Authentication of security documents and mobile device to carry out the authentication

Patent metadata

Field	Value
Publication number	US-9947163-B2
Application number	US-201414764061-A
Country	US
Kind code	B2
Filing date	Feb 4, 2014
Priority date	Feb 4, 2013
Publication date	Apr 17, 2018
Grant date	Apr 17, 2018

Abstract

A method of authenticating security documents and a mobile device, especially a smartphone, programmed to carry out the method, based on an analysis of features which are produced by intaglio printing, which analysis involves a decomposition of sample images of a candidate document to be authenticated based on Wavelets, each sample image being digitally processed by performing a Wavelet transform of the sample image in order to derive a set of classification features. The method is based on an adaptive approach, which includes the following steps: —prior to carrying out the Wavelet transform, defining a categorization map containing local information about different intaglio line structures that are found on the security documents; —carrying out a Wavelet selection amongst a pool of Wavelet types based on the categorization map; and —performing the Wavelet transform of the sample image on the basis of the selected Wavelet.

First claim

The invention claimed is:

1. A method of authenticating security documents based on an analysis of intrinsic features of the security documents which are produced by intaglio printing, which analysis involves a decomposition of one or more sample images of at least a part of a candidate document to be authenticated based on Wavelets, each sample image being digitally processed by performing a Wavelet transform of the sample image in order to derive a set of classification features allowing a classification of the candidate document within a multidimensional feature space,
wherein the method is based on an adaptive Wavelet approach, which adaptive Wavelet approach includes the following steps:
prior to carrying out the Wavelet transform, defining a categorization map containing local information about different intaglio line structures that are found on the security documents and allocating a pool of Wavelet types to the categorization map;
carrying out a Wavelet selection among the pool of Wavelet types based on the categorization map; and
performing the Wavelet transform of the sample image on the basis of the selected Wavelet,

wherein the step of defining the categorization map includes defining a statistical model of each given intaglio line structure, wherein the statistical model is built from measurements of a line width and of a line distance within each given intaglio line structure, wherein the statistical model includes a 4-tuple of parameters characterizing four histograms representative of each given intaglio line structure, and wherein the four histograms comprise: a histogram of the statistical distribution of line widths in a horizontal direction, a histogram of the statistical distribution of line distances in the horizontal direction, a histogram of the statistical distribution of line widths in a vertical direction, and a histogram of the statistical distribution of line distances in the vertical direction.

2. The method according to claim 1 , wherein the at least one parameter is a shape parameter describing a shape of the corresponding histogram.

3. The method according to claim 1 , wherein the at least one parameter is determined on the basis of a Maximum Likelihood Estimation (MLE) approach.

4. The method according to claim 1 , wherein the pool of Wavelet types includes a baseline Wavelet which is used as baseline for the Wavelet selection.

5. The method according to claim 4 , wherein the baseline Wavelet is the db2-Wavelet.

6. The method according to claim 4 , wherein the baseline Wavelet is replaced by another Wavelet type, if a separation ability of that other Wavelet type in the feature space, for a given intaglio line structure, is better than that of the baseline Wavelet.

7. The method according to claim 1 , wherein the set of classification features includes statistical moments descriptive of a statistical distribution, or histograms, of Wavelet coefficients resulting from the Wavelet transform.

8. The method according to claim 7 , wherein the set of classification features includes the variance, the skewness and the excess.

9. The method according to claim 1 , applied in a mobile device environment.

10. A mobile device comprising an image processing unit programmed to carry out the method according to claim 1 .

11. The method according to claim 9 , applied in a smartphone.

12. The method according to claim 1 , wherein the security documents are banknotes.

13. The mobile device according to claim 10 , wherein the mobile device is a smartphone.

Assignees

• Kba Notasys Sa

Inventors

• Lohweg Volker
• Hoffmann Jan Leif
• Dörksen Helene
• Hildebrand Roland
• Gillich Eugen
• Hofmann Jürg
• Schaede Johannes Georg

Classifications

• G06V10/52

  Scale-space analysis, e.g. wavelet analysis (multi-scale boundary representations G06V10/42) · CPC title

• G06F2218/06

  by applying a scale-space analysis, e.g. using wavelet analysis · CPC title

• G06V30/19173

  Classification techniques · CPC title

• G06V30/40

  Document-oriented image-based pattern recognition · CPC title

• G06T7/168

  involving transform domain methods · CPC title