Apparatus and method for identifying speech and call-progression signals

• US 5,319,703 A
• Filed: 05/26/1992
• Issued: 06/07/1994
• Est. Priority Date: 05/26/1992
• Status: Expired due to Fees

• Current Assignee: VMX, Inc. (Innospec Incorporated)
• Original Assignee: VMX, Inc. (Innospec Incorporated)
• Inventors: Drory, Eatamar
• Primary Examiner(s): Dwyer, James L.
• Assistant Examiner(s): SAINT SURIN, JACQUES M
• Application Number:
  US07/889,513
• Time in Patent Office: 742 Days
• Field of Search: 379/84, 379/386, 379/351, 379/86, 381/46, 381/41, 381/110, 370/110.3
• US Class Current: 379/351
• CPC Class Codes: H04Q 1/46 comprising means for distinguishing between a signalling current of predetermined frequency and a complex current containing that frequency, e.g. speech current

Abstract

A method for real-time recognition of speech and call-progression tone energy includes the steps of digitizing an analog signal present on a signal line, correlating N neighboring portions of the digitized signal, performing FFT analysis on the digitized signal, identifying the three largest frequency-domain maxima, determining whether the two largest maxima are above a threshold frequency, and determining whether the ratio of the largest to the third largest maxima exceeds a predetermined value. According to the present invention, the steps of the method may be performed in real time using fixed-point hardware by approximating the correlation and FFT functions.

Claims

1. A method for performing real-time recognition of speech and call-progression tones from electrical energy on a signal line including the steps of:

a. providing a digitized signal representing electrical energy on a signal line;

b. correlating N neighboring portions of the digitized signal and reporting no speech present if the correlation does not exceed a predetermined threshold;

c. performing Fast Fourier Transform analysis on the digitized signal and identifying a first, a second, and a third largest frequency-domain maxima;

d. determining whether the first and second largest maxima are above a threshold frequency and reporting speech present if either the first or the second largest maxima are below the threshold frequency; and

e. determining whether a ratio of the first largest maxima to the third largest maxima exceeds a predetermined value and reporting a tone present if the ratio is larger than the predetermined value, and reporting noise present if the ratio is smaller than the predetermined value.

2. The method as claimed in claim 1 wherein the threshold frequency is 300 Hz.

3. An apparatus for performing real-time recognition of speech and call-progression tones from electrical energy on a signal line comprising:

a. means for receiving a first plurality of digitized signal samples representing electrical energy on a signal line;

b. means for attenuating the digitized signal samples and forming a filtered signal, the means for attenuating coupled to the means for receiving;

c. means for decimating the filtered signal and forming a decimated signal, the means for decimating coupled to the means for attenuating;

d. means for correlating a second plurality of neighboring portions of the decimated signal and calculating a maximum, the means for correlating and calculating coupled to the means for decimating;

e. first means for determining if the maximum exceeds a first predetermined threshold and reporting noise present if the predetermined threshold is not exceeded, the means for determining coupled to the means for correlating and calculating;

f. means for performing a Fast Fourier Transform analysis of the decimated signal and identifying a first, a second and a third largest frequency-domain maxima, the means for performing coupled to the first means for determining;

g. second means for determining if both the first and second largest frequency-domain maxima exceed a second predetermined threshold and reporting speech present if either the first or the second largest frequency-domain maxima exceed the second predetermined threshold, the second means for determining coupled to the means for performing a Fast Fourier Transform; and

h. third means for determining whether a ratio of the first largest to the third largest frequency-domain maxima exceeds a third predetermined value and reporting a tone present if the ratio is larger than the third predetermined value, the third means for determining coupled to the second means for determining.

4. The apparatus as claimed in claim 3 wherein the means for attenuating comprises a low-pass digital filter where attenuation begins at 700 Hz and is complete at 800 Hz.

5. The apparatus as claimed in claim 4 wherein the means for decimating takes every fifth sample.

6. The apparatus as claimed in claim 5 further comprising a register for storing data samples while they are being processed, the register coupled to the means for decimating and the means for correlating.

7. The apparatus as claimed in claim 6 wherein the first predetermined threshold is a value taken from the range of values beginning at 0.6 and ending at 0.8.

8. The apparatus as claimed in claim 7 wherein the plurality of neighboring portions comprises 3 neighboring portions.

9. The apparatus as claimed in claim 8 wherein the second predetermined threshold is equal to 300 Hz.

10. An apparatus for performing real-time recognition of speech and call-progression tones from electrical energy on a signal line comprising:

a. means for receiving a first plurality of digitized signal samples representing electrical energy on a signal line;

b. means for attenuating the digitized signal and forming a filtered signal, the means for attenuating coupled to the means for receiving; and

c. microprocessor means for determining whether the electrical energy represents a tone, noise or speech coupled to the means for receiving and the means for attenuating, the microprocessor means comprising;

i. means for decimating the filtered signal and forming a decimated signal;

ii. means for correlating a second plurality of neighboring portions of the decimated signal and calculating a maximum;

iii. first means for determining if the maximum exceeds a first predetermined threshold and reporting noise present if the predetermined threshold is not exceeded;

iv. means for performing a Fast Fourier Transform analysis of the decimated signal and identifying a first, a second and a third largest frequency-domain maxima;

v. second means for determining if both the first and second largest frequency-domain maxima exceed a second predetermined threshold and reporting speech present if either the first or the second largest frequency-domain maxima exceed the second predetermined threshold; and

vi. third means for determining whether a ratio of the first largest to the third largest frequency-domain maxima exceeds a third predetermined value and reporting a tone present if the ratio is larger than the third predetermined value.

11. The apparatus as claimed in claim 10 wherein the means for attenuating comprises a low-pass digital filter.

12. The apparatus as claimed in claim 11 further comprising a storage register for storing data samples while the data samples are being processed, the storage register coupled to the microprocessor means.

13. The apparatus as claimed in claim 12 wherein the second plurality of neighboring portions comprises three neighboring portions.