Portable data storage medium for biometric user identification

• US 8,255,699 B2
• Filed: 12/06/2007
• Issued: 08/28/2012
• Est. Priority Date: 12/08/2006
• Status: Active Grant

• Current Assignee: Giesecke + Devrient Mobile Security GmBH (Giesecke & Devrient GmbH)
• Original Assignee: Giesecke & Devrient GmbH
• Inventors: Tagscherer, Michael
• Primary Examiner(s): Barron, Jr., Gilberto
• Assistant Examiner(s): Cribbs, Malcolm
• Application Number:
  US12/518,264
• Publication Number:
  US 20100054547A1
• Time in Patent Office: 1,727 Days
• Field of Search: 726 1- 4, 726 16- 19, 726 26- 28, 713/186, 382115-127
• US Class Current: 713/186
• CPC Class Codes: G06F 18/2137 based on criteria of topology preservation, e.g. multidimensional scaling or self-organising maps
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06V 40/10 Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands

Abstract

A portable data carrier (1) comprises an executable training module (8) which provides in a memory (4, 5) of the data carrier (1) a reference model (10; M) for the biometric recognition of a user of the data carrier (1) by a recognition module (9), whereby a biometric comparison vector (16) which lies within the acceptance range (A) of the reference model (10; M) is accepted by the recognition module (9) as coming from the user. In so doing, the training module (8) defines the reference model (10; M) by model nodes (P, P0-P4) which are formed from biometric reference vectors (17; N) in each case coming from the user, whereby a new model node (P, P0-P4) is added to the reference model (10; M) when the associated reference vector (17; N) lies outside the acceptance range (A) of the reference model (10; M).

Claims

1. A method for providing a reference model having an acceptance range on a portable data carrier for biometrically recognizing a user of the data carrier, wherein a biometric comparison vector which lies within the acceptance range of the reference model is accepted as coming from the user, comprising the steps of:

defining, by a processor of the portable data carrier, the reference model by model nodes which are formed from biometric reference vectors in each case coming from the user, during an initial training phase, wherein for generating the reference model a multitude of biometric reference vectors is recorded, and wherein the reference model is defined piecewise by local models each having a validity range, and adding a model node to the reference model, during an application phase of the reference model upon checking, by the processor of the portable data carrier, a comparison vector when the comparison vector lies outside the acceptance range of the reference model and it is certain that the comparison vector actually comes from the user,

wherein a local model is formed by at least one model node, and that for recognizing the user that local model of the reference model is used in whose validity range a biometric comparison vector of the user lies,

wherein each local model is assigned in each case a validity range which is determined by a probability distribution which is assigned to the at least one model node forming the local model, wherein the width of the probability distribution of the model node results from the probability distributions of neighboring model nodes, and that the probability distribution assigned to a model node indicates a recognition quality of a biometric comparison vector accepted as coming from the user which lies within the validity range of that local model which is formed by at least the model node.

2. The method according to claim 1, wherein each local model is assigned in each case a tolerance range, the tolerance ranges of all local models forming the acceptance range of the reference model, and that a biometric comparison vector which lies in the tolerance range of a local model is accepted as coming from the user.

3. The method according to claim 1, wherein, upon addition of a model node to the reference model the widths of the probability distribution of the model node and of the probability distributions of neighboring model nodes are adjusted to each other.

4. The method according to claim 1, wherein when the number of model nodes forming a local model exceeds a predetermined maximum number, a further model node is not added to this local model and/or the local model is split into two new local models.

5. The method according to claim 1, wherein a decision about whether a biometric comparison vector is a biometric reference vector coming from the user is effected by an authorized body outside the data carrier.

6. A portable data carrier comprising a memory, a communication interface, a processor, and a recognition module and a training module each executable by the processor, wherein the training module is configured to provide in the memory a reference model having an acceptance range for the biometric recognition of a user of the data carrier by the recognition module, wherein a biometric comparison vector which lies within the acceptance range of the reference model is accepted by the recognition module as coming from the user;

wherein the training module is configured to define the reference model by model nodes during an initial training phase,

wherein for generating the reference model a multitude of biometric reference vectors is recorded, which in each case are formed from biometric reference vectors received via the communication interface and coming from the user,

wherein the training module during an application phase of the reference model for checking a comparison vector adds a model node to the reference model when the comparison vector lies outside the acceptance range of the reference model and it is certain that the comparison vector actually comes from the user,

wherein the training module is configured to piecewise define the reference model by local models each having a validity range, to form a local model by at least one model node, and to store the local models in the memory,

wherein the recognition module is configured to use that local model of the reference model in whose validity range a biometric comparison vector of the user received via the communication interface lies,

wherein the training module is configured to assign to a local model a validity range which is determined by a probability distribution which is assigned to the at least one model node forming the local model, and to fix the width of the probability distribution of the model node in dependence on the probability distributions of neighboring model nodes, and

wherein the recognition module is configured to ascertain from the probability distribution assigned to a model node a recognition quality of a biometric comparison vector received via the communication interface and accepted as coming from the user which lies within the validity range of that local model which is formed by at least the model node.

7. The data carrier according to claim 6,

wherein the training module is configured to assign a tolerance range to each local model and to form the acceptance range of the reference model from the tolerance ranges of all local models, and

wherein the recognition module is configured to accept as coming from the user a biometric comparison vector, received via the communication interface, which lies within the tolerance range of a local model.

8. The data carrier according to claim 6, wherein the training module is set up such that upon addition of a model node to the reference model it adjusts to each other the widths of the probability distribution of the model node and of the probability distributions of neighboring model nodes.

9. The data carrier according to claim 6, wherein the training module is configured to not add a further model node to the local model and/or to split the local model into two new local models, and to store them in the memory, when the number of model nodes forming a local model exceeds a predetermined maximum number stored in the memory.

10. The data carrier according to claim 6, wherein the training module is configured to store in the memory, as a model node to be added to the reference model, the associated biometric reference vector of the user as a new model node.

11. The data carrier according to claim 6,

wherein the recognition module is configured to carry out a recognition of a biometric comparison vector by matching the comparison vector with a reference model of the user stored in the memory;

and

wherein the training module is configured to generate the reference model within the framework of an initial training phase of the recognition module from a plurality of reference vectors received via the communication interface and to store it in the memory.

12. The data carrier according to claim 6,

wherein the recognition module is configured to carry out a recognition of a biometric comparison vector by matching the comparison vector with an adaptive reference model of the user stored in the memory;

and

wherein the training module is configured to continuously adjust the reference model within the framework of an application phase of the recognition module with biometric reference vectors received via the communication interface and to store it in the memory.

13. The data carrier according to claim 6, wherein the training module is configured such that upon addition of a model node to the reference model it gets information from the recognition module that the associated reference vector lies outside the acceptance range of the reference model, and receives information from an authorized body via the communication interface that the associated reference vector comes from the user.

14. The data carrier according to claim 6, wherein the data carrier is a chip card personalized for the user.