JPS63255793A - Discrimination of sheet paper - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] (Industrial Application Field) The present invention relates to a pattern discrimination method, and in particular detects patterns such as patterns printed on paper sheets such as banknotes and hand stamps, watermarks, etc. The present invention relates to a paper sheet discrimination method that determines the direction in which the paper sheet is placed, its authenticity, etc. based on the detected pattern.

(Prior art) Generally, a sensor, for example, a light source, irradiates light onto a paper sheet to be identified, and the reflected light or transmitted light from the paper sheet is converted into an electrical signal by a photoelectric converter element. A specific track on a paper sheet such as a banknote can be tracked using an optical sensor that detects a magnetic component contained in the ink of a design, character, etc. printed on the paper sheet using an electromagnetic transducer. When scanning, the output of the sensor, that is, the pattern of the paper sheets, may not be the same even if the paper sheets are of the same type, due to variations in individual print densities, differences in the degree of dirt, etc. it is obvious.

Conventionally, as a device using this type of paper leaf discrimination method,
There was a bill validating device as shown in Figure 2. Light sources 2-1.2-2 arranged in a line on a conveyance path that conveys paper sheets 1 printed with a certain pattern in the direction indicated by arrow 8.
, . = 2-n, and photoelectric conversion elements 3-1 .
A reflection sensor is constituted by the combination with 3-2, . . . 3-n. Photoelectric conversion element 3-1.3-2, =-3-
n converts the pattern of the sheet 1 into a current signal and supplies this current signal to the amplifiers 4-1, 4-2, .=4-n.

The amplifiers 4-1, 4-2, -4-n amplify the manually input current signal and transfer it to the signal processing circuit 5 as data on paper g1, that is, as a detection pattern.

The signal processing circuit 5 uses the timing pulse TP output from the pulse generator 15 and having a pulse speed proportional to the transport speed of the sheet 1 to generate an amplifier 4-1.4-2, = 4-
n detection patterns are sequentially read and each photoelectric conversion element 3-
1. The detection pattern is analyzed every 3-2 and -3-n, and the results are transferred to the comparison circuit 6. The comparison circuit 6 compares the result of the analysis by the signal processing circuit 5 with the determination reference value stored in advance in the dictionary memory 7, and outputs the determination result, that is, the determination result of the sheet 1.

FIG. 3 is a waveform diagram illustrating the detection pattern analysis process. The signal processing circuit 5 receives the timing pulses TP supplied from the pulse generator 15 every time when the timing pulses TP reach a predetermined number.
For example, as shown in the figure, four blocks B1, B2, B3,
Amplifier 4-
1. Integrate the detection pattern from 4-2, -4-n,
Obtain each integral value. Furthermore, the signal processing circuit 5 calculates the integral values of the blocks old, B2, B3, and B4, and outputs them as analysis results.

The comparison circuit 6 calculates the analysis results of each block B1, B2, B3, and B4 obtained by the signal processing circuit 5 within the range from the minimum judgment reference value to the large judgment reference value stored in advance in the power storage memory 7. It is determined by comparison whether or not the paper sheet 1 exists, and the result of the determination is output as a determination of whether the paper sheet 1 is front or back, forward or reverse, true or false, etc.

As another determination method, the signal processing circuit 5 uses the detection pattern and p settings for blocks B1, B2, B3, and B4.
2, SL3, and Sth in terms of size, the results (binary values) are counted according to the timing pulse TP, and the counted total value is output as an analysis result. Further, a standard detection pattern is stored in the dictionary memory 7 in advance, and this standard detection pattern and the signal processing circuit 5
The comparator circuit 6 compares the detected pattern with the detection pattern from the above, and if the degree of matching is greater than a predetermined value, it is determined to be true.

FIG. 4 shows an output level diagram of the detection pattern, and in this case, the output level of the detection pattern obtained from the paper sheet 1 is shown in correspondence with time.

Normally, all the output levels of the pattern data of sheet 1 exist between waveforms a and waveforms.

Furthermore, when the output level of a detection pattern is low in a certain region, the detection pattern tends to be at a low level in all regions, that is, close to waveform a. On the other hand, when the output level of the detection pattern tends to be high, the entire area of the detection pattern also tends to be close to the waveform.

Furthermore, when the output level of the detection pattern is at an intermediate level, the detection pattern indicates a tendency that exists between waveforms a and waveforms A and S. However, when the output level difference between the detection patterns is small, the output level may fluctuate throughout the entire area of the detection pattern.

When extreme dirt is partially present on a sheet of paper, it may show a certain tendency as described above, or the output level of the detection pattern may be statistically correlated with the height relationship of the partial detection pattern. known to.

(Problems to be Solved by the Invention) However, the conventional method for identifying paper sheets having the above-mentioned configuration is based on the output level of the detection pattern due to the difference in printing density of the banknotes, dirt attached during the distribution process, etc. When variations exist, there is a problem in that errors are likely to occur in determining front and back, forward and reverse, etc.

In addition, in conventional paper sheet discrimination methods, if there is a judgment error like this, the authenticity is judged based on a judgment standard value that is different from the actual conveyance direction of the banknote, so it is incorrectly judged as fake. Even if there is no error in determining whether the paper sheet is front or back, right or wrong, there is a problem that it is easy to misjudge a genuine note as a counterfeit note when determining authenticity. There was also the undesirable problem that in order to improve the acceptance rate of genuine notes, the tolerance for judgment had to be set wide.

Furthermore, in the conventional method of identifying paper sheets based on the degree of matching between a reference pattern and a detection pattern, a standard detection pattern is set to include all patterns that can be genuine notes, and the standard detection pattern is set to include all patterns that can be genuine notes. If the counter is set to a high firing rate, it will not be possible to make the judgment reference value or standard detection pattern correspond well to the characteristics of the pattern of genuine banknotes, and this will result in the detection of non-genuine banknotes, counterfeit banknotes, etc. There was also the problem that there was a high possibility that so-called counterfeit notes would be mistakenly judged as genuine.

This invention was made to eliminate the above-mentioned problems, and it is possible to obtain highly reliable discrimination even if there are differences in the density of printing ink on paper sheets, stains that have adhered during the distribution process, etc. The purpose is to provide a method for identifying paper sheets.

Another object of the present invention is to provide a paper sheet discrimination method that can increase the acceptance rate of counterfeit bills without impairing the ability to detect counterfeit bills.

(Means for solving the problem) In order to solve the above problem, the present invention compares a detection pattern obtained by scanning paper sheets in a specific direction with a predetermined reference value, The paper sheet discrimination method for determining matters related to the paper sheet includes a first step of determining the overall detection level of the detection pattern and setting a corresponding reference value; The method includes a second step of comparing detected patterns, and a third step of determining matters related to the paper sheet based on the comparison result in the second step.

According to the present invention, since the discrimination method is configured as described below, after grasping the overall pattern from the detection pattern of paper sheets detected by a plurality of sensors, By first selecting a reference value, and then performing processing to determine the conveyance direction of the paper sheet, authenticity, etc. of the detected pattern according to this reference value, the print density difference of the paper sheet,
To obtain highly reliable judgments that are not affected by the degree of contamination.

(Embodiment) FIG. 1 is a block diagram of a paper sheet discrimination device according to the present invention. In FIG. 1, light sources 2-1, 2-2, -2-n are arranged on a conveyance path corresponding to paper sheets 1 conveyed by a conveyance mechanism (not shown) in the direction of an arrow 8, and Photoelectric conversion elements 3-1, 3-2 and -3-n are arranged correspondingly and receive the light reflected by the paper i1.
In combination, n pairs of light reflection sensors are constructed.

The photoelectric conversion element 3-1, 3-2, = 3-n scans the paper sheet 1, converts the pattern into a detection signal of n (n is an integer greater than or equal to l) tree scanning lines, and detects this. The pattern is manually applied to amplifiers 4-1, 4-2, . . . 4-n. Amplifiers 4-1, 4-2, . . . 4-n amplify these detection patterns, and supply their respective outputs to a multiplexer (MX) 9. The multiplexer 9 inputs the outputs of the amplifiers 4-1, 4-2, .
) Enter 10.

Analog-digital conversion circuit 1O is multiplexer 9
Timing pulse TP detects the detection pattern manually from
is sequentially converted into a digital signal by data bus D.
According to the control of the computer (CPU) 11 via B,
Random access memory (RAM) 13 is stored in a predetermined area.

A read-only memory (ROM) 12 includes a storage area for storing a control program for controlling processing operations of the computer 11, and a storage area for dividing a detection pattern into a plurality of blocks and storing slice-level data corresponding to each block. Address X -X+I to store as shown in Figure 7
, X+n+I~2n+I, X+2n+I~, and a data area for storing standard analysis data and standard detection pattern data.

Analog-digital conversion circuit 10, read-only
Memory 12 and random access memory 13 are interconnected to computer 11 via data bus DB and address bus ^1.

A pulse generator (PG) 15 generates timing pulses TP of a frequency proportional to the transport speed of the sheet 1. Timing pulse TP is analog-to-digital conversion circuit 1
0, computer 11 and pulse counter (PCN)
16. The pulse counter 16 divides the timing pulse TP by n.

Next, a description will be given with reference to a flowchart showing the operation of the paper sheet discriminating apparatus shown in FIG. Start processing (5TAR
T) Shi, paper sheet 1 enters a paper sheet thrusting means (not shown),
When the paper sheet 1 is conveyed to the photoelectric conversion elements 3-1, 3-2, .
1.3-2, -3-0, it is converted into an electrical signal, and is amplified by amplifier 4-1.4-2, -4-n,
The signal is inputted via the multiplexer 9 to the analog-to-digital conversion circuit IO. Analog-digital conversion circuit 1
0 converts the detection pattern of the analog signal supplied from the multiplexer 9 into a digital signal and sends it to the computer 1.
1 via the data bus DB according to the control of
It is temporarily stored in the access memory 13.

In step +02, the detection pattern stored in the random access memory 13 is used to determine the level of the entire detection pattern. This is achieved, for example, by determining the sum obtained by integrating each detection pattern. In this case, photoelectric conversion elements 3-1, 3-2, -
Although all the detection patterns output from 3-n may be used, it is also possible to process only a part of them or those corresponding to a specific area of the paper sheet I. Alternatively, the determination may be made by comparing the detected detection pattern with a predetermined level and checking how many times the detected pattern is larger than this level. In either case, the overall characteristics of the detected pattern can be understood.

In step 102, each detected detection pattern is divided into a plurality of data blocks, for example blocks old, B2, B3 and B4 as shown in FIG.
．． The detected patterns may be analyzed by integrating the detected patterns at B2, B3, and B4 and calculating the sum. In addition, as shown in FIG. 3, block B1,
Slice levels SLI, SL2, SL3 and SL4 respectively determined as p corresponding to B2, B3 and B4;
The detection pattern may be compared with each other to determine the magnitude between them, and the detection pattern may be analyzed based on the result.

In step 103, based on the analysis results obtained through the above processing and a predetermined algorithm stored in the read-only memory 12, the transport direction of the paper sheet 1 in the front and back and forward and reverse states is determined. When making this determination,
A predetermined judgment level, that is, a reference value, is set corresponding to the detection pattern obtained in step 101, some typical discrimination reference data are selected from the detection pattern according to this judgment level, and the paper sheet is The direction of transport is determined.

In this case, − types of discrimination reference data are predetermined, this discrimination reference data is corrected in accordance with the overall level of the detection pattern obtained in the process of step +01, and based on the corrected discrimination reference data, The conveyance direction may also be determined by

In general, the designs on paper sheets are not symmetrical either on the left or right side or on the front and back, so the photoelectric conversion element 3-1.3-
The detection patterns of 2,=-3-n have different waveforms in the transport direction.

For this reason, in the process of step 104, step lO
According to the conveyance direction determined by the processing in step 3, the reference analysis value described above and the analysis value of block 11F of the detection pattern are compared, and the J^ quasi-detection pattern and detection pattern described above are compared. A comparison is made, and the comparison result indicating the degree of matching between the two is subjected to a process of determining whether it is true or false.

FIG. 6 is a distribution diagram of the output level, and is a diagram showing the integral value of the detection pattern and its frequency. The overall trend of the detection level becomes clear from the distribution of integral values shown in FIG. For this reason, the integral value is a”-b
, bNc, and c to d.

For example, if the integral value of the detection pattern belongs to the range a%b, the memory allocation in the read-only memory 12 in FIG. l-5l stored in advance from X+n
, I-32, -I-Sn are read out and these are used as the standard analysis value or the standard detection pattern. Similarly, if the integral value belongs to the range b - c, address X+n+lNX
+2n to ll-5l, If -S2, -II-Sn
are read out and set as the reference analysis value or reference detection pattern.

In step +05, it is determined whether or not each comparison result based on the standard analysis value and the standard detection pattern in step 104 has a degree of coincidence of a predetermined degree or more, and when the result is YES, the process proceeds to step 106.

In step 106, a truth determination output process is performed. On the other hand, if the determination in step 105 is negative, the process proceeds to step 107 to output a false determination.

In the above embodiment, a sensor that detects reflected light from paper sheets was used as a sensor for detecting paper sheets, but the sensor may be a sensor that detects transmitted light of paper sheets or a sensor that detects the magnetic field of paper sheets. It may also be a magnetic sensor that detects.

Furthermore, these sensors may be appropriately combined to determine whether the detected patterns are true or false independently or in association with each other, and in either case, the same effects as in the above embodiments can be obtained.

(Effects of the Invention) As described above in detail, according to the present invention, the overall pattern of the paper sheet is grasped from the detection pattern of the paper sheet detected by a plurality of sensors, and based on this, the conveyance direction, When determining authenticity, select the reference value and J^ quasi-detection pattern from the preset ones corresponding to the entire detection level, and select each pattern corresponding to the entire detection level. After correcting the discrimination reference value and standard detection pattern, these are used for judgment, so it is not affected by the print density difference between the five sheets, the degree of staining, etc., and it is possible to judge the conveyance direction, authenticity, etc. It also has the effect of increasing the reliability of the judgment by strictly setting the reference value for the judgment range of the condition that is true in the judgment of authenticity according to the condition of the paper sheet. There is.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a paper sheet discriminating device according to an embodiment of the present invention, FIG. 2 is a block diagram showing a conventional paper sheet discriminating device, and FIG. 3 explains processing for analyzing detected patterns. FIG. 4 is a waveform diagram, FIG. 4 is a diagram of the output level of the detected pattern, FIG. 5 is a flowchart explaining the processing of the paper sheet discrimination device according to the present invention shown in FIG. 1, and FIG. 6 is a statistical diagram of the output level of the detected pattern. FIG. 7 is a diagram showing the memory allocation of the leave-only memory shown in FIG. 1. 11--Computer, 12--Read-only memory, 13--Random access memory.

Claims (5)

[Claims] (1) In a method for identifying paper sheets, in which a detection pattern obtained by scanning paper sheets in a specific direction is compared with a predetermined reference value, and a determination is made regarding matters related to the paper sheets, a first step of determining the overall detection level of the detection pattern and setting a corresponding reference value; a second step of comparing the reference value and the detection pattern; and a result of the comparison in the second step. A method for determining paper sheets, comprising: a third step of determining matters related to the paper sheets by using the following methods. (2) The discrimination method according to claim 1, wherein in the first step, an integral value obtained by integrating the entire detection pattern is used as a reference value. (3) The first step is to compare the preset slice level and the detection pattern in terms of size, count a specific binary value that indicates the comparison result, and use the counted binary value as a reference value. A determination method according to claim 1, characterized in that: (4) The first step is to divide the detection pattern into multiple blocks consisting of a series of preset periods, integrate the detection pattern within each block, and use the integrated value obtained by the integration as a reference value. A determination method according to claim 1, characterized in that: (5) The first step is to divide the detection pattern into a plurality of blocks consisting of a series of preset periods, integrate the detection pattern within each block, and slice the integral value obtained by the integration into preset slices. - The determination method according to claim 1, characterized in that the level and the magnitude are compared, a specific binary value indicating the comparison result of the magnitude is counted, and the counted binary value is used as a reference value.