JPS62205492A - Reference pattern generation for sheet paper discriminating machine - 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] [Table of Contents] Overview Industrial Field of Application Conventional Technology Problems to be Solved by the Invention Means for Solving the Problems (Fig. 1) Working Examples Fat Overall Description (Second Part) (Fig., Fig. 3) Explanation of fb1 group division (Fig. 4, Fig. 5, Fig. 6, Fig. 7, Fig. 8) (C1 Explanation of other embodiments Effects of the invention [(already required)] Paper sheet In a standard pattern creation method for a paper sheet discrimination machine that identifies the paper sheet by comparing the reading output of the same sheet with a reference pattern, a group of patterns with a high degree of approximation is created based on the output distribution of the reading output of multiple sheets. By extracting as follows and creating a reference pattern from the extracted group, it is possible to accurately discriminate with a small number of reference patterns.

[Industrial application field]

The present invention relates to a method for creating a reference pattern for a machine for discriminating paper sheets such as banknotes, and in particular, to a method for creating a standard pattern for paper sheets such as banknotes, etc., which can reduce the number of reference patterns created by taking into consideration the traveling conditions of paper sheets. This article relates to a method for creating a reference pattern for a machine.

Banknote validators are widely used in automatic service machines such as automatic teller machines and automatic ticket vending machines.

As shown in FIG. 9(A), such a banknote validating machine has reading heads HDI to HD3 arranged so as to read a specific trace line of a banknote CC to be inserted, and a guide F.
The banknote CC inserted between GD as shown in the figure is sent to sensors S1 and S.
Position correspondence pattern T1 of the banknote CC detected by 2 and sent by a feed roller (not shown) by heads HD1 to HD3.
~Read T3. Then, the authenticity determining section GJ compares the read output with the reference pattern (recognition unit) in the memory PM to determine the authenticity of the banknote CC.

For this banknote CC, if a reference pattern is prepared for each trace line Tl-73, identification is possible in principle, but in reality, it is possible to identify the same banknote depending on the running state of the banknote with respect to the head MDI-HD3. The read signal changes even if the power is turned on.

That is, when the heads (IDI to HD3) read a pattern outside the predetermined trace lines T1 to T3 of the banknote CC by shifting the drawing of the banknote CC in the X direction or obliquely moving it in the R direction,
The read signal changes as shown in FIG. 9(B) depending on the condition of the banknotes, such as new banknotes and old banknotes.

Therefore, it is desired to create a reference pattern that can be used to identify these banknotes.

[Conventional technology]

Conventionally, to create this reference pattern, multiple bills are read in advance in various running conditions and the reading outputs are collected. The standard pattern was created for each group.

For example, when considering the shifting of banknotes, obtain the reading output for 15 trace lines within the maximum range of shifting for the banknote CC as shown in FIG. Physical recognition zone A-E depending on the state of
It was divided into groups and defined as groups. That is, the banknotes are read by giving the banknote CC a forced amount of gathering, and the reading output of the 15) lace line is obtained, and this reading output is shown in FIG. 10(B).
As shown in the figure, the groups were physically divided into groups A to A.

Furthermore, in consideration of skew, banknotes CC are forced to have a skew amount to be read, and the banknotes are similarly divided into groups.

These are performed one after another on a plurality of different banknotes, and the reading outputs are collected, and a reference pattern, which is a recognition unit, is created from the reading outputs classified into groups according to such physical position parameters.

For example, considering 15 trace lines for shift and 8 trace lines for skew, it is necessary to create 120 reference patterns.

Furthermore, if the input direction is not regulated, it is necessary to consider four input directions, and in addition, if multiple denominations can be discriminated, this is necessary for each denomination.

[Problem that the invention seeks to solve]

According to such conventional reference pattern creation methods, the read output for creating the reference pattern is divided into groups based on physical parameters such as offset and skew, and a reference pattern is created for each group, resulting in a huge amount of work. There is a problem in that 49 patterns have to be created, which also causes problems in that the capacity of the reference pattern memory increases and the time required for matching increases.

SUMMARY OF THE INVENTION An object of the present invention is to provide a reference pattern creation method for a Gi'Jtft discrimination machine that can reduce the number of reference patterns without reducing discrimination accuracy.

[Means for solving problems]

FIG. 1 is a diagram explaining the principle of the present invention.

In the present invention, as shown in FIG. 1(A), reading paper sheets while changing a plurality of physical parameters in step 31 and collecting various reading outputs is the same as in the conventional method.
When dividing this into groups in step S2, the groups are divided based on the output distribution of the read output, and from the group divisions, step S3
This is an attempt to create a reference pattern.

That is, whereas conventional group division is based on physical position elements, the present invention attempts to divide groups based on the obtained output distribution.

[Effect]

The starting point of the present invention is to focus on the output distribution of read output. In other words, if the intensity or density output of the collected reading output (hereinafter referred to as sample) obtained from each paper sheet is plotted, excluding the physical position parameter, and the distribution is calculated, Figure 1 (
As shown in B), each sample is separated by a certain circle (
The distribution is not scattered over the entire area within the circle, but is concentrated at a plurality of specific positions within the circle.

Therefore, groups can be divided by extracting individual concentrated groups (recognition units) a, b, c, d, . . . from the output distribution of the entire sample based on the degree of approximation of the output.

Such a distribution is not a recognition standard for each position based on conventional position parameters, but represents data obtained from the entire surface of the specimen as one recognition standard. For example, Fig. 10 (A)
Recognition unit bands B and D, which have a high degree of approximation, belong to different recognition standards in the conventional sample distribution based on position parameters, but in the present invention, the same Li! Become a standard of knowledge.

Therefore, the number of divided groups becomes small. Also, trace line 13
Such broken images belong to an independent recognition standard separate from the trace line 14, and can be divided into groups with significant differences.

〔Example〕

(Al-General description of embodiment) FIG. 2 is a block diagram of one embodiment of the present invention, and FIG. 3 is an explanatory diagram of read output.

In the figure, ■ is a banknote data collection unit which reads patterns on a plurality of trace lines of the inserted banknote CC and outputs the read output as in FIG. 9(A), and 2 is a memory; 3 is a group division/reference pattern creation unit that stores the read output from the data collection unit 1;
It consists of a combinator and divides the read output (sample) collected in the memory 2 into groups based on the output distribution, and creates a reference pattern for the divided groups.

In this embodiment, the banknote data collection unit 1 is provided with a head that reads two trace lines on each of the front and back sides of the input banknote.
Obtain reading output for 4 trace lines in one run. As shown in FIG. 3, each read output is represented by a voltage value from the head, D0 to D.

divided into nine zones, each zone D o '-D
The voltage of a becomes data for one trace line.

Then, the banknotes are inserted in multiple amounts of stacking and skewing, each with 9
Obtain a large number of zone specimens.

Also, a sample is obtained in the same manner by changing the feeding direction.

For example, a total of 1000 specimens 4000 are obtained and stored in the memory 2 for data collection.

Each sample in the memory 2 is divided into groups by a group division/reference pattern creation section 3 based on the output distribution by calculation, and a reference pattern for the divided group is further created.

(Explanation of b1 group division FIG. 4 is a flow diagram of group division processing according to an embodiment of the present invention.
Fig. 5 is an explanatory diagram of distance calculation from the center of distribution in Fig. 4, Fig. 6 is an explanatory diagram of a histogram of the calculated distance in Fig. 5, Fig. 7 is an explanatory diagram of optimal group selection in Fig. 4, and Fig. 8 is an explanatory diagram of distance calculation from the distribution center in Fig. 4.
The figure is an explanatory diagram of group division.

In the present invention, in order to obtain optimal groups (recognition criteria) a, b, c, and d from the samples distributed on the sample plane as shown in FIG. 1(B), the degree of approximation of the samples is determined. For this purpose, the following statistical calculation process is performed in the group division/reference pattern creation section 3.

■ First, find the distribution center and distance distribution/dispersion of the target sample.

For this reason, the average of each zone D o '' D s of the target sample (a D O to D8 and their variances ■. to ■8 are determined.

That is, when the number of samples is N, the sum a0 to a of N samples of each zone D0 to D is obtained as follows.

Next, the sum of squares b0 to b for N samples of each zone D0 to D,
, find.

From formula (1), the average values D0 to D8 of each zone D0 to D6
From Equation (2), is the variance of N samples in each zone D0 to D8.

~V, ha, becomes.

In this way, the average value for each zone D6 to Ds, l
(k-0 to 8) and the variance v1 are determined, then the variance V from the center O (FIG. 8(A)) of the distribution of the entire sample is determined.

This can be done by calculating the statistical distance l from the center and finding the variance v5 of the distance β.

Calculation of this distance is multidimensional due to the number of zones, so it cannot be shown in the drawing, but for simplicity,
When displayed in two dimensions, it becomes as shown in FIG.

For example, if zone D0 is taken as X and zone D is taken as Y, then the coordinates M-, Mv of the distribution center AC for N samples are D in equation (3).
o, DI, and the respective variances V, , V7 are V, , V, in equation (4).

If this variance V=Vv is a normal distribution, based on well-known probability theory, an effective sample exists within the dotted line in FIG. 5 when the distribution center is set to 0.

Therefore, in order to check whether each sample is within the dotted line in FIG. 5, the distance β between each sample and the distribution center is determined.

This distance #ρ is not the usual Euclidean distance, but the Mahalanobis distance shown below, which takes probability into consideration.

V o V + In other words, the center of the distribution of the total number of samples N (Do, . . . Ds)
, and calculate the Mahalanobis distance from this center to each sample.

Using the Mahalanobis distances thus obtained, a histogram (distance distribution) for the number of samples N is created as shown in FIG.

This histogram shows a chi-square distribution when n variables are normally distributed independently of each other in statistics.

Next, the variance of this histogram■. is calculated using the following formula.

−・−・−・・・・−(6) ■ Next, check whether step ■ has been performed J times a predetermined number of times.
If the process has not been repeated a predetermined number of times, a predetermined number of samples farthest from the center of the distribution in FIG. 6 are removed and step (2) is repeated.

The sample BNII to be removed at this time is 6. >K, and sets K such that the value of T n -N x /N becomes a predetermined value, and removes NK.

Removing the farthest sample from the distribution center in this way is
As shown in FIG. 8(A), the sample object is limited to the area inside the dotted line in the figure.

Therefore, when the sample center is found again in step (2), the sample center moves to OI as shown in FIG. 8(B). Similarly,
The farthest sample from this distribution center 0 is removed, as shown in Figure 8 (B
) is limited to samples within the dotted line.

■ Repeat steps ■ and ■, and when the number of removals reaches the specified number of times, the dispersion diagram of the distance distribution for j times is changed to the seventh
Find the lowest value of the variance (2) as shown in the figure, and extract the target samples as a group.

For example, in the example in Figure 8, by repeating steps ■ and ■, the distribution center 0 moves as shown in Figure 8 (B), (C), and (D), and within the circle where the variance is the lowest value. extract the sample as one of the optimal groups.

■ Next, in Figure 8 (D), the removed specimens are a, c,
Steps ■, ■, ■ are repeated for the samples in d, for example, the samples in circle C are extracted as one of the optimal groups, and then the removed samples (in Figure 8 (D), in a and d For example, the samples in circle a are extracted as the optimal group, and the samples in circle d are extracted in the same way.

In this way, optimal groups with a high degree of approximation, that is, recognition criteria, are extracted from the samples as the output distribution, and all samples are divided into groups based on the output distribution.

As is well known, the reference pattern is created using the average value of each zone of the divided sample group.

(C1 Description of Other Embodiments In the above-mentioned embodiments, the discrimination standard pattern for banknotes was explained as an example, but it can also be applied to paper sheets distributed to an unspecified number of people, such as checks and securities.

Further, the reading head may be either a magnetic head or an optical head, and can be applied with or without regulation of the input direction.

Although the present invention has been described above using one embodiment, the present invention can be modified in various ways according to the gist of the present invention, and these are not excluded from the present invention.

〔Effect of the invention〕

As explained above, according to the present invention, those having a high degree of approximation are extracted as a group from the distribution of reading outputs of paper sheets,
Since it is divided into groups, it is possible to significantly reduce the number of reference patterns to be created.

In addition, since the output distribution of the read output is used, it also has the effect of making it possible to create a reference pattern with high accuracy.
It can be used as a reference pattern for paper sheet identification machines, contributing to high-accuracy and high-speed identification.

[Brief explanation of drawings]

Fig. 1 is a diagram explaining the principle of the present invention, Fig. 2 is a configuration diagram of an embodiment of the present invention, and Fig. 3 is a diagram explaining the principle of the present invention.
Figure 4 is an explanatory diagram of the reading output of the embodiment, Figure 4 is a processing flow diagram of an embodiment of the present invention, Figure 5 is an explanatory diagram of distance calculation in the embodiment of Figure 4, and Figure 6 is an illustration of the embodiment of Figure 4. 7 is an explanatory diagram of optimal group selection in the embodiment shown in FIG. FIG. 2 is an explanatory diagram of a discrimination reference pattern creation method. In the figure, CC-...-Banknote, HDI~HD3--Het', 1---Data collection section, 2---Memory, 3---Group division/reference pattern creation section . Patent distributor Fujitsu Limited

Claims (1)

[Scope of Claims] A method for creating a reference pattern for a paper sheet discrimination machine that reads paper sheets to be classified and performs discrimination by comparing the read output with a reference pattern, comprising: reading a plurality of paper sheets; a step (S1) of obtaining a plurality of read outputs; a step (S2) of extracting as a group those having a high degree of approximation from the output distribution of the plurality of read outputs; and a reference pattern from the read outputs belonging to the extracted group. 1. A method for creating a reference pattern for a paper sheet discriminator, comprising the step of creating a reference pattern (S3).