing data processing on the divided gradation data for each block area.each
Histogram data generation for generating histogram data
Means and for each block areaeachSet histogram data in advance
Standard histogram data for each billComparison operation
And a bill determining means for determining a bill.
Bill discriminating device. 2. A bill discriminating apparatus comprising: a sensor unit for reading all print patterns of a bill to be discriminated; and a data storage memory for storing discriminated data including bill tone data obtained from the sensor unit. Means for searching for the banknote gradation data from the discriminated data and extracting the banknote gradation data; data division means for dividing the selected banknote gradation data into a plurality of block areas; and division gradation for each block area. A data calculating means for integrating the data; and weighting the integrated data for each block area based on the weighting data set for each banknote set in advance to determine a denomination and a reading direction of the banknote. the denomination direction determining means, said division gray scale data of each block area - de the data - data processing to the histogram de - histogram de generates the data - data generation hands If, each histogram data of each block area - the data, the denomination
A bill discriminating apparatus comprising: a bill authenticity judging means for judging the authenticity of a bill by performing a comparison operation with standard histogram data set in advance for the bill whose reading direction has been judged. 3. The bill validity judging means compresses and processes the histogram data, and sets a predetermined standard compression history.
3. The bill discriminating apparatus according to claim 2, wherein the bill discriminating apparatus performs a comparison operation with the togram data .