Method for determining alcohol consumption, and recording medium and terminal for carrying out same

• US 9,916,844 B2
• Filed: 01/28/2014
• Issued: 03/13/2018
• Est. Priority Date: 01/28/2014
• Status: Active Grant

• Current Assignee: Foundation of Soongsil University-Industry Cooperation
• Original Assignee: Foundation of Soongsil University-Industry Cooperation
• Inventors: Bae, Myung Jin, Lee, Sang Gil, Baek, Geum Ran
• Primary Examiner(s): Desir, Pierre-Louis
• Assistant Examiner(s): Kim, Jonathan
• Application Number:
  US15/115,118
• Publication Number:
  US 20160379669A1
• Time in Patent Office: 1,505 Days
• Field of Search: None
• CPC Class Codes: A61B 5/4803 Speech analysis specially adapted for diagnostic purposes
• CPC Class Codes: A61B 5/4845 Toxicology, e.g. by detection of alcohol, drug or toxic products A61B5/4848 takes precedence
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit
• CPC Class Codes: G10L 15/16 using artificial neural networks
• CPC Class Codes: G10L 17/26 Recognition of special voice characteristics, e.g. for use in lie detectors; Recognition of animal voices
• CPC Class Codes: G10L 19/12 the excitation function being a code excitation, e.g. in code excited linear prediction [CELP] vocoders
• CPC Class Codes: G10L 25/15 the extracted parameters being formant information
• CPC Class Codes: G10L 25/66 for extracting parameters related to health condition detecting or measuring for diagnostic purposes A61B5/00
• CPC Class Codes: G10L 25/84 for discriminating voice from noise

Abstract

Disclosed are a method for determining whether a person is drunk after consuming alcohol on the basis of a difference among a plurality of formant energy energies, which are generated by applying linear predictive coding according to a plurality of linear prediction orders, and a recording medium and a terminal for carrying out the method. The alcohol consumption determining terminal comprises: a voice input unit for receiving voice signals and converting same into voice frames and outputting the voice frames; a voiced/unvoiced sound analysis unit for extracting voice frames corresponding to a voiced sound from among the voice frames; an LPC processing unit for calculating a plurality of formant energy energies by applying linear predictive cording according to the plurality of linear prediction orders to the voice frames corresponding to the voiced sound; and an alcohol consumption determining unit for determining whether a person is drunk after consuming alcohol on the basis of a difference among the plurality of formant energy energies which have been calculated by the LPC processing unit, thereby determining whether a person is drunk after consuming alcohol depending on a change in the formant energy energies generated by applying linear predictive coding according to the plurality of linear prediction orders to voice signals.

Claims

1. A method for determining whether alcohol is consumed by a person in a vehicle, the method comprising:

receiving a voice signal from the said person in the vehicle and converting the received voice signal into a plurality of voice frames;

extracting a voice frame corresponding to a voiced sound from among the plurality of voice frames;

calculating a plurality of formant energies by applying linear predictive coding according to a plurality of linear prediction orders to the extracted voice frame;

computing differences among the plurality of formant energies;

determining whether alcohol is consumed by said person in the vehicle according to the computed differences; and

enabling or disabling the vehicle based on the determination.

2. The method of claim 1, wherein the extracting a voice frame corresponding to a voiced sound among the plurality of voice frames comprises:

extracting predetermined features from a voice frame among the plurality of voice frames, and

determining whether said voice frame is from a voiced sound, an unvoiced sound, or background noise.

3. The method of claim 2, wherein the predetermined features comprise periodic characteristics of harmonics, root mean square energy (RMSE), or zero-crossing count (ZC) of a low-band voice signal energy area.

4. The method of claim 2, wherein the determining whether said voice frame is from a voiced sound, an unvoiced sound, or background noise comprises using neural network.

5. The method of claim 1, wherein the calculating a plurality of formant energies by applying linear predictive coding according to a plurality of linear prediction orders to the extracted voice frame comprises applying a low order linear prediction order and a high order linear prediction order to the extracted voice frame thereby calculating formant energies according to the applied linear prediction orders.

6. The method of claim 5, wherein the computing differences among the plurality of formant energies comprises calculating a difference between the formant energy according to the applied low order linear prediction order and the formant energy according to the applied high order linear prediction order.

7. The method of claim 1, wherein the determining whether alcohol is consumed by said person according to the computed differences comprises:

calculating an average energy of the differences among the plurality of formant energies for the voice frame,

calculating a ratio of average energies between a previous voice frame and a current voice frame for each of the voice frames

counting the number of voice frames each having the ratio of average energies greater than a predetermined threshold, and

determining whether alcohol is consumed by said person according to the counted number.

8. The method of claim 7, wherein the determining whether alcohol is consumed by said person according to the counted number comprises:

calculating a ratio between the counted number and the total number of voice frames determined as the voiced sounds,

determining that alcohol is consumed by said person when the calculated ratio is less than a predetermined value, and

determining that alcohol is not consumed by said person when the calculated ratio is greater than the predetermined value.

9. The method of claim 1, further comprising filtering out harmonics of a certain frequency or higher by applying a low pass filter to the extracted voice frame.

10. A non-transitory computer-readable recording medium having a program recorded thereon for performing a method for determining whether alcohol is consumed by a person in a vehicle, the method comprising:

receiving a voice signal from the said person in the vehicle and converting the received voice signal into a plurality of voice frames;

extracting a voice frame corresponding to a voiced sound from among the plurality of voice frames;

calculating a plurality of formant energies by applying linear predictive coding according to a plurality of linear prediction orders to the extracted voice frame;

computing differences among the plurality of formant energies;

determining whether alcohol is consumed by said person in the vehicle according to the computed differences; and

enabling or disabling the vehicle based on the determination.

11. The non-transitory computer-readable recording medium of claim 10, wherein the extracting a voice frame corresponding to a voiced sound among the plurality of voice frames comprises:

extracting predetermined features from a voice frame among the plurality of voice frames, and

determining whether said voice frame is from a voiced sound, an unvoiced sound, or background noise.

12. The non-transitory computer-readable recording medium of claim 11, wherein the predetermined features comprise periodic characteristics of harmonics, root mean square energy (RMSE), or zero-crossing count (ZC) of a low-band voice signal energy area.

13. The non-transitory computer-readable recording medium of claim 11, wherein the determining whether said voice frame is from a voiced sound, an unvoiced sound, or background noise comprises using neural network.

14. The non-transitory computer-readable recording medium of claim 10, wherein the calculating a plurality of formant energies by applying linear predictive coding according to a plurality of linear prediction orders to the extracted voice frame comprises applying a low order linear prediction order and a high order linear prediction order to the extracted voice frame thereby calculating formant energies according to the applied linear prediction orders.

15. The non-transitory computer-readable recording medium of claim 14, wherein the computing differences among the plurality of formant energies comprises calculating a difference between the formant energy according to the applied low order linear prediction order and the formant energy according to the applied high order linear prediction order.

16. The non-transitory computer-readable recording medium of claim 10, wherein the determining whether alcohol is consumed by said person according to the computed differences comprises:

calculating an average energy of the differences among the plurality of formant energies for the voice frame,

calculating a ratio of average energies between a previous voice frame and a current voice frame for each of the voice frames

counting the number of voice frames each having the ratio of average energies greater than a predetermined threshold, and

determining whether alcohol is consumed by said person according to the counted number.

17. The non-transitory computer-readable recording medium of claim 16, wherein the determining whether alcohol is consumed by said person according to the counted number comprises:

calculating a ratio between the counted number and the total number of voice frames determined as the voiced sounds, determining that alcohol is consumed by said person when the calculated ratio is less than a predetermined value, and determining that alcohol is not consumed by said person when the calculated ratio is greater than the predetermined value.

18. The non-transitory computer-readable recording medium of claim 10, further comprising filtering out harmonics of a certain frequency or higher by applying a low pass filter to the extracted voice frame.