Method for dynamic reconstruction of handwritten data

• US 5,933,514 A
• Filed: 02/23/1996
• Issued: 08/03/1999
• Est. Priority Date: 08/20/1992
• Status: Expired due to Term

• Current Assignee: Communications Intelligence Corporation
• Original Assignee: Communications Intelligence Corporation
• Inventors: Austin, Norman A., Ostrem, John S., Crane, Hewitt D.
• Primary Examiner(s): Kelley, Christopher S.
• Application Number:
  US08/605,972
• Time in Patent Office: 1,257 Days
• Field of Search: 382/119, 382/120, 382/121, 382/122, 382/123, 382/187, 382/189, 382/293, 382/276
• US Class Current: 382/122
• CPC Class Codes: G06V 40/30 Writer recognition; Reading and verifying signatures
• CPC Class Codes: G07C 9/35 by means of a handwritten signature

Abstract

A method is provided for reconstructing data generated on a digitizer tablet based on the dynamics of the data generation. The digitizer tablet samples the position of the writing pen, continuously transmitting data to a computer in the form of x and y coordinates plus an indicator of whether or not the pen is touching the surface of the tablet. After preprocessing, the data is reconstructed based on the dynamics of the data'"'"'s creation. The values of the dynamic reconstruction are created by accumulating increments that are some function of the original data. To enhance the dynamics a nonlinear function is required.

Claims

1. A method for authenticating handwritten data using a computer, the computer comprising a processor, a memory, and a digitizer tablet, the method comprising:

capturing a stroke of the handwritten data as a first sequence of data points as a function of time, in response to a physical movement of a pen on the digitizer tablet, the data points in the first sequence having positional coordinates;

forming a second sequence of data points by determining positional coordinates of a previous data point in the second sequence of data points and the spatial differences between adjacent data points in the first sequence according to the following relationship;

space="preserve" listing-type="equation">X.sub.i =X.sub.i-1 +(x.sub.i -x.sub.i-1).sup.a,

space="preserve" listing-type="equation">Y.sub.i =Y.sub.i-1 +(y.sub.i -y.sub.i-1).sup.b,

where

x_i and x_i-1 over the range i=0 to i=N-1 are positional coordinates of data points in the first sequence of data points,

y_i and y_i-1 over the range i=0 to i=N-1 are positional coordinates of data points in the first sequence of data points,

X_i-1 and Y_i-1 over the range i=0 to i=N-1 are the positional coordinates of the previous data points in the second sequence of data points,

X_i and Y_i over the range i=0 to i=N-1 are the positional coordinates of the data points in the second sequence of data points,

a is any real number excluding -1, 0, or 1,

b is any real number excluding -1, 0, or 1, and

N is a number of positional data points in the first sequence; and

authenticating the handwritten data on the basis of the second sequence of data points.

10. The method according to claim 1 wherein the first sequence of data points from the digitizer tablet is a sequence of x and y positional coordinate values in a Cartesian coordinate system.

11. The method according to claim 1, further comprising the step of spacing out individual strokes of a dynamic reconstruction of strokes.

2. A method for comparing a stroke of an unverified handwriting sample to a stroke of a verified handwriting sample comprising:

capturing a parametric representation of the stroke of the unverified handwriting sample produced by a physical movement of a pen on a digitizer tablet as a first sequence of positional data points as a function of time;

determining the spatial differences between adjacent positional data points of the first sequence;

constructing the stroke of the unverified handwriting sample as a second sequence of positional data points in response to the spatial differences between adjacent positional data points of the first sequence, according to the following relationship;

space="preserve" listing-type="equation">X.sub.i =X.sub.i-1 +(x.sub.i -x.sub.i-1).sup.a,

space="preserve" listing-type="equation">Y.sub.i =Y.sub.i-1 +(y.sub.i -y.sub.i-1).sup.b,

where

x_i -x_i-1 and y_i -y_i-1 over the range i=0 to i=N-1 are the spatial differences between positional data points in the first sequence,

X_i and Y_i over the range i=0 to i=N-1 are positional data points in the second sequence,

a is any real number excluding -1, 0, or 1,

b is any real number excluding -1, 0, or 1, and

N is a number of positional data points in the first sequence; and

comparing the constructed stroke of the unverified handwriting sample with a constructed stroke of the verified handwriting sample.

12. The method according to claim 2 wherein the first sequence of data points is a sequence of x and y coordinate values in a Cartesian coordinate system.

13. The method according to claim 2, further comprising the step of spacing out individual strokes of the dynamic reconstruction.

3. A method for authentication using a computer to dynamically reconstruct a representation of handwritten data as a dynamic reconstruction, the computer comprising a processor, a memory, and a digitizer tablet, the method comprising:

capturing handwritten data produced by a physical movement of a pen on the digitizer tablet, and using the processor to create a parametric representation of the handwritten data as a first sequence of positional data points as a function of time;

enhancing dimensional components of the parametric representation with temporal components of the parametric representation to produce the dynamic reconstruction in the form of a second sequence of positional data points according to the following relationship;

space="preserve" listing-type="equation">X.sub.i =X.sub.i-1 +(x.sub.i -x.sub.i-1).sup.a,

space="preserve" listing-type="equation">Y.sub.i =Y.sub.i-1 +(y.sub.i -y.sub.i-1).sup.b,

where

x_i and y_i over the range i=0 to i=N-1 is a discrete representation of a stroke,

X_i and Y_i over the range i=0 to i=N-1 is a discrete representation of the dynamic reconstruction of the stroke,

a is any real number excluding -1, 0, or 1,

b is any real number excluding -1, 0, or 1, and

N is a number of points in the stroke; and

authenticating the handwritten data on the basis of the dynamic reconstruction of the handwriting data.

4. A method for comparing an unverified handwriting sample to a verified handwriting sample comprising:

capturing a parametric representation of the unverified handwriting sample produced by a physical movement of a pen on a digitizer tablet as a first sequence of positional data points as a function of time;

computing the linear distances between adjacent data points of the first sequence;

creating the unverified handwriting sample as a second sequence of positional data points from the first sequence of positional data points according to the following relationship;

space="preserve" listing-type="equation">X.sub.i =X.sub.i-1 +(x.sub.i -x.sub.i-1).sup.a,

space="preserve" listing-type="equation">Y.sub.i =Y.sub.i-1 +(y.sub.i -y.sub.i-1).sup.b,

where

x_i -x_i-1 and y_i -y_i-1 over the range i=0 to i=N-1 are the linear distances between positional data points in the first sequence,

X_i and Y_i over the range i=0 to i=N-1 is the second sequence of the positional data points,

a is any real number excluding -1, 0, or 1,

b is any real number excluding -1, 0, or 1, and

N is a number of points in the stroke; and

comparing the created unverified handwriting sample with a created verified handwriting sample.

5. A computer program product for authenticating handwritten data, the computer program product comprising:

a computer-readable memory including;

code that captures a stroke of the handwritten data as a first sequence of data points as a function of time, in response to a physical movement of a pen on the digitizer tablet, the data points in the first sequence having positional coordinates;

code that forms a second sequence of data points by determining positional coordinates of a previous data point in the second sequence of data points and the spatial differences between adjacent data points in the first sequence according to the following relationship;

space="preserve" listing-type="equation">X.sub.i =X.sub.i-1 +(x.sub.i -x.sub.i-1).sup.a,

space="preserve" listing-type="equation">Y.sub.i =Y.sub.i-1 +(y.sub.i -y.sub.i-1).sup.b,

where

x_i and x_i-1 over the range i=0 to i=N-1 are positional coordinates of data points in the first sequence of data points,

y_i and y_i-1 over the range i=0 to i=N-1 are positional coordinates of data points in the first sequence of data points,

X_i-1 and Y_i-1 over the range i=0 to i=N-1 are the positional coordinates of the previous data points in the second sequence of data points,

X_i and Y_i over the range i=0 to i=N-1 are the positional coordinates of the data points in the second sequence of data points,

a is any real number excluding -1, 0, or 1,

b is any real number excluding -1, 0, or 1, and

N is a number of positional data points in the first sequence; and

code that authenticates the handwritten data on the basis of the second sequence of data points.

6. A computer program product for comparing a stroke of an unverified handwriting sample to a stroke of a verified handwriting sample comprising:

a computer-readable memory including;

code that directs a processor to capture a parametric representation of the stroke of the unverified handwriting sample, produced by a physical movement of a pen on a digitizer tablet, as a first sequence of positional data points as a function of time;

code that directs the processor to apply a process to positional data points of the first sequence to obtain a dynamic reconstruction of the stroke of the unverified handwriting sample as a second sequence of positional data points according to the following relationship;

space="preserve" listing-type="equation">X.sub.i =X.sub.i-1 +(x.sub.i -x.sub.i-1).sup.a,

space="preserve" listing-type="equation">Y.sub.i =Y.sub.i-1 +(y.sub.i -y.sub.i-1).sup.b,

where

x_i and y_i over the range i=0 to i=N-1 are positional data points in the first sequence,

X_i and Y_i over the range i=0 to i=N-1 are positional data points in the second sequence,

a is any real number excluding -1, 0, or 1,

b is any real number excluding -1, 0, or 1, and

N is a number of positional data points in the stroke; and

code that directs the processor to compare the dynamic reconstruction of the stroke of the unverified handwriting sample with a dynamic reconstruction of the stroke of the verified handwriting sample.

7. A computer system for comparing a stroke of an unverified handwriting sample to a stroke of a verified handwriting sample, the computer system including a digitizer tablet, the computer system further comprising:

a memory for capturing the unverified handwriting sample produced by a physical movement of a pen on the digitizer tablet; and

a processor for creating a parametric representation of a stroke of the unverified handwriting sample as a first sequence of positional data points as a function of time, for computing the linear distances between adjacent data points of the first sequence, for enhancing dimensional components of the parametric representation with temporal components of the parametric representation to produce a dynamic reconstruction of the stroke of the unverified handwriting sample in the form of a second sequence of positional data points according to the following relationship;

space="preserve" listing-type="equation">X.sub.i =X.sub.i-1 +(x.sub.i -x.sub.i-1).sup.a,

space="preserve" listing-type="equation">Y.sub.i =Y.sub.i-1 +(y.sub.i -y.sub.i-1).sup.b,

where

x_i and y_i over the range i=0 to i=N-1 are positional data points in the first sequence,

X_i and Y_i over the range i=0 to i=N-1 are positional data points in the second sequence,

a is any real number excluding -1, 0, or 1,

b is any real number excluding -1, 0, or 1, and

N is a number of positional data points in the first sequence, and

for comparing the dynamic reconstruction of the stroke of the unverified handwriting sample with a dynamic reconstruction of a stroke of the verified handwriting sample.

8. A method of authenticating handwritten data using a computer, the computer having a processor, a memory, and a digitizer tablet, the method comprising:

capturing a stroke of the handwritten data as a first sequence of data points as a function of time, said capturing being performed in response to a physical movement of a pen on the digitizer tablet, the data points in the first sequence having positional coordinates;

generating a second sequence of data points based on the linear distances between the data points in the first sequence such that the second sequence of data points is altered along substantially the same direction as defined by the first sequence of data points; and

authenticating the handwritten data on the basis of the second sequence of data points.

14. The method of claim 8 wherein the second sequence of data points is altered by increasing the linear distances between the data points in the second sequence.

15. The method of claim 8 wherein the second sequence of data points is altered by decreasing the linear distances between the data points in the second sequence.

9. A method of authenticating handwritten data using a computer, the computer having a processor, a memory, and a digitizer tablet, the method comprising:

capturing a stroke of the handwritten data as a first pair of data points as a function of time, said capturing being performed in response to a physical movement of an input device on the digitizer tablet, the first pair of data points having positional coordinates;

calculating a linear distance between the first pair of data points;

generating a second pair of data points from the first pair of data points by increasing or decreasing the linear distance between the first pair of data points to generate the second pair of data points along substantially the same direction as the first pair of data points; and

authenticating the handwritten data on the basis of the second pair of data points.

16. The method of claim 9 wherein authenticating the handwritten data includes displaying the second pair of data points.

17. The method of claim 9 wherein authenticating the handwritten data includes comparing the second pair of data points with a reference pair of data points.