EP2143081B1 - Method and device for testing value documents - Google Patents

Description

The invention relates to a method for checking value documents, in particular for detecting forged value documents, and to an apparatus for carrying out the method. The forged value documents to be recognized are composed counterfeit orders, which are composed of parts of different value documents. The composed counterfeits can be composed of parts of genuine and forged value documents, but there are also known composite counterfeits, which are composed exclusively of parts of real value documents.

Various methods for detecting counterfeit banknotes are known from the prior art. Composed counterfeits whose individual parts are glued together with adhesive strips can in some cases be detected indirectly by detecting the adhesive strips with the help of a reflection measurement. However, this is not possible in every case of a glued together compositional counterfeit. In addition, for authenticity testing, the banknotes are checked, for example, for properties that distinguish genuine banknote paper from ordinary paper, such as its fluorescence properties. Many composed counterfeits are partly made of genuine and partly fake paper, but have similar fluorescence properties to real banknotes. In addition, counterfeits consisting solely of parts of genuine banknotes are also put together. With the previous methods, it is not possible to reliably detect such composite counterfeits which have comparable measurement signals, e.g. Deliver fluorescence signals, like real banknotes.

The US4464786A discloses to use a line sensor for checking banknotes, by which the banknote is optically scanned line by line.

From the signals D (i) detected along a line, a kind of gradient value S (i) along the line (X) is calculated for each "picture element" i. Each gradient value S (i) is compared with the predetermined value 50, and for each gradient value S (i) exceeding 50, C (i) = 1 is set and the counter G is incremented by one. The counter G provides information about how often the detected intensity changes along the respective line. At the unprinted edges of the banknote, G is small because the detected signal hardly changes along such a line. For the lines in which there are printed characters of the banknote ("character row"), G is large. Based on the comparison of the counter G with the "second standard data", the "character rows" 11a, 11b, 11c of the banknote are found. The position of the printed area of the banknote is calculated from the position of the "character rows." The pattern found in the printed area of the banknote is compared with a standard pattern, and the similarity of the two patterns determines the authenticity or type of banknote. In the US4464786A For example, the gradient values S (i) are used only for locating the printed area of the banknote, but not for checking the banknote itself. Instead, the banknote is checked on the basis of the detected signals themselves ("detected pattern" of the printed image of the banknote). The US4464786A but does not deal with checking the bill for the presence of dividing lines where the parts of a composed counterfeit are composed.

The EP0947964 A discloses a method of examining securities in which samples of the security are detected in rows and columns. The samples of each column are added to a total value each, reducing the area to a single resulting row. The sum values of the resulting row are subjected to pattern recognition. Not every single one of these pattern recognition Sample value must be compared with a corresponding default value, but only the respective sum value, thus simplifying the evaluation of the samples. However, only the samples themselves are added up or linked together, but no gradient values.

The DE202005021081 U1 discloses a sensor for checking the authenticity of a banknote which is movable relative to a transport device which transports the banknote in order to allow different scanning curves on the banknote.

It is therefore an object of the invention to provide an easy way to reliable detection of forged value documents, in particular of composed counterfeits.

This object is solved by the features of the independent claims. In dependent claims advantageous embodiments and developments of the invention are given.

In the method according to the invention, in a first step, the signal intensities of a measurement signal are determined at a plurality of measurement points on a value document. Subsequently, a group of these measuring points is selected, whereby the selection of the group takes place in dependence on a typical separating line position on the value document. Alternatively, the group of measurement points can be selected even before the determination of the signal intensities. In a further step, gradient values of the signal intensities for the measuring points of the group are determined. To determine a connection strength of the group, the Gradient values within the group are linked together, whereby the connection strength of the group is determined by forming a product or a sum of the gradient values of the group. The connection strength is evaluated, for example, by comparing the connection strength to a reference connection strength that applies to the group.

The method according to the invention serves to detect forged value documents, for example for the detection of composed counterfeits. The value document is thus checked for the presence of dividing lines on which the value document is composed or in which individual components are connected to form the value document. In general, a composed counterfeit may have one or more dividing lines on which it is composed.

Comparing the connection strength with the associated reference connection strength results in a difference that is specific to the selected group of measurement points. The evaluation of the size of the difference can be done in addition to a simple comparison, whether the ever average connection strength is less than or greater than the corresponding reference connection strength. From the difference or from the size of the difference, a probability can be derived that a dividing line or at least a partial section of a dividing line is arranged in the respective group of measuring points. If the connection strength greatly exceeds the reference connection strength, for example above a certain threshold value, a higher probability of the existence of a separation line can be concluded within the measurement points of the respective group than, for example, if the reference connection strength is just exceeded.

For determining the signal intensities, at least one profile of the signal intensity of the measurement signal is determined as a function of the location on the value document, e.g. a two-dimensional distribution of the signal intensity. The two-dimensional distribution of the signal intensity can be determined over the entire value document or also over one or more subareas of the value document.

The connection strength of the respective group of measuring points provides a quantitative statement as to how far a large gradient value exists throughout the group, in particular along a specific direction on the value document. To determine the connection strength, the product of the gradient values within the group or also their sum is formed.

In a further development of the method, the signal intensities are normalized to reference intensities, which are preferably specific to the respective Are measuring point. In this case, the signal intensities are standardized, for example at each of the measuring points, to a reference intensity valid for the respective measuring point. The reference intensities can be determined on the basis of a large number of genuine value documents or determined in advance of the test. The reference intensity of a measuring point can be given by an average value of the signal intensities determined for the respective measuring point on the basis of the plurality of value documents.

In addition to the reference intensities of the measuring points, the reference connecting strength valid for a group of measuring points can also be determined on the basis of a multiplicity of genuine value documents or determined in advance of the test. The reference connection strength of a group may be given by an average of connection strengths determined for the respective group of measurement points from the plurality of value documents.

Preferably, the reference intensities and / or the reference connection strengths are determined on the basis of value documents of the type of the value document to be checked, in the case of banknotes, for example, on the basis of banknotes of the same denomination. Specific reference intensities and / or specific reference connection strengths can each be stored for the different types of value documents. The reference intensities and / or the reference connection strengths may be selected based on the type of value document, for example the currency and denomination of a banknote. In order to select the reference intensities and / or the reference connection strengths that apply to the value document to be checked, the type of value document is identified, for example, before the value document is checked. In the case of banknotes can this identification, for example, be a denomination determination preceding the method according to the invention.

For determining the gradient values, in one embodiment, for each of the measurement points of the selected group, the first derivative of the signal intensity along a first direction is formed on the value document. For example, the gradient value of the signal intensity at the respective measurement point may be proportional to the absolute value of the first derivative of the signal intensity at the respective measurement point, the first derivative being formed along the first direction on the value document.

In a further embodiment, at least one difference from the signal intensity at the respective measuring point and the signal intensity at at least one adjacent measuring point is formed for each of the measuring points of the selected group for determining the respective gradient value, wherein the adjacent measuring points along a first direction on the value document to the respective measuring point are adjacent. For example, the gradient value of the signal intensity at the respective measuring point may be proportional to the absolute value of the difference, which is formed from the signal intensity at the respective measuring point and the signal intensity at at least one adjacent measuring point.

The first direction preferably runs along a transport direction of the value document, in particular approximately parallel to a longitudinal direction of the document of value or approximately perpendicular to the longitudinal direction, ie approximately parallel to the shorter side of the document of value. The value document is guided past a sensor along the transport direction, with which the signal intensities of the measurement signal are determined.

In a special embodiment, the measuring points of the group are arranged along a second direction on the value document. The second direction preferably runs at a non-zero angle to the first direction, for example, vertically to the first direction.

In the method according to the invention, for example, an optical measuring signal is used as the measuring signal, which is in particular in the visible or in the non-visible spectral range. The measurement signal may be a luminescence signal which is emitted by the value document, for example a luminescence signal excited by UV light, in particular a fluorescence signal.

In a development of the method, additionally at least one further group of the measuring points is selected, at which the signal intensity of the measuring signal is determined. For example, selecting the further groups may be done immediately after selecting the previously selected group. Alternatively, the selection of the further groups may also be made during or after performing one or more of the method steps of the invention following selection of the previously selected group, such as after determining the connection strength for the previously selected group. After selecting a further group, further gradient values of the signal intensities for the measuring points of the respective further group are determined. In order to determine a further connection strength of the respective further group, the further gradient values are linked to one another. Subsequently, the further connection strength is evaluated. For evaluation, the further connection strength is compared, for example, with a further reference connection strength, which was determined for the measurement points of the respective further group, eg on the basis of genuine value documents. For the different selected groups can use the same or individual reference connection strengths.

If the connection strength and / or the further connection strength exceeds the reference connection strength valid for the respective group, then there is a high probability that within the measurement points of this group at least a partial section of a separation line runs. The tested value document can then be sorted out with suspicion of the existence of a composed forgery.

For example, the selection of which measurement points are grouped together is based on the positions of a value document in which the dividing lines of composed forgeries are typically placed. In a development of the method, the selection of the group therefore takes place as a function of a typical parting line position on the value document, the typical parting line position being determined on the basis of a plurality of forged value documents. To determine a typical parting line position, for example, the positions of the parting lines of several known composed counterfeits are detected and evaluated statistically. Depending on this, the groups of the measuring points can then be selected for the method according to the invention. The selection of groups is e.g. in such a way that the entire value document or a subarea of the value document is checked for the presence of parting lines.

The selection of the further groups can also be made depending on the connection strength of the previously selected group or of several previously selected groups. In addition, the selection of the further groups can also be dependent on at least one difference between the connection strength of at least one previously selected group and the Reference connection strength of at least one previously selected group.

The further groups can be used to check a part of the area and / or an immediate area of the area that has already been checked by the first group. If, for example, a relatively large connecting strength of a first group of measuring points indicates an indication of a possible dividing line - if e.g. a subsection of a non-rectilinear separation line is detected - so further groups of measurement points in the immediate vicinity of the first group could be selected. With the aid of the further groups, the suspect area of the value document and / or its immediate surroundings can be checked at different angles

As further groups for checking the suspicious area, it is also possible to select measuring points which are not arranged over the entire value document but only over a subsection of the value document. The further groups may contain or be a subset of one or more previously selected groups. With the help of the other groups, a subarea of a previously tested area can also be checked.

In a specific embodiment, the measuring points of a plurality of selected groups are arranged parallel to one another on the value document. However, the measurement points of the selected groups may also be arranged along different directions on the value document.

Another aspect of the invention relates to the device used to carry out the method according to the invention. This device preferably has a sensor for determining the signal intensities of the measurement signal. The sensor may be an image sensor for detecting optical features of value documents, for example banknotes, and preferably has at least one detector line for determining the signal intensities of the measurement signal.

The invention will be described by way of example with reference to the accompanying drawings.

Fig. 1a

Schematische Abbildung einer composed-Fälschung, die aus zwei Teilen zusammengesetzt ist,

Fig. 1b

Skizzierter Verlauf der Signalintensität s_m der Messspur m und der Gradientenwerte g_m der Messspur m,

Fig. 1c

Skizzierter Verlauf der Signalintensität s_m+1 der Messspur m+1 und der Gradientenwerte g_m+1 der Messspur m+1,

Fig. 2a

Zweidimensionale Anordnung der Messpunkte in Form eines Gitters auf der composed-Fälschung,

Fig. 2b

Drei Gruppen (a, b, c) von Messpunkten mit parallel zueinander angeordneten Messpunkten,

Fig. 2c

Tabelle mit beispielhaften Gradientenwerten der Gruppen a, b, c sowie daraus ermittelte Verbindungsstärken V_a, V_b, V_c und zugehörige Referenz-Verbindungsstärken R_a, R_b, R_c, und

Fig. 3

Fünf Gruppen (a', b', c', d', e') von Messpunkten mit schräg zur Transportrichtung, entlang verschiedener Richtungen angeordneten Messpunkten.

Show it:

Fig. 1a

Schematic illustration of a composed counterfeit composed of two parts

Fig. 1b

Sketched course of the signal intensity s _{m of} the measuring track m and the gradient values g _{m of} the measuring track m,

Fig. 1c

Sketched course of the signal intensity s _{m + 1 of} the measuring track m + 1 and of the gradient values g _{m + 1 of} the measuring track m + 1,

Fig. 2a

Two-dimensional arrangement of the measurement points in the form of a grid on the composed counterfeit,

Fig. 2b

Three groups (a, b, c) of measuring points with measuring points arranged parallel to one another,

Fig. 2c

Table with exemplary gradient values of groups a, b, c and connection strengths V _a , V _b , V _c determined therefrom and associated reference connection strengths R _a , R _b , R _c , and

Fig. 3

Five groups (a ', b', c ', d', e ') of measuring points with measuring points obliquely to the transport direction, arranged along different directions.

In FIG. 1a schematically a composed forgery 1 is shown, which is composed of two parts, for example, a left, real part banknote 1a and a right, wrong partial banknote 1b. At an approximately perpendicular through the composed counterfeiting 1 ongoing dividing line 2, the two sub-bank notes 1a and 1b are glued together. On the real sub-bank note 1a, a security element 3 is shown by way of example, which fluoresces under UV illumination. To check for authenticity, the composed counterfeit 1 is moved under a detector line 5 along the transport direction x marked with the arrow 4. The detector row 5 is part of a sensor for checking value documents which detects the signal intensity of the fluorescence of the composed forgery 1 to be tested as a function of time or as a function of the location x on the composed counterfeit 1. The detector row 5 has a plurality of measuring tracks, which are arranged perpendicular to the transport direction (in the y-direction), inter alia the two measuring tracks m and m + 1, the signals of which are considered below by way of example.

The Figures 1b and 1c show the signal intensity s _{m of} the measuring track m and the gradient values g _{m of} the measuring track m and the signal intensity s _{m + 1 of} the measuring track m + 1 and the gradient values g _{m + 1 of} the measuring track m + 1 as a function of the location x on the composed counterfeit 1. The signal intensities s _m and s _{m + 1} have a clear jump at the location x _T , the intersection of the x-axis with the dividing line 2, which results from different fluorescence properties of the real partial bank note 1 a and the incorrect partial bank note ₁ b. In addition, in the measuring track m is a significantly increased Signal intensity in the area of the fluorescent security element 3 to recognize. The signal intensity s _{m + 1} has no increased signal intensity at the x-coordinates of the security element 3, since the measuring track m + 1 no longer detects the fluorescence of the security element 3. The gradient values g _m and g _{m + 1} result from the signal intensities s _m and s _{m + 1,} respectively, by calculating the absolute value of the first derivative of the respective signal intensity in the transport direction x.

In addition, each of the signal intensities also from the other measuring tracks of the detector line 5 s _1, s _2, ... and the corresponding gradient values g _1, g ₂ determines .... This yields a two-dimensional distribution of the signal intensities and the gradient values over the entire composed counterfeiting 1. At the location x _T , not only the gradient values g _m and g _{m + 1} but also the gradient values of the remaining measurement tracks have a peak corresponding to a jump of the respective signal intensity.

FIG. 2a illustrates the two-dimensional arrangement of the measuring points, which are represented as cells of a grid in the xy-plane, and their position with respect to the composed counterfeiting 1. Each measuring line of the detector line 5 corresponds to one line of the grid. Following the determination of the gradient values at each of the measuring points, a plurality of groups of measuring points are formed. The measuring points of each group are each arranged along a specific direction on the value document. FIG. 2b shows by way of example three groups a, b, c whose measuring points are arranged along the y-direction and parallel to one another. In this example, the measuring points of the three groups a, b, c are chosen such that the signal intensities in the area of the dividing line 2 and in their immediate surroundings are detected on them.

In the table of Figure 2c For example, in lines 1-12, gradient values of the measuring points of the three groups a, b, c are listed. From the gradient values of the individual groups a, b, c, in each case a connection strength V _a , V _b , V _{c is} determined by multiplying the gradient values within each group of measuring points. From the gradient values of the group b, a very large connection strength Vb results compared to the groups a and c. For evaluation, the connection strengths V _a, V _b are then compared V _c of the groups with the reference connection strengths Ra, Rb, Rc, which apply to the respective group and for example have been determined in advance of the examination based real value documents. The connection strengths V _a and V _c are significantly smaller than the respective reference connection strength Ra or Rc. In contrast, the connection strength Vb is significantly greater than the corresponding reference connection strength Rb. It can be concluded from the comparatively large connection strength Vb that within the measuring points of the group b a separation line 2 is likely to run. The tested value document can thus be sorted out with suspicion of the presence of a composed counterfeit 1.

In FIG. 3 Examples of further groups a'-e 'of measuring points are shown. The measuring points of the groups a'-c 'are arranged in a star shape, here from the second measuring point of the first line of the measuring point grid, at different angles over the document of value 1. In the same way, several, arbitrarily selected measuring points can be used as output measuring point. The measuring points of the groups d 'and e' are arranged parallel to one another and run at a non-zero angle to the transport direction x over a subsection of the value document 1. The measuring points lying at the edge of the value document 1 were excluded in the groups d 'and e'. With the help of these groups can also composed fakes be recognized, whose dividing lines run obliquely over the composed counterfeiting.

Not only groups with measuring points arranged transversely or obliquely to the banknote can be selected for the method according to the invention, but also those groups whose measuring points are arranged in the longitudinal direction of the banknote. In addition, the measurement points of a group can also be arranged so that they do not lie in a line. Instead, the measurement points of a group may also lie on a curve and / or the arrangement of the measurement points may have an offset, e.g. to detect composed counterfeits with corresponding dividing lines.

Claims (13)

1. A method for checking value documents (1), in particular for recognizing forged value documents (1), having the steps of:

   - determining signal intensities (s) of a measuring signal at a plurality of measuring points on a value document (1),

   - selecting a group (a-c, a'-e') of the measuring points, the selection of the group being effected in dependence on a typical separating line position on the value document (1),

   - determining gradient values (g) of the signal intensities (s) for the measuring points of the group (a-c, a'-e'),

   - linking the gradient values (g) of the group (a-c, a'-e') for ascertaining a connection strength (V_a-c, _a'-e') of the group (a-c, a'-e'), the connection strength (V_a-c, _a'-e') of the group (a-c, a'-e') being ascertained by forming a product or a sum of the gradient values (g) of the group (a-c, a'-e'),

   - evaluating the connection strength (V_a-c, _a'-e'),

   wherein the value document (1) is checked for the presence of separating lines (2) at which the value document (1) is assembled.
2. The method according to claim 1, characterized in that for evaluating the connection strength (V_a-c, _a'-e') the connection strength (V_a-c, _a'-e') is compared to a reference connection strength (R_a-c, _a'-e').
3. The method according to claim 2, characterized in that there is derived from a difference between the connection strength (V_a-c, _a'-e') and the reference connection strength (R_a-c, _a'-e') a probability of at least a segment of a separating line (2) being disposed in the group (a-c, a'-e') of measuring points.
4. The method according to one or more of the preceding claims, characterized in that for determining the signal intensities (s) there is determined at least one curve of the signal intensity (s) as a function of the place (x, y) on the value document (1), in particular a two-dimensional distribution of the signal intensity.
5. The method according to one or more of the preceding claims, characterized in that for determining the gradient values (g) of the group (a-c, a'-e') at least a first derivative of the signal intensity (s) is formed along a first direction (x) on the value document (1), the gradient value (g) at least at one measuring point of the group (a-c, a'-e') being in particular proportional to the absolute value of the first derivative at the measuring point.
6. The method according to one or more of the preceding claims, characterized in that for determining the gradient values (g) of the group at least one difference of the signal intensity (s) at a measuring point of the group and of the signal intensity (s) at least at one neighboring measuring point is formed, the neighboring measuring point being neighboring to the measuring point along a first direction (x) on the value document (1), the gradient value (g) at least at one measuring point of the group being in particular proportional to the absolute value of the difference.
7. The method according to either of claims 5 to 6, characterized in that the first direction (x) extends along a transport direction of the value document (1) which extends in particular approximately parallel to a longitudinal direction of the value document (1) or approximately perpendicular to the longitudinal direction.
8. The method according to claim 7, characterized in that the measuring points of the group (a-c, a'-e') are disposed along a second direction on the value document (1), the second direction extending preferably at a non-zero angle to the first direction (x), in particular vertically to the first direction (x).
9. The method according to one or more of the preceding claims, characterized in that the measuring signal is an optical measuring signal, preferably a luminescence signal excited by UV light, in particular a fluorescence signal.
10. The method according to one or more of the preceding claims, characterized by the further steps of:

    - selecting at least one further group (a-c, a'-e') of the measuring points,

    - determining further gradient values (g) of the signal intensities (s) for the measuring points of at least one of the further groups (a-c, a'-e'),

    - linking the further gradient values (g) of at least one of the further groups (a-c, a'-e') for ascertaining at least one further connection strength (V_a-c, _a'-e'),

    - evaluating at least one of the further connection strengths (V_a-c, _a'-e').
11. The method according to claim 10, characterized in that the selection of at least one of the further groups (a-c, a'-e') is effected in dependence of the connection strength of the group according to claim 1 (a-c, a'-e') and/or in dependence of a difference between the connection strength (V_a-c, _a'-e') of the group (a-c, a'-e') according to claim 1 and the reference connection strength (R_a-c, _a'-e') according to claim 2.
12. The method according to one or more of claims 10 to 11, characterized in that through the group (a-c, a'-e') according to claim 1 an area of the value document (1) is checked and through at least one of the further groups (a-c, a'-e') a part of the area and/or immediate surroundings of the area are checked, the area and/or the immediate surroundings of the area being checked in particular at different angles.
13. An apparatus for checking value documents (1) for the presence of separating lines (2) at which the value document (1) is assembled, wherein the apparatus preferably has at least one sensor for determining signal intensities (s) of a measuring signal at a plurality of measuring points on a value document (1) and the apparatus is adapted for carrying out the following steps of:

    - selecting a group (a-c, a'-e') of the measuring points, the selection of the group being effected in dependence on a typical separating line position on the value document (1),

    - determining gradient values (g) of the signal intensities (s) for the measuring points of the group (a-c, a'-e'),

    - linking the gradient values (g) of the group (a-c, a'-e') for ascertaining a connection strength (V_a-c, _a'-e') of the group (a-c, a'-e'), the connection strength (V_{a-c, a'-e'}) of the group (a-c, a'-e') being ascertained by forming a product or a sum of the gradient values (g) of the group (a-c, a'-e'),

    - evaluating the connection strength (V_a-c, _a'-e'), the value document (1) being checked for the presence of separating lines (2) at which the value document (1) is assembled.

Images