Method for checking a value document

Patent metadata

Field	Value
Publication number	US-9582953-B2
Application number	US-201415021548-A
Country	US
Kind code	B2
Filing date	Sep 11, 2014
Priority date	Sep 13, 2013
Publication date	Feb 28, 2017
Grant date	Feb 28, 2017

Abstract

A method is described for checking a value document, wherein measuring values of the value document are detected in spatially resolved fashion. The measuring value detected in the respective measuring point is allocated to a node corresponding to this measuring point, and a two-dimensional network of nodes is formed therefrom. A network is formed from the two-dimensional network of the nodes and a source node and a sink node. By means of the maximally possible flow through the network, the value document is classified as suspected of forgery or not suspected of forgery. The maximally possible flow through the network is a measure for the degree of probability of the value document having a continuous object along a direction transverse to the direction of the network, said object indicating a manipulation of the value document, such as e.g. an adhesive strip or a separating line of a composed forgery.

First claim

The invention claimed is:

1. A method for checking a value document, with the following steps of:
detecting in spatially resolved fashion measuring values (Ii, Ij) in a plurality of different measuring points on the value document which are distributed over the value document two-dimensionally;
forming a two-dimensional network of nodes (Ki, Kj), wherein each node corresponds to at least one measuring point on the value document, and allocating the measuring value detected in the respective measuring point to the node corresponding to this measuring point;
creating two additional nodes arranged on mutually opposing sides of the two-dimensional network, and connecting the two additional nodes with the two-dimensional network of nodes (Ki, Kj), wherein one of the additional nodes is a source node (KQ) forming a source for a flow through the network, and the other one is a sink node (KS) forming a sink for a flow through the network;
determining respectively one capacity value for each pair of two adjacent nodes (Ki, Kj) of the network, said capacity value being a measure of the maximally possible flow between the two adjacent nodes, by mutually comparing the measuring values (Ii, Ij) of respectively two adjacent nodes (Ki, Kj) of the network, ascertaining respectively one capacity value (cij) for each of the pairs of adjacent nodes on the basis of this comparison and allocating the respective capacity value (cij) to a connecting line between the two adjacent nodes (Ki, Kj) of the respective pair of adjacent nodes whose measuring values have been mutually compared;
computing the maximally possible flow from the source node (KQ) through the network to the sink node (KS) on the basis of the ascertained capacity values (cij);
classifying the value document, in particular in view of the presence of a composed forgery, as suspected of forgery or not suspected of forgery in dependence on the computed maximally possible flow through the network.

2. The method according to claim 1 , wherein, for checking the value document, the maximally possible flow through the network is compared to a flow threshold and the value document is classified as suspected of forgery in view of the presence of a composed forgery, if the maximally possible flow undershoots or reaches the flow threshold, and as not suspected of forgery in view of the presence of a composed forgery, if the maximally possible flow overshoots the flow threshold.

3. The method according to claim 1 , wherein for computing the maximally possible flow a numerical optimization method is utilized in which a minimum cut through the network is ascertained that forms a “bottleneck” for the flow through the network, and that in the case that the value document is classified as suspected of forgery, the position of the minimum cut through the network is utilized for determining the position of an adhesive strip ( 80 )/ a separating line (T) of the value document, wherein the position of the adhesive strip ( 80 )/ the separating line (T) ascertained by means of the minimum cut is utilized in particular to check the value document for the presence of a composed forgery with the aid of further methods.

4. The method according to claim 1 , wherein to the respective node there is allocated the measuring value detected in the respective measuring point minus a measuring value to be expected for the respective measuring point within the framework of a parametric model.

5. The method according to claim 1 , wherein the maximally possible flow through the network is computed along a direction which on the value document corresponds to that direction which extends perpendicularly to the longitudinal direction of typical adhesive strips ( 80 )/ typical separating lines (T), wherein the maximally possible flow is computed along a direction that corresponds to the longitudinal direction of the value document.

6. The method according to claim 1 , wherein the maximally possible flow through the network is computed both for a first direction through the network and for a second direction extending perpendicularly to the first direction, wherein the first direction corresponds in particular to the longitudinal direction of the value document and the second direction corresponds in particular to the transverse direction of the value document.

7. The method according to claim 6 , wherein
the first direction a first maximally possible flow through the network is computed, and for the second direction a second maximally possible flow through the network is computed;
the first maximally possible flow through the network is normalized on the basis of the number of nodes which the network has along the first direction;
the second maximally possible flow through the network is normalized on the basis of the number of nodes which the network has along the second direction;
the normalized first maximally possible flow and the normalized second maximally possible flow are mutually compared and the value document is classified as suspected of forgery or not suspected of forgery in dependence on the smaller one of the two normalized maximally possible flows.

8. The method according to claim 1 , wherein for determining the respective capacity value (cij) of the respective two adjacent nodes (Ki, Kj) of the network, the measuring values (Ii, Ij) of these two nodes are compared to a target value (S)/ a target range (B) of the respective measuring value and the capacity value (cij) is chosen in dependence on the one of these two measuring values (Ii, Ij) which deviates more strongly from the target value (S)/ from the target range (B) than the other one of the two measuring values (Ii, Ij).

9. The method according to claim 8 , wherein the capacity value (cij) is chosen by assuming for the capacity value (cij) as a function of the measuring value that deviates more strongly from the target value/ from the target range a stepped function which has its maximal value (C) in a target range (B) surrounding the target value (S) and which, as a function of the measuring value deviating more strongly, declines in stepped fashion on one side or on both sides of the target range (B).

10. The method according to claim 9 , wherein for determining the respective capacity value (cij) of the two adjacent nodes of the network, if the more strongly deviating measuring value lies within the target range (B), a nominal capacity C>0 is utilized as capacity value (cij), and, if the more strongly deviating measuring value lies outside of the target range (B), a capacity c is utilized as capacity value (cij) which is smaller than the nominal capacity C, wherein 0<c<C.

11. The method according to claim 8 , wherein the position of the target value (S)/ of the target range (B) is chosen in dependence on an average measuring value of several measuring values detected on the value document.

12. The method according to claim 1 , wherein the measuring values are measuring values of the electromagnetic radiation that are detected by a spatially resolved electromagnetic measuring of the value document, in particular by a spatially resolved transmission, remission or luminescence measuring of the value document, e.g. in the visible, UV or IR spectral range.

13. The method according to claim 1 , wherein the measuring values are ultrasound measuring values detected through a spatially resolved ultrasound transmission or ultrasound remission measurement of the value document.

14. The method according to claim 1 , wherein the measuring values are combined measuring values, in which respectively at least two different measuring values of the value document are incorporated which wer

Assignees

• Giesecke & Devrient Gmbh

Inventors

• Feulner Johannes
• Domke Jan
• Holl Norbert

Classifications

• G07D11/50

  Sorting or counting valuable papers · CPC title

• G07D7/08 Primary

  Acoustic waves · CPC title

• G07D2207/00

  Paper-money testing devices · CPC title

• G01N29/04

  Analysing solids (using acoustic emission techniques G01N29/14) · CPC title

• G01N2291/048

  Transmission, i.e. analysed material between transmitter and receiver · CPC title