Recordation of handwriting and hand movement using electromyography

• US 8,280,169 B2
• Filed: 12/19/2006
• Issued: 10/02/2012
• Est. Priority Date: 12/21/2005
• Status: Expired due to Fees

• Current Assignee: Michael Linderman
• Original Assignee: Michael Linderman
• Inventors: Linderman, Michael
• Primary Examiner(s): Akhavannik, Hadi
• Application Number:
  US11/640,954
• Publication Number:
  US 20070140562A1
• Time in Patent Office: 2,114 Days
• Field of Search: 382/186, 382/187
• US Class Current: 382/187
• CPC Class Codes: G06V 30/1423 the instrument generating sequences of position coordinates corresponding to handwriting preprocessing or recognising digital ink G06V30/32

Abstract

Handwriting activity is recorded by use of electromyography (EMG) signals detected from muscles at several locations on the hand. The EMG signals are sensed and registered. The sensed signals are processed and stored, after which the signals are analyzed to reconstruct handwriting activity into a digital format. Machine edible text is generated and displayed along with a graphical depiction of the handwriting.

Claims

1. A method to record handwriting activity, comprising:

registering electromyography (EMG) signals from one or more muscles at selected locations proximate the hand during handwriting activity;

wherein the EMG signals are full-wave rectified and band-pass filtered in the range of 0.2-20 HZ;

correlating the EMG signals with hand movement activity;

processing and storing said EMG signals in a memory for the purposes of handwriting recording;

analyzing said EMG signals and providing an output in a digital format; and

generating and providing an output of a correlation between the hand movements during handwriting and the EMG signals;

wherein one or more EMG sensor electrodes are configured to be disposed on a user'"'"'s hand;

wherein the EMG sensors are dry sensors;

wherein the EMG signals are sampled at a number of time lags with temporary time spacing between samples from several hand muscles and in different combinations;

wherein a pattern recognition algorithm consisting of a mathematical algorithm is trained to reconstruct handwriting patterns from the EMG signals;

wherein the mathematical algorithm is expressed by the equation V(t)=b+Σ

_r=mⁿw(τ

)n(t+τ

) +ε

(t);

wherein n(t+τ

) is a vector of input EMG signals at time t and time lag τ

;

V is the parameter of interest at time t;

w(τ

)is a vector of weights for each input at time lag τ

;

b is the y-intercept;

ε

(t) is the residual error;

wherein the equation is solved using linear least square regression;

and wherein the algorithm is trained by adiusting the parameters of the equation until handwriting patterns are reconstructed from the EMG signals.

2. Non-transitory-machine readable digital media including instructions for the steps as set forth in claim 1.

10. The method of claim 1 wherein the EMG signals are registered with in a glove, watch, wristband or similar article of apparel for housing the means for registration, processing, and storing of EMG signals to record the handwriting activity.

3. A system to record handwriting activity, comprising:

means for registration of electromyography (EMG) signals from one or more muscles at selected locations proximate the hand during handwriting activity;

means for correlating the EMG signals with hand movement activity;

wherein the EMG signals are full-wave rectified and band-pass filtered in the range of 0.2-20 HZ;

means for providing a digital output of a reconstruction of the EMG signals, the output selected from the group consisting of a digital representation of handwritten text, machine editable text, a drawing, and a digital correlation output; and

means for generating a display corresponding to the digital output;

wherein one or more EMG sensor electrodes are configured to be disposed on a user'"'"'s hand;

wherein the EMG sensors are dry sensors;

wherein the EMG signals are sampled at a number of time lags with temporary time spacing between samples from several hand muscles and in different combinations;

wherein a pattern recognition algorithm consisting of a mathematical algorithm is trained to reconstruct handwriting patterns from the EMG signals;

wherein the mathematical algorithm is expressed by the equation V(t)=b+Σ

<sub>τ

=m</sub>ⁿw(τ

)n(t+τ

)+ε

(t) ;

wherein n(t+τ

) is a vector of input EMG signals at time t and time lag τ

;

V is the parameter of interest at time t;

w(τ

)is a vector of weights for each input at time lag τ

;

b is the y-intercept;

ε

(t) is the residual error;

wherein the equation is solved using linear least square regression;

and wherein the algorithm is trained by adiusting the parameters of the equation until handwriting patterns are reconstructed from the EMG signals.

4. The system of claim 3 wherein said means for generating a display corresponding to the digital output generates a display of machine-editable text corresponding to the digital output and in the case of a drawing generates the drawing in an editable format.

5. The system of claim 3 wherein:

the means for registration of EMG signals comprises at least one sensor, electrode, amplifier, bandpass filter, multiplexer, analog to digital converter, and processing unit; and

the means for registration of EMG signals includes a routine to improve signal to noise ratio.

6. The system of claim 5 wherein the processing unit includes data processing equipment including a central processing unit and a memory unit suitable for data processing.

7. The system of claim 3 comprising a glove, watch, wristband or similar article of apparel for housing the means for registration, processing, and storing of EMG signals to record the handwriting activity.

8. The system of claim 3 wherein the means for registration comprises a processor provided as a chipset including at least one monolithic integrated circuit, the processor executing a routine that reconstructs the EMG signals.

9. A handwriting digital reconstruction system, comprising:

a plurality of electromyography (EMG) sensors configured to be disposed proximate a user'"'"'s hand during handwriting activity, each sensor being configured to transmit a signal representative of a physiological property of the user;

a processor;

a pattern recognition program operable by the processor to correlate signals representative of a physiological property of the user with hand movement activity of the user; and

data storage means storing program instructions that, when implemented by the processor, are configured to provide a digital representation of the user'"'"'s handwriting based on EMG signals transmitted during the handwriting activity;

wherein the EMG signals are full-wave rectified and band-pass filtered in the range of 0.2-20 HZ;

wherein one or more EMG sensor electrodes are configured to be disposed on a user'"'"'s hand;

wherein the EMG sensors are dry sensors;

wherein the EMG signals are sampled at a number of time lags with temporary time spacing between samples from several hand muscles and in different combinations;

wherein a pattern recognition algorithm consisting of a mathematical algorithm is trained to reconstruct handwriting patterns from the EMG signals;

wherein the mathematical algorithm is expressed by the equation V(t)=b+Σ

<sub>τ

=m</sub>w(τ

)n(t+τ

) +ε

(t);

wherein n(t+τ

) is a vector of input EMG signals at time t and time lag τ

;

V is the parameter of interest at time t;

w(τ

)is a vector of weights for each input at time lag ε

;

b is the y-intercept;

ε

(t) is the residual error;

wherein the equation is solved using linear least square regression;

and wherein the algorithm is trained by adiusting the parameters of the equation until handwriting patterns are reconstructed from the EMG signals.

11. The system of claim 9 wherein the plurality of electromyography (EMG) sensors configured to be disposed proximate a user'"'"'s hand by means of a glove, watch, wristband or similar article of apparel for housing the means for registration, processing, and storing of EMG signals to record the handwriting activity.