EP3516634B1 - Method and device for detecting color fading on a banknote, and value-document processing system - Google Patents

Description

The invention relates to a method and a device for detecting color wear on a value document according to the independent claims and a value document processing system.

In the case of documents of value in circulation, in particular banknotes, it can happen that the printed image or other color elements printed on a document of value are at least partially or partially removed or changed, for example by fading, wearing away or washing out.

This so-called ink wear, which is also referred to as “inkwear”, can occur in particular in the case of documents of value in which ink, in particular printing ink, is applied to a polymer substrate. If such a banknote is subjected to certain mechanical stresses, for example due to folding or rubbing, it can happen that the ink on the banknote comes off at least partially from these areas.

When automatically checking the condition, the so-called fitness, of banknotes, it is desirable to also check them for possible color wear.

the EP 2 256 698 A1 describes an apparatus for handling documents.

the DE 27 52 412 A1 describes a testing device for dynamically measuring the degree of soiling of banknotes.

It is the object of the invention to specify a method and a corresponding device which enables a simple and reliable detection of color wear on a document of value, in particular a banknote.

In a method according to the invention, color wear on a document of value, namely a banknote, is detected using an image captured from the document of value, which has a large number of pixels, each of which is assigned at least one intensity value, with a selection of pixels for each pixel and/or for each pixel, the following steps are carried out in at least one partial area of the image: determination of an intensity value from intensity values which are associated with the pixel and pixels in the vicinity of the pixel; calculating a contrast value based on intensity values associated with the pixel and pixels in the vicinity of the pixel; and classification of the respective pixel with regard to color wear of the document of value depending on the intensity value determined in each case and the calculated contrast value.

A device according to the invention for detecting color wear on a document of value, namely a bank note, based on an image captured from the document of value, which has a large number of pixels, each of which is assigned at least one intensity value, has an evaluation unit which is set up for each pixel carrying out the following steps for a selection of pixels and/or for each pixel in at least a partial region of the captured image: determining an intensity value from intensity values which are assigned to the pixel and pixels in the vicinity of the pixel; calculating a contrast value based on intensity values associated with the pixel and pixels in the vicinity of the pixel; and classification of the respective pixel with regard to color wear of the document of value depending on the intensity value determined in each case and the calculated contrast value.

A value document processing system according to the invention has: a sensor device for capturing an image of a value document, in particular a bank note; a device according to the invention for detecting color wear on the document of value; and a device for processing, in particular conveying and/or sorting and/or counting documents of value as a function of color wear detected on the respective document of value.

One aspect of the invention is based on the approach of grading or classifying individual pixels of the image captured from the document of value with regard to possible color wear of the document of value, taking into account both a local brightness and a local contrast of the image in the area of the respective pixel.

For this purpose, for each pixel that is located in at least a partial area of the image and/or belongs to at least one selection of pixels from the image, an intensity value representing the local brightness is determined from the intensity values that correspond to the respective pixel and other pixels in the vicinity of the image Pixels are assigned.

The determined intensity value can also be a value which is derived from two or more intensity values which are assigned to the pixel and/or the pixels in the vicinity of the pixel.

In a corresponding manner, a contrast value characterizing the local contrast of the image in the area of the respective pixel is obtained from intensity values calculated, which are associated with the pixel and/or the pixels in the vicinity of the pixel.

Both the intensity value determined for this pixel and the contrast value calculated for this pixel are then used in the classification of the respective pixel with regard to possible color wear of the document of value. Among other things, this ensures that, for example, pixels in relatively bright areas of the image with relatively low contrast values and pixels in relatively dark areas of the image with relatively high contrast values are reliably classified as pixels that indicate color wear at the corresponding points on the document of value to let.

Advantageously, different areas of the document of value, in particular IR areas, do not have to be treated separately. Furthermore, a smaller number of reference documents of value is required to adapt the method or the device than with known methods for detecting color wear. The effort for any subsequent adaptation of the method or the device is also lower. Last but not least, the influence of quality fluctuations in value documents on the reliability of the color wear detection is also reduced.

Overall, the invention thus enables a simple and reliable detection of color wear on documents of value, in particular banknotes.

A selection of pixels of the image within the meaning of the invention can be a part of all pixels of the image or also all or essentially all pixels of the captured image.

The pixels, for which an intensity value is determined, a contrast value is calculated and finally a classification is made with regard to color wear of the document of value, can be located in a partial area or on the entire area of the captured image. The method can be carried out for all or essentially all of the pixels located in the partial area or on the entire area of the image, or only for a selection of pixels from the partial area or the entire area.

A selection of pixels is preferably obtained by double or n-fold undersampling, so-called subsampling, of the partial area or the entire area of the captured image, in which only every second or nth pixel in both dimensions of the document of value is selected. In this way, the computing time for carrying out the method can be reduced without the recognition or classification of the pixels with regard to possible color wear of the document of value being significantly impaired.

The determined intensity value corresponds to the respective highest intensity value from the intensity values that are associated with the pixel and/or the pixels in the vicinity of this pixel. The brightness used in the classification of the respective pixel is represented here by the highest intensity value of the pixels found in the area of the respective pixel. The classification of the respective pixel with regard to a possible color wear of the document of value is particularly reliable.

In an embodiment not according to the invention, the intensity value determined corresponds to a mean value which is formed from the intensity values which are assigned to the pixel and/or pixels in the vicinity of the pixel. The intensity value determined here thus corresponds to an average brightness of the image of the document of value in the area of the respective pixel, the classification of which with regard to possible color wear of the document of value is particularly reliable as a result.

It is also preferred that the determination of the intensity value and/or the calculation of the contrast value take place using intensity values that are assigned to the pixel and/or the pixels directly adjacent to this pixel. In the case of rectangular, in particular square, detector elements when capturing the image of the document of value, the picture elements of the image are likewise represented by rectangular or square picture elements (pixels). The pixels whose intensity values are used to determine a maximum, minimum or average intensity value and/or to calculate a contrast value preferably form a group of 3 × 3 pixels, with the pixel to be classified in each case being in the middle of the group which the remaining eight pixels border. The latter thus form the immediate vicinity of the pixel.

When determining the maximum, minimum or average intensity value and/or calculating the contrast value, a so-called local operator is preferably used in each case, through which not the intensity value assigned to the respective pixel but also the intensity values of neighboring pixels or pixels from the environment are taken into account. The local image section is therefore also referred to as an operator window, which preferably comprises 3×3 pixels. In principle, however, it is also possible to define larger operator windows with, for example, 5×5, 7×7, 9×9 or 11×11 pixels. The number of pixels in the operator window is preferably odd, so that the pixel under consideration lies in the center of the operator window.

By taking into account the pixels immediately adjacent to the pixel under consideration or the pixels located in a defined operator window when determining the maximum, minimum or average intensity value and/or calculating the contrast value, the values obtained in this way allow a particularly reliable classification of the respective Make pixel in terms of color wear of the document of value.

The calculated contrast value is preferably a measure of a relative change in intensity in the area of the image that is being viewed. The contrast value is preferably calculated taking into account a highest intensity value and a lowest intensity value, the highest and lowest intensity value being determined from the intensity values assigned to the pixel and the pixels in the vicinity of the pixel. The contrast value preferably corresponds to a quotient of a difference between the maximum and minimum intensity value in the numerator and a further intensity value in the denominator, with the further intensity value preferably corresponding to the sum of the maximum and minimum intensity value. Alternatively, the further intensity value can also be one of the intensity values assigned to the pixel and pixels in the vicinity of the pixel are, correspond to the mean value formed. Depending on the application, the most meaningful contrast value can be calculated and used together with the determined intensity value, in particular the mean value formed, in the classification of the respective pixel with regard to color wear of the document of value, which means that this has a particularly high level of reliability.

A classification parameter is preferably calculated using the determined intensity value, in particular the mean value, and the contrast value, and the classification of the respective pixel is carried out depending on the result of a comparison of the classification parameter with a comparison value. As a result, the number of parameters to be compared in the classification is reduced from the original two, namely the determined intensity value or mean value and the calculated contrast value, to just one classification parameter, so that a simple comparison of the calculated classification parameter with a single comparison value enables classification of the respective Pixel takes place in a reliable and particularly simple manner.

The classification parameter is preferably calculated using a differentiation function with one variable each for the determined intensity value, in particular the mean value, and the contrast value. The differentiation function can preferably be a linear function, a step function or else a polynomial, in particular a polynomial of the second order. The respective parameters of the differentiation function, which are preferably constant addends, subtrahends and/or factors, can preferably be defined as part of a test run that takes place in advance or an adaptation of the method or the device. Depending on the use case Such a differentiation function allows a particularly reliable differentiation between image points where there is color wear at the corresponding points of the document of value and image points where this is not the case.

Provision can preferably be made for those pixels which are in the region of an edge of the value document and/or an element characterizing the value document, in particular an element printed on the value document, to be removed from the pixels classified with regard to color wear of the value document. As a result, in the classification of the document of value, pixels are reliably excluded for which the probability that these are incorrectly identified as pixels lying in the area of color wear is comparatively high. Accordingly, a classification of the document of value on the basis of the pixels cleaned in this way is particularly reliable.

Preferably, a measure of the color wear at the corresponding location on the document of value is taken into account in each case for the pixels classified with regard to color wear of the document of value, if necessary after removing pixels located in the area of an edge of the document of value and/or in the area of an element applied to the document of value determines at least one of the following parameters: area of ink wear, brightness of ink wear, contrast of ink wear, position of ink wear. Preferably, the level of ink wear is calculated using a linear function in which one or more of the above parameters are variables. The number of parameters as variables of the linear function and the associated coefficients, in particular factors, can be be determined or determined in test runs with value documents of known properties. The determination of a measure characterizing the color wear on the document of value provides additional information about the extent and/or the relevance of a recognized color wear, so that an even more precise classification of the document of value as “fit” or as “unfit” due to color wear is made possible.

It is also preferred that the image captured from the document of value is a transmission image, in particular a dark-field transmission image. By using the intensity values determined using a transmission image or the calculated contrast values, a classification of the respective pixel with regard to color wear of the document of value can be carried out particularly reliably. This is especially true for dark field transmission images.

Fig. 1

ein Beispiel eines Wertdokumentbearbeitungssystems mit einer Vorrichtung zur Erkennung von Farbabnutzungen an Wertdokumenten;

Fig. 2

ein Beispiel eines Transmissionsbildes einer Banknote;

Fig. 3

ein Diagramm mit einem ersten und zweiten Beispiel einer Unterscheidungsfunktion; und

Fig. 4

ein Beispiel eines aus dem in Figur 2 gezeigten Transmissionsbild abgeleiteten Bildes mit Bildpunkten, die als auf einer Farbabnutzung der Banknote beruhende Bildpunkte eingestuft wurden.

Further advantages, features and application possibilities of the present invention result from the following description in connection with the figures. Show it:

1

an example of a value-document processing system with a device for detecting color wear on value documents;

2

an example of a transmission image of a banknote;

3

a diagram with a first and second example of a distinguishing function; and

4

an example of one from the in figure 2 shown transmission image derived image with pixels that have been classified as due to color wear of the banknote pixels.

1 shows an example of a value-document processing system with a device 10 for detecting color wear on a value document 1, which is in particular a banknote, and a device 30, indicated only schematically, for processing, in particular conveying and/or sorting and/or counting Documents of value, in particular as a function of color wear and tear detected on the respective document of value 1.

The device 10 has a radiation source 2 which is designed to generate electromagnetic radiation with which the bank note 1 is irradiated. The electromagnetic radiation is preferably infrared radiation.

Furthermore, an image sensor 3 is provided, which is designed to detect electromagnetic radiation emanating from the bank note 1 in a spatially resolved manner. In the arrangement of radiation source 2 and image sensor 3 shown in the present example, the image of banknote 1 captured by image sensor 3 is a so-called dark-field transmission image. Alternatively, however, it is also possible to arrange the radiation source 2 relative to the image sensor 3 in such a way that the latter captures a bright-field transmission image or a remission image.

The image of the banknote 1 captured by the image sensor 3 has a large number of image points, so-called pixels, each of which has at least one intensity value is assigned, which is a measure of the intensity of the transmitted or reflected from a variety of locations on the banknote 1 and detected by the image sensor 3 electromagnetic radiation.

The image data 4 generated by the image sensor 3, which contain the intensity values assigned to the individual pixels, are forwarded to an evaluation unit 5 and analyzed there in particular with regard to possible color wear on the banknote 1 and classified accordingly.

The evaluation unit 5 is preferably designed to control the device 30 for processing banknotes in such a way that banknotes 1 are output into different output compartments 31 and 32 depending on the result of the grading or classification with regard to color wear.

The analysis of the image data 4 with regard to possible color wear of the banknote 1 and its corresponding rating or classification is explained in more detail below.

2 FIG. 4 shows an example of a dark-field transmission image 4 of a polymer banknote, which typically has a transparent substrate film that is primed with white lacquer. A print image, which is also referred to as a background print image, is printed on the primer, for example by means of offset printing. Alternatively or additionally, digits and/or infrared-sensitive areas, for example, are printed using intaglio printing.

Due to the typical structure of polymer banknotes described, color wear and other defects appear in the dark field transmission image 4 differently noticeable. For example, areas 6 where the white paint primer has been partially removed appear very dark, whereas color wear in printed areas usually appears as relatively light areas 7, since the electromagnetic radiation incident from below at an angle is scattered by the white paint primer . Holes and tears in the bank note are usually noticeable in the dark-field transmission image 4 as relatively dark areas 8 since there is no scattering of the electromagnetic radiation and it is therefore not detected by the image sensor 3 .

In order to automatically determine the pixels of image 4 in areas 6 and 7, which correspond to the spots on the banknote with color wear, first for each pixel of image 4 or at least for each pixel of a selection of pixels from image 4 , for example in the area of a specific security feature, a mean value m and a contrast value c are determined taking into account the intensity values of the respective pixel itself and the pixels surrounding this pixel.

As illustrated by an enlarged section of an area 9 of image 4, area 9 contains a total of nine pixels, with the pixel (shaded) for which the mean value m and the contrast value c are calculated being in the center of area 9. The mean m is preferably the arithmetic mean, which is calculated as follows: $$\begin{array}{cc}m=\frac{1}{9}& {\displaystyle \sum _{i\in \left(0.8\right)}{P}_i}\end{array}$$

P ₀ to P ₈ denote the intensity values of the observed pixel and the neighboring pixels.

It is also preferred that the contrast value c corresponds to the so-called Michelsen contrast, which is calculated as follows: $$c=\frac{\mathrm{Max}\left\{P_i\right\}-\mathrm{at\; least}\left\{P_i\right\}}{\mathrm{Max}\left\{P_i\right\}+\mathrm{at\; least}\left\{P_i\right\}}\forall i\in \left(0.8\right)$$

This means that the contrast value c corresponds to the quotient of the difference between the highest intensity value and the lowest intensity value on the one hand and the sum of the highest intensity value and lowest intensity value on the other hand, specifically from the intensity values P ₀ to P ₈ in the region 9 under consideration.

In principle it is possible to calculate a corresponding mean value m and contrast value c for all pixels of the image 4 . In order to save computing time without losing relevant information at the same time, it can be advantageous to use undersampling, the so-called subsampling, to use corresponding values m and c only for every second pixel along the two dimensions of the image 4, taking into account the respective surroundings of the pixel to identify what's in 2 is illustrated by way of example using the enlarged region 11 of image 4 . A mean value m and a contrast value c are calculated for each of the hatched pixels.

In a further step, based on the mean value m and contrast value c determined for each pixel, a first classification, which can also be referred to as a preliminary differentiation, is carried out to determine whether the respective pixel indicates color wear at the corresponding point on the banknote.

This is preferably done by computing a grading parameter f using a discriminating function f(m,c) defined as follows: $$f\left(m,c\right)\equiv c-a_0+a_1\times m^1+a_2\times {\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="imgf0001.png"\; state="\{\{state\}\}">m</figure-callout>}}^2=c-{\displaystyle \sum _{i\in \left(0.2\right)}\left(a_i\times m^i\right)\ge 0}$$

In the present example, the pre-discrimination function f(m, c) is a polynomial of the second order in the variables m and c, with the parameters ai with i = 0, 1 and 2 of the polynomial being defined or changed by a previously performed adaptation of the method or the device .can be estimated.

If the classification parameter f calculated using the differentiation function f(m, c) for a pixel is greater than or equal to a comparison value, which in the present example is zero, then this pixel is classified as a so-called inkwear candidate, i.e. the pixel then possibly corresponds to a position on the banknote with a color abrasion.

This is done using an in 3 first example shown (dashed) of a course of the contrast c over the mean value m, which forms the boundary between pixels that were not classified as an inkwear candidate (area below the dashed line) and pixels that were classified as an inkwear candidate ( area above the dashed line). Inkwear candidates, which correspond to color wear in the field of intaglio printing, typically have a relatively low mean intensity m and a relatively high contrast c and are accordingly in the left area of the diagram above the dashed line. Inkwear candidates, which correspond to color wear in the area of the bank note's watermark, have a relatively high average intensity m with comparatively low contrast values c at the same time and are therefore in the right-hand area of the diagram above the dashed line. Inkwear candidates, which are due to color wear in the offset printing area of the bank note, typically have medium brightness and medium contrast c and are accordingly in the middle area of the diagram above the dashed line.

Depending on the application, other functions can also be considered as a differentiation function, for example a linear function or a step function, for which the in 3 shown second example (dash-dotted) of a course of the contrast c over the mean value m is obtained. Pixels that lie above the dot-dash curve based on the respectively calculated mean value m and contrast value c are classified as inkwear candidates, whereas pixels lying below the curve are classified as non-inkwear candidates. Otherwise, the above explanations for the first example of the curve apply accordingly.

In an optional further step, the pixels classified as inkwear candidates can be checked to determine whether they are in specific areas of the banknote, such as in the area of the edge of the banknote and/or in the area of specific printing elements. Since pixels from corresponding image areas of the image 4 can under certain circumstances be incorrectly classified as inkwear candidates, provision can be made for these pixels to be removed from the group of pixels classified as inkwear candidates. Since this is the Inkwear candidates are checked or validated for their significance based on their position in image 4 of the bank note, this step can also be regarded as a validation step or validation.

In a further optional step, the intensity value I(x, y) of the respective pixel (x, y) can be set to zero if the differentiation function f(m, c) calculated from the associated mean value m and contrast value c is smaller as zero, and set it to a specific intensity value, for example the highest intensity value I _max or max{P _i } determined in the area of the respective pixel, if the differentiation function is greater than or equal to zero: $$l\left(x,y\right)=\{\begin{array}{lll}=I_{\mathit{Max}}\left(x,y\right),& \mathit{if}& f\left(m,c\right)\ge 0\\ =0,& \mathit{if}& f\left(m,c\right)<0\end{array}$$

In this way, all pixels classified as inkwear candidates are highlighted particularly well against a black background.

In a further optional step, provision can be made for scaling the image of color wear obtained as described above with regard to brightness by assigning a constant, in particular maximum, intensity value to all pixels with an intensity other than zero, so that the image finally obtained has a black White image in which pixels corresponding to color wear at corresponding locations on the banknote are represented as white areas against a black or very dark background.

Such a picture is in 4 shown as an example. In the 1 Areas 6 and 7 drawn in, which correspond to places on the banknote with color wear, are very clearly recognizable as white areas against a homogeneous dark background, whereas areas 8 (see 1 ) that are due to other defects such as tears or holes in the banknote do not appear. Areas that have been printed and/or provided with a watermark are also no longer visible (see 1 ).

wobei ai angibt, in welchem Bereich des Bildes der Banknote der als Inkwear-Kandidat eingestufte Bildpunkt liegt, und bi ein Maß für die Helligkeit, c_i ein Maß für den Kontrast und d_i die Position des als Inkwear-Kandidaten eingestuften Bildpunkts angeben. Die zugehörigen Koeffizienten InkA, InkB, InkC und InkP können bei vorab durchgeführten Testläufen anhand von Banknoten mit bekannten Eigenschaften berechnet werden. Ebenso kann festgelegt bzw. konfiguriert werden, welche bzw. wie viele der vorstehend genannten Informationen ai, bi, ci bzw. di bei der Berechnung des Inkwear Levels berücksichtigt werden.In an optional further step, which is also referred to as classification, it can be provided that a measure (so-called . Inkwear Level) for the respective ink wear at the corresponding point on the banknote. This can preferably be done by calculating an Inkwear Level (InkL) for each of these pixels using the following linear function: $$\mathit{Incl}\left(i\right)=a_i\times \mathit{IncA}+b_i\times \mathit{InkB}+c_i\times \mathit{InkC}+{\mathrm{i.e}}_i\times \mathit{InkP}$$

where ai indicates in which area of the image of the banknote the pixel classified as an inkwear candidate is located, and bi indicates a measure of the brightness, c _i a measure of the contrast and d _i indicates the position of the pixel classified as an inkwear candidate. The associated coefficients InkA, InkB, InkC and InkP can be calculated in test runs carried out in advance using banknotes with known properties. It can also be specified or configured which or how much of the above information ai, bi, ci or di is taken into account when calculating the inkwear level.

As an alternative to determining the respective contrast value c described above by taking into account the local surroundings of the respective pixel using a so-called local operation or a local operator, contrast values can also be determined using a histogram, in which the number or frequency of pixels via their intensity is applied. Certain areas of the bank note in which there is color wear in the form of abrasion or smearing of the printing ink can be identified in that a transition area occurs in the respective histogram between an area of light pixels and an area of dark pixels, in which only very few pixels can be found . This transition area depends on the respective state of the bank note. By using a dynamically calculated threshold, the respective transition area can be determined in a histogram. A corresponding contrast value c can then be determined using equation (2) on the basis of the respective lowest and highest intensity value in this transition region.

Preferably, the method or the device is also designed to analyze or classify banknotes 1 that may be at an angle or are passing at an angle on the image sensor 3 with regard to any color wear. For this purpose, a skew correction, so-called skew correction, is preferably carried out according to the Bresenham algorithm in the case of undersampling, as a result of which any rounding errors which arise as a result of a discretization of continuous coordinates are minimized.

Claims (11)

1. Method for detecting colour degradation on a document of value (1), namely a banknote, on the basis of an image (4) which is detected by the document of value (1) and has a multiplicity of pixels to each of which at least one intensity value (P_i) is assigned, the following steps being carried out for each pixel of a selection of pixels of the image (4):

   a) determining an intensity value (P_i , m) from intensity values (P₀ - P₈ ) which are assigned to the pixel and pixels in the vicinity of the pixel,

   b) calculating a contrast value (c) on the basis of intensity values (P₀ - P₈ ) which are assigned to the pixel and pixels in the vicinity of the pixel, and

   c) classification of the respective pixel with regard to a colour degradation of the value document (1) as a function of the respectively determined intensity value (Pi, m) and calculated contrast value (c), wherein the determined intensity value (Pi) corresponds to the highest intensity value (max{P_i }) from the intensity values (P₀ - P₈ ) which are assigned to the pixel and pixels in the environment of the pixel.
2. Method according to one of the preceding claims, wherein the determination of the intensity value (Pi, m) and/or the calculation of the contrast value (c) is carried out on the basis of intensity values (Po - P₈ ) which are assigned to the image point and the image points directly adjacent to this image point.
3. Method according to one of the preceding claims, wherein a highest intensity value (max{Pi }) and a lowest intensity value (min{Pi }) are determined from the intensity values (P₀ - P₈ ) assigned to the pixel and the pixels in the vicinity of the pixel, and the contrast value (c) is calculated from the highest intensity value (max{P_i }) and the lowest intensity value (min{P_i }).
4. Method according to one of the preceding claims, wherein a classification parameter (f) is calculated on the basis of the determined intensity value (Pi, m), in particular the mean value (m), and the contrast value (c), and the classification of the respective pixel takes place as a function of the result of a comparison of the classification parameter (f) with a comparison value.
5. Method according to claim 4, wherein the classification parameter (f) is calculated on the basis of a discrimination function (f(m, c)) with one variable each for the determined intensity value (Pi, m), in particular the mean value (m), and the contrast value (c).
6. The method according to claim 5, wherein the discrimination function (f(m, c)) is given by a linear function, a staircase function or a polynomial, in particular a polynomial of the second degree.
7. Method according to one of the preceding claims, wherein those pixels are removed from the pixels classified with regard to colour wear of the value document (1) which lie in the region of an edge of the value document (1) and/or in the region of an element applied, in particular printed, on the value document (1).
8. Method according to one of the preceding claims, wherein for the pixels classified with respect to a colour wear of the value document (1) a measure for the colour wear at the corresponding location of the value document (1) is determined based on at least one of the following parameters: Area of colour wear (a_i ), brightness of colour wear (b_i ), contrast of colour wear (c_i ), position of colour wear (d_i ).
9. Method according to one of the preceding claims, wherein the image (4) detected by the value document (1) is a transmission image, in particular a dark field transmission image.
10. Device for detecting colour degradation on a document of value (1), namely a banknote, on the basis of an image which is detected by the document of value (1) and has a multiplicity of pixels to each of which at least one intensity value (Pi) is assigned, having an evaluation unit (5) which is set up to carry out the following steps for each pixel of a selection of pixels of the image (4):

    a) determining an intensity value (Pi, m) from intensity values (P₀ - P₈ ) assigned to the pixel and pixels in the vicinity of the pixel,

    b) calculating a contrast value (c) on the basis of intensity values (P₀ - P₈ ) assigned to the pixel and pixels in the vicinity of the pixel, and

    c) classification of the respective pixel with regard to a colour degradation of the value document (1) as a function of the respectively determined intensity value (Pi, m) and calculated contrast value (c), wherein the determined intensity value (P_i) corresponds to the highest intensity value (max{Pi }) from the intensity values (Po - P₈ ) which are assigned to the pixel and pixels in the environment of the pixel.
11. A value document processing system comprising

    - a sensor device (2, 3) for detecting an image (4) of a value document (1), in particular a banknote,

    - a device (10) for detecting colour wear on the value document (1) according to claim 11, and

    - a device (30) for processing, in particular conveying and/or sorting and/or counting, value documents in dependence on colour wear detected on the respective value document (1).

Images