Systems and methods for assessing the authenticity of dynamic handwritten signature

• US 8,907,932 B2
• Filed: 10/21/2010
• Issued: 12/09/2014
• Est. Priority Date: 10/26/2009
• Status: Active Grant

• Current Assignee: Softwin Srl
• Original Assignee: Softwin Srl
• Inventors: Rusu, Mircea Sorin, Dinescu, Adrian, Diaconescu, Stefan Stelian
• Primary Examiner(s): Lao, Lun-Yi
• Assistant Examiner(s): Hegarty, Kelly B
• Application Number:
  US13/504,198
• Publication Number:
  US 20120212459A1
• Time in Patent Office: 1,510 Days
• Field of Search: 345/179, 345/158, 178 1901- 1905
• US Class Current: 345/179
• CPC Class Codes: G06F 3/0317 in co-operation with a patterned surface, e.g. absolute position or relative movement detection for an optical mouse or pen positioned with respect to a coded surface
• CPC Class Codes: G06F 3/03545 Pens or stylus
• CPC Class Codes: G06V 30/1423 the instrument generating sequences of position coordinates corresponding to handwriting preprocessing or recognising digital ink G06V30/32
• CPC Class Codes: G06V 40/30 Writer recognition; Reading and verifying signatures

Abstract

A system includes an electronic pen, functionally integrated with a personal computer for the acquisition and processing of signals associated with signatures, which can be network connected together with other personal computers. Each personal computer has connected as peripherals an electronic pen that includes two groups of inertial accelerometers that capture kinetic data and data about contact microvibrations with the writing support. The electronic pen also includes one self-referential optical navigation sensor that captures a series of data pairs, as momentary movements necessary in the reconstruction of the trajectory of the electronic pen and which, together with the kinetic data captured by the set of inertial accelerometers, represents personal computer input data for sensorial fusion processing and for creating the conditions to extract the information from the sensorial and psychomotric representation of a user'"'"'s perspective.

Claims

1. A system for acquiring and processing data associated with a dynamic handwritten signature, made on a writing paper support or any other support with similar texture having a quasi-uniform pattern printed thereon, the system comprising:

an electronic pen comprising two groups of inertial acceleration sensors that capture kinetic data and data about contact microvibrations with the writing support; and

an integrated personal computer for the acquisition and processing of the signals associated with the signatures, which can be network connected together with other personal computers, each personal computer having connected as peripherals an electronic pen, the electronic pen further comprising;

a self-referential optical navigation sensor that captures a series of data pairs (d_x, d_y), as momentary movements necessary for a reconstruction of a trajectory of the electronic pen, which, together with the kinetic data (a_x, a_y, b_x, b_y) captured by the set of inertial acceleration sensors constitute personal computer input data for sensorial fusion processing and for creating conditions to extract information from the sensorial and psychomotric representation of a user'"'"'s perspective;

an infrared LED to illuminate the pattern;

a lens that focuses and projects a current image of the pattern on a sensitive area of the optical navigation sensor,

wherein the lens is positioned so that the projection of a pattern region image is obtained on the sensitive area of the optical navigation sensor,

wherein a condition of spatio-temporal synchronism required for the acquisition through sensorial fusion of the series of data pairs d_x, d_y, of a graphic type, and kinetic data a_x, a_y, b_x, b_y is achieved by;

axial and topological quasi-alignment of three origins of coordinate axes of the set of inertial acceleration sensors, of the optical navigation sensor, and of a refill tip of the electronic pen;

topological plane-parallel quasi-alignment of coordinate/sensitive axes of the set of inertial acceleration sensors and of the optical navigation sensor with a writing plane; and

sampling the data captured by the set of inertial acceleration sensors and the optical navigation sensor, at a constant rate ranging between 1-8 ms, which assures, through high frequency, a time synchronism of the acquisition of graphic and kinetic details, the acquisition being controlled by an acquisition microcontroller that sends the information in real time to the personal computer that hosts algorithmic methods for visualization, for writing thickness reconstruction, and for processing and comparison, and

wherein, in order to capture the data from the psychometric and sensorial representation from the user'"'"'s perspective, the set of inertial acceleration sensors and the optical navigation sensor, integrated in the electronic pen handled by the user, capture the six associated signals a_x, a_y, b_x, b_y, d_x, d_y in self-referential coordinate systems, which are the x, y co-ordinates axes of sensitivity.

2. A method of assessing a dynamic handwritten signature authentication, the method consisting of:

acquiring a signature using the system claimed in claim 1;

determining characteristics of the signature;

comparing the characteristics of the signature with those stored in a database; and

combining the comparison results in order to assess the handwritten signature authentication,

wherein, in the acquiring step, a calligraphic reconstruction and a writing thickness visualization based on the signals a_x, a_y, b_x, b_y, d_x, d_y are performed, and

wherein, in the determining step, various processing of the signals a_x, a_y, b_x, b_y, d_x, d_y associated with the signature into a sequence of invariants C_i is performed.

3. The method according to claim 2, wherein the calligraphic reconstruction of handwriting thickness comprises the steps of:

calculating a trajectory by an algebraic summing of a series of values of the signals d_x and d_y, thus obtaining coordinates of momentary movements captured by an optical navigation sensor,

wherein the trajectory is graphically reproduced on a personal computer monitor;

filtering digitized signals a_x and a_y from an electronic pen with a high pass filter with a cutoff frequency of 150 Hz, thus resulting in two signals c_x and c_y;

computing a RMS value in a time period [n−

i, n] of i consecutive values, experimentally established, and incrementing with 1 after each stage of calculation, thus obtaining v_x(n) and v_y(n) signals, which, in turn, are arithmetically mediated for each incremental step, thus resulting in a m(n) signal proportional with the writing thickness;

determining a situation where segments of the trajectory are eliminated from an initial graphic representation, corresponding to zero thickness, by verifying a condition according to which any of the v_x(n), v_y(n) or

4. The method according to claim 2, named SRA3, wherein the sequence of invariants C_i is obtained starting from conversion of the signals d_x and d_y, which are momentary movements of the input signature that describe one plane curve.

5. The method according to claim 4, named SRA5, wherein the sequence of invariants C_i is obtained starting from conversion of both the signals a_x, a_y, and respectively, the signals b_x, b_y of the input signature, which describe two plane curves.

6. The method according to claim 2, named SRA7, sequence of invariants C_i is obtained starting from conversion of all of the signals a_x, a_y, b_x, b_y, d_x, d_y of the input signature, which describe one curve in six dimensional space.

7. The method according to claim 6, named SRA8, sequence of invariants C_i is obtained starting from conversion of the signals a_x, a_y, b_x, b_y of the input signature, which describe one curve in a four dimensional space.