Dialing pulse signature recognizing method and device

• US 5,177,781 A
• Filed: 09/21/1990
• Issued: 01/05/1993
• Est. Priority Date: 08/07/1990
• Status: Expired due to Fees

• Current Assignee: Teleliaison
• Original Assignee: Teleliaison
• Inventors: Lemieux, Yves, Ponton, Michel, Girard, Normand, Gareau, Marc
• Primary Examiner(s): Dwyer, James L.
• Assistant Examiner(s): MATAR, AHMAD
• Application Number:
  US07/586,007
• Time in Patent Office: 837 Days
• Field of Search: 379/102, 379/104, 379/105, 379/281, 379/283, 379/339, 379/353, 379/386, 379/342, 379/31, 379/1, 379/201
• US Class Current: 379/339
• CPC Class Codes: H04M 11/007 with remote control systems remote control in general G08C
• CPC Class Codes: H04Q 1/30 Signalling arrangements; Manipulation of signalling currents multiplex systems providing for calling or supervisory signals H04J1/14, H04J3/12
• CPC Class Codes: H04Q 1/453 in which m-out-of-n signalling frequencies are transmitted
• CPC Class Codes: H04Q 1/50 Conversion between different kinds of signals

Abstract

A method and device for recognizing a signature of dialing pulse present in a communication signal isolate from the latter signal the signature. The so isolated signature is correlated with a statistic model obtained from a large number of dialing pulse signatures to determine whether the isolated signature corresponds to a dialing pulse. During this correlation, a first complex function representative of the isolated signature is calculated and correlated with a second complex function representative of the statistic model. The result of the correlation is normalized and whether the isolated signature is a signature of dialing pulse is decided in function of this normalization. The method and device can be used to convert into a tone frequency signal a series of dialing pulse signatures present in the communication signal and corresponding to a dialed digit. Each recognized signature is counted in order to determine the number of signatures in the series and a tone signal having a frequency which is function of this counted number is generated.

Claims

1. A method for recognizing a signature of dialing signal present in a communication signal, comprising the steps of:

storing a statistic model of said signature previously obtained from a large number of dialing signal signatures present in communication signals;

isolating said signature from said communication signal; and

correlating said isolated signature with said statistic model to determine whether said isolated signature corresponds to said dialing signal, said correlating step comprising calculating a first complex function representative of said isolated signature and correlating said first function with a second complex function representative of the statistic model.

2. A signature recognizing method as defined in claim 1, in which the signature isolating step comprises automatically adjusting the amplitude level of the isolated signature to a predetermined level.

3. A signature recognizing method as defined in claim 1, wherein said communication signal is an audio signal and said dialing signal is a dialing pulse, and wherein said signature isolating step comprises the steps of:

low-pass filtering the audio signal with a transition frequency of about 540 Hz;

automatically adjusting the low-pass filtered signal to a predetermined amplitude level;

high-pass filtering the signal at said predetermined amplitude level with a transition frequency of about 220 Hz;

full-wave rectifying the high-pass filtered signal; and

again low-pass filtering the full-wave rectified signal with a transition frequency of about 250 Hz.

4. A signature recognizing method as defined in claim 1, wherein the said step of correlating the isolated signature with a statistic model further comprises calculating the energy of said first function, normalizing the so calculated energy, normalizing the result of the correlation of the first function with the second function using the said normalized energy, and deciding in accordance with the result of the normalization of the correlation of the first function with the second function whether the isolated signature is a signature of dialing signal.

5. A signature recognizing method as defined in claim 4, wherein the isolated signature is an analog signature, wherein the first function calculating step comprises converting the analog signature into digital samples and calculating the first function using the following expression:

##EQU10## where x(n) are digital samples of the analog signature, ##EQU11##

6. A signature recognizing method as defined in claim 5, wherein the first function is correlated with the second function using the following expression:

##EQU12## and r(n)=n samples of a reference dialing signal signature obtained statistically from said large number of dialing signal signatures.

7. A signature recognizing method as defined in claim 6, in which the energy of the first complex function is calculated using the following expression:

##EQU13##

8. A signature recognizing method as defined in claim 7, wherein the normalized energy E_p² of the first complex function is related to at least one precedent value of the said normalized energy E_p².

9. A signature recognizing method as defined in claim 8, wherein the result of the correlation of the first function with the second function is normalized using the following expression:

##EQU14##

10. A device for recognizing a signature of dialing signal present in a communication signal, comprising:

means for storing a statistic model of said signature previously obtained from a large number of dialing signal signatures present in communication signals;

means for isolating said signature from said communication signal; and

means for correlating the isolated signature with said statistic model to determine whether said isolated signature corresponds to said dialing signal, said correlating means comprising means for calculating a first complex function representative of said isolated signature and means for correlating said first function with a second complex function representative of the statistic model.

11. A signature recognizing device as defined in claim 10, in which the signature isolating means comprises an automatic gain control circuit for adjusting the amplitude level of the isolated signature to a predetermined amplitude level.

12. A signature recognizing device as defined in claim 10, wherein said communication signal is an audio signal and said dialing signal is a dialing pulse, and wherein said signature isolating means comprises:

a first low-pass filter for filtering the audio signal, said first low-pass filter having a transition frequency of about 540 Hz;

an automatic gain control circuit for automatically adjusting the low-pass filtered signal to a predetermined amplitude level;

a high-pass filter for filtering the signal at said predetermined amplitude level, said high-pass filter having a transition frequency of about 220 Hz;

a full-wave rectifying means for rectifying the high-pass filtered signal; and

a second low-pass filter for filtering the full-wave rectified signal, said second low-pass filter having a transition frequency of about 250 Hz.

13. A signature recognizing device as defined in claim 10, wherein the said means for correlating the isolated signature with a statistic model comprises means for normalizing the result of the correlation of the first function with the second function, and means for deciding in accordance with the result of the normalization of the correlation of the first function with the second function whether the isolated signature is a signature of dialing signal.

14. A method of converting into a tone frequency signal a series of dialing pulse signatures present in a communication signal and corresponding to a dialed digit, comprising the steps of:

storing a statistic model of said dialing pulse signatures previously obtained from a large number of dialing pulse signatures present in communication signals;

isolating each pulse signature from said communication signal;

correlating each isolated signature with said statistic model to determine whether said isolated signature is effectively a dialing pulse signature, said correlating step comprising calculating a first complex function representative of said isolated signature and correlating said first function with a second complex function representative of the statistic model;

counting each isolated signature recognized as being effectively a dialing pulse signature to determine the number of signatures in the series; and

generating a tone frequency signal having a frequency which is function of the number of pulse signatures in the series.

15. A device for converting into a tone frequency signal a series of dialing pulse signatures present in a communication signal and corresponding to a dialed digit, comprising:

means for storing a statistic model of said dialing pulse signatures previously obtained from a large number of dialing pulse signatures present in communication signals;

means for isolating each pulse signature from said communication signal;

means for correlating each isolated signature with said statistic model to determine whether said isolated signature is effectively a dialing pulse signature, said correlating means comprising means for calculating a first complex function representative of said isolated signature and correlating said first function with a second complex function representative of the statistic model;

means for counting each isolated signature recognized as being effectively a dialing pulse signature to determine the number of pulse signatures in the series; and

means for generating a tone frequency signal having a frequency which is function of the number of pulse signatures in the series.