Behaviometric signature authentication system and method

• US 9,053,309 B2
• Filed: 03/14/2013
• Issued: 06/09/2015
• Est. Priority Date: 03/14/2013
• Status: Expired due to Fees

• Current Assignee: Artificial Intelligence Research Group Limited
• Original Assignee: Applied Neural Technologies Limited
• Inventors: Avni, Yossi, Suchard, Eytan
• Primary Examiner(s): Gelagay, Shewaye
• Assistant Examiner(s): LE, KHOI V
• Application Number:
  US13/831,140
• Publication Number:
  US 20140283005A1
• Time in Patent Office: 817 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 21/32 using biometric data, e.g. fingerprints, iris scans or voiceprints
• CPC Class Codes: G06F 21/64 Protecting data integrity, e.g. using checksums, certificates or signatures
• CPC Class Codes: G06F 2221/2113 Multi-level security, e.g. mandatory access control

Abstract

The present invention discloses a method of verifying the authenticity of a provided signature, comprising the steps of: receiving a set of sampled data points, each sampled data point being associated with a different position along the signature; identifying a set of characterizing nodes within the set of sampled data points using a set of predetermined characterizing nodes comprised in a pre-stored user profile; determining if each identified characterizing node lies within a predetermined threshold range of a corresponding predetermined characterizing node; and generating a positive verification when the characterizing nodes lie within the predetermined threshold range. A system arranged to carry out the method is also disclosed.

Claims

1. A method comprising:

receiving a set of sampled data points, each sampled data point being associated with a different position along a signature;

identifying a subset of nodes within the set of sampled data points using a set of predetermined nodes of a user profile;

identifying, with a processor, nodes of the user profile that correlate with nodes of the subset based on optimization matching;

determining if each identified node of the subset lies within a predetermined threshold range of a corresponding predetermined node;

calculating an acceleration value and a velocity value associated with line segments present between adjacent sampled data points comprised in the set of sampled data points;

defining a geometrical complexity rating of the provided signature on the basis of the calculated acceleration value and the calculated velocity value;

generating a positive verification when the nodes of the subset lie within the predetermined threshold range; and

rejecting the received signature when the defined geometrical complexity rating is below a minimum predetermined required geometrical complexity rating threshold,

wherein each sampled data point associated with a visible portion of the signature and each node of the subset is represented by a vector comprising a time coordinate and a spatial component, the spatial component being indicative of a relative position of the vector along the signature.

2. The method of claim 1, wherein each sampled data point is associated with a time coordinate value, and the said receiving comprises for each sampled data point:

calculating a time interval between the sampled data point and an adjacently located sampled data point, by comparing the time coordinate values associated with respectively the sampled data point and the adjacently located sampled data point;

determining if the time interval lies within a predetermined time interval threshold value; and

interpolating the position and time coordinate of one or more further data points located between the sampled data point and the adjacently located sampled data point when the calculated time interval exceeds the predetermined time interval threshold value, the interpolated position being selected such that the time interval between the sampled data point and the interpolated time coordinate associated with the one or more further data points lies within the predetermined time interval threshold value.

3. The method of claim 1, wherein receiving comprises:

calculating a distance of separation between a sampled data point and an adjacently located sampled data point;

determining if the distance of separation between the sampled data point and the adjacently located sampled data point lies within a predetermined distance interval threshold value; and

interpolating the position of one or more further data points located between the sampled data point and the adjacently located sampled data point, when the calculated distance of separation exceeds the predetermined distance interval threshold value, such that the distance of separation between the sampled and the interpolated position associated with the one or more further data points lies within the predetermined distance interval threshold value.

4. The method of claim 1, wherein the optimization matching comprises:

selecting a first predetermined node from the set of predetermined nodes;

calculating a vector dot product value between the selected first predetermined node and each sampled data point comprised within the set of sampled data points;

identifying the sampled data point associated with the largest vector dot product value as the data point that is most correlated with the first predetermined node, and designing the sampled data point based on such identification.

5. The method of claim 4, wherein the most correlated sampled data point is the data point which is oriented in substantially the same direction as the predetermined node, such that an angle of divergence θ

_j between the two vectors associated with respectively the predetermined node and the sampled data point is minimised.

6. The method of claim 1, wherein the optimization matching comprises using a matching function M_j to identify the sampled data point which is most correlated with the predetermined node, the matching function being a function of three differentiable functions F(θ

_j), G(d_j,d_j+1), and Q(r_j*d_j), where the following definitions apply;

θ

_j is the angle formed between the vector associated with the predetermined node and the vector associated with the sampled data point;

r_j is the scalar component of the vector associated with the predetermined node;

d_j is the scalar component of the vector associated with the sampled data point;

d_j+1 is the scalar component of the vector associated with an adjacent sampled data point;

F(θ

_j) and G(d_j,d_j+1) are positive and have an upper positive value; and

Q(r_j*d_j) is a convex function selected such that whilst it is monotonically increasing, its derivative monotonically decreases to zero.

7. The method of claim 6, wherein the matching function M_j is proportional to the product of the functions F(θ

_j), G(d_j,d_j+1), and Q(r_j*d_j), such that

M_j−

F(θ

_j)*G(d_j,d_j+1)*Q(r_j*d_j).

8. The method of claim 1 further comprising:

selecting a first one of the identified nodes of the subset;

calculating a geometric relationship of the selected first node with respect to one or more adjacently located identified nodes; and

wherein

said determining comprises verifying if each calculated geometric relationship lies within a predetermined threshold value range comprised in the pre-stored user profile; and

wherein

a positive verification result is generated when one or more calculated geometric relationships lie within the predetermined threshold value range.

9. The method of claim 8, wherein the geometric relationship is calculated between the identified node and each one of two adjacent, sequentially-located identified nodes, in order to define two different geometric relationships associated with the identified node.

10. The method of claim 9, wherein the geometric relationship is calculated between the identified node and each one of seven adjacent, sequentially-located identified nodes, in order to define seven different geometric relationships associated with the identified node.

11. The method of claim 9, wherein the number of identified nodes m is less than or equal to half the number of sampled data points n:

12. The method of claim 9, wherein the number of identified nodes m is less than or equal to a quarter the number of sampled data points n:

13. The method of claim 1 further comprising sampling the provided signature with a variable sampling rate, such that at least a portion of the sampled data points comprised in the set of received sampled data points are associated with different sampling rates.

14. The method of claim 1, wherein the method comprises:

generating a hash value on the basis of the set of sampled data points;

comparing the generated hash value with a set of pre-stored hash values to determine if the generated hash value is unique; and

wherein

a positive verification result is generated when the generated hash value is unique.

15. The method of claim 14, wherein sampling comprises normalising the provided signature.

16. The method of claim 1, wherein the method further comprises:

calculating a lapsed time interval between each identified node;

determining if the calculated time lapse value lies within a predetermined threshold value range associated with the pre-stored user profile; and

generating the positive verification result when the calculated time lapse value lies within the predetermined threshold value range.

17. The method of claim 1 further comprising:

calculating a velocity vector for each identified node, using spatial coordinates and a temporal coordinate associated with each node;

determining if each calculated velocity vector lies within a predetermined threshold value range comprised in the pre-stored user profile; and

generating the positive verification result when the calculated velocity vectors lie within the predetermined threshold value range.

18. The method of claim 1 further comprising:

calculating an acceleration vector for each identified node, using spatial coordinates and a temporal coordinate associated with each identified node;

determining if each calculated acceleration vector lies within a predetermined threshold value range comprised in the pre-stored user profile; and

generating the positive verification result when the calculated acceleration vectors lie within the predetermined threshold value range.

19. The method of claim 1 comprising:

maintaining a record of node values that resulted in positive verification results, the node values being associated with a plurality of different received sets of sampled data points associated with different copies of the same signature;

calculating a statistical variance between the node values and the corresponding predetermined nodes for each different provided copy of the same signature; and

amending the predetermined threshold value range of the corresponding predetermined node to be consistent with the calculated statistical variance.

20. The method of claim 19, wherein the statistical variance is calculated using node values that resulted in positive verification results associated with different copies of the same signature provided over the course of a time period.

21. The method of claim 1 further comprising authorizing a transaction between two remotely located entities based on said positive verification.

22. A system for verifying the authenticity of a provided signature, the system comprising:

an input device arranged to receive a set of sampled data points, each sampled data point being associated with a different position along the signature;

a processor;

instructions accessible and executable by the processor, said instructions comprising;

identifying a set of characterising nodes within the set of sampled data points using a set of predetermined characterising nodes comprised in a pre-stored user profile;

determining if each identified characterising node lies within a predetermined threshold value range of a corresponding predetermined characterising node;

calculating first order and second order derivatives associated with line segments present between adjacent sampled data points comprised in the set of sampled data points;

defining a geometrical complexity rating of the provided signature on the basis of the calculated first and second order derivatives;

rejecting the received signature when the defined geometrical complexity rating is below a minimum predetermined required geometrical complexity rating threshold; and

generating a positive verification result when the characterising nodes lie within the predetermined threshold value range,

wherein each sampled data point associated with a visible portion of the signature and each node of the subset is represented by a vector comprising a time coordinate and a spatial component, the spatial component being indicative of a relative position of the vector along the signature.

23. The system of claim 22, wherein the instructions further comprise obtaining the set of predetermined characterising nodes comprised in the pre-stored user profile, and identifying the sampled data point which is most correlated with each predetermined characterising node from the set of sampled data points, using optimization matching.

24. The system of claim 23, wherein the instructions further comprise:

selecting a first predetermined characterising node from the set of predetermined characterising nodes;

calculating a vector dot product value between the selected first predetermined characterising node and each sampled data point comprised within the set of sampled data points; and

identifying the sampled data point associated with the largest vector dot product value as the data point that is most correlated with the first predetermined characterising node and designating the sampled data point as a characterising node comprised within the set of identified characterising nodes.

25. The system of claim 24, wherein the instructions further comprise identifying the most correlated sampled data point as the data point which is oriented in substantially the same direction as the predetermined characterising node, such that an angle of divergence θ

_j between the two vectors associated with respectively the characterising node and the sampled data point is minimised.

26. The system of claim 25, wherein the instructions further comprise using a matching function M_j to identify the sampled data point which is most correlated with the predetermined characterising node, the matching function being a function of three differentiable functions F(θ

_j), G(d_j,d_j+1), and Q(r_j*d_j), where the following definitions apply;

θ

_j is the angle formed between the vector associated with the predetermined characterising node and the vector associated with the sampled data point;

r_j is the scalar component of the vector associated with the predetermined characterising node;

d_j is the scalar component of the vector associated with the sampled data point;

d_j+1 is the scalar component of the vector associated with an adjacent sampled data point;

F(θ

_j) and G(d_j,d_j+1) are positive and have an upper positive value; and

Q(r_j*d_j) is a convex function selected such that whilst it is monotonically increasing, its derivative monotonically decreases to zero.

27. The system of claim 26, wherein the matching function M_j is proportional to the product of the functions F(θ

_j), G(d_j,d_j+1) and Q(r_j*d_j), such that M_j=F(θ

_j)*G(d_j,d_j+1)*Q(r_j*d_j).

28. The system of claim 22 further comprising an interpolator operatively coupled to the input device, and arranged to interpolate one or more data points.

29. The system of claim 28, wherein the input device is arranged to calculate a time interval between a sampled data point and an adjacently located sampled data point, by comparing a time coordinate value associated with respectively the sampled data point and the adjacently located sampled data point, and determine if the time interval lies within a predetermined time interval threshold value;

and

the interpolator is arranged to interpolate the position and time coordinate of one or more further data points located between the sampled data point and the adjacently located sampled data point when the calculated time interval exceeds the predetermined time interval threshold value, the interpolator being arranged to interpolate the interpolated position such that the time interval between the sampled data point and the interpolated time coordinate associated with the one or more further data points lies within the predetermined time interval threshold value.

30. The system of claim 29, wherein the input device is arranged to calculate a distance of separation between a sampled data point and an adjacently located sampled data point, determine if the distance of separation between the sampled data point and the adjacently located sampled data point lies within a predetermined distance interval threshold value;

and

the interpolator is arranged to interpolate the position of one or more further data points located between the sampled data point and the adjacently located sampled data point, when the calculated distance of separation exceeds the predetermined distance interval threshold value, such that the distance of separation between the sampled and the interpolated position associated with the one or more further data points lies within the predetermined distance interval threshold value.

31. The system of claim 22, wherein the instructions further comprise selecting a first one of the identified characterising nodes, calculating a geometric relationship of the selected first node with respect to one or more adjacently located identified characterising nodes, determining if each calculated geometric relationship lies within a predetermined threshold value range comprised in the pre-stored user profile, and generating a positive verification result when one or more calculated geometric relationships lie within the predetermined threshold value range.

32. The system of claim 22 further comprising:

a sampling device operatively coupled to the input device, the sampling device being arranged to sample the provided signature with a variable sampling rate, such that at least a portion of the sampled data points comprised in the set of sampled data points received by the input device are associated with different sampling rates.

33. The system of claim 32, wherein the sampling device is arranged to generate a hash value on the basis of the set of sampled data points;

and

the instructions further comprising comparing the generated hash value with a set of pre-stored hash values to determine if the generated hash value is unique, and generating a positive verification result when the generated hash value is unique.

34. The system of claim 32, wherein the sampling device is arranged to normalise the provided signature.

35. The system of claim 22, wherein the input device and the processor are remote to each other.

36. The system of claim 35, wherein the processor is at a server located remotely to the input, and the server is operatively coupled to the input via a communication channel.

37. The system of claim 35, wherein the input device comprises a touch-pad arranged to receive a signature.

38. The system of claim 35, wherein the input device comprises a mobile telephone provided with a touch-sensitive screen.

39. The system of claim 35, wherein the input device comprises a personal computer.

40. The system of claim 35, wherein the input device comprises a tablet computer.

41. The system of claim 22, the instructions further comprise controlling access to a secure resource.

42. The system of claim 41, wherein the secure resource is a bank account.

43. The system of claim 41, wherein the secure resource is an automobile configured with a touch-pad arranged to receive the signature.

44. The system of claim 41, wherein the instructions further comprise controlling a transaction between two remotely located entities.

45. The system of claim 44, wherein the transaction is a financial transaction.

46. The system of claim 44, wherein the two remotely located entities comprise a payee and a recipient.