Detection and classification of running vehicles based on acoustic signatures

• US 8,164,484 B2
• Filed: 10/03/2008
• Issued: 04/24/2012
• Est. Priority Date: 10/03/2007
• Status: Active Grant

• Current Assignee: University of Southern California
• Original Assignee: University of Southern California
• Inventors: Berger, Theodore W., Lu, Bing, Dibazar, Alircza
• Primary Examiner(s): Nguyen, Hung T.
• Application Number:
  US12/245,564
• Publication Number:
  US 20090115635A1
• Time in Patent Office: 1,299 Days
• Field of Search: 340/943, 340/941, 340/933, 340/539.1, 340/825.71, 340/825.72, 367/127, 367/128, 381/56, 381/86, 701/207, 701/213, 701/209, 701/212, 701/117, 701/119
• US Class Current: 340/943
• CPC Class Codes: G01H 3/08 Analysing frequencies present in complex vibrations, e.g. comparing harmonics present acoustic presence detection G01V1/001
• CPC Class Codes: G08G 1/015 with provision for distinguishing between two or more types of vehicles, e.g. between motor-cars and cycles
• CPC Class Codes: G08G 1/04 using optical or ultrasonic detectors
• CPC Class Codes: G10L 25/30 using neural networks
• CPC Class Codes: G10L 25/51 for comparison or discrimination

Abstract

A method and apparatus for identifying running vehicles in an area to be monitored using acoustic signature recognition. The apparatus includes an input sensor for capturing an acoustic waveform produced by a vehicle source, and a processing system. The waveform is digitized and divided into frames. Each frame is filtered into a plurality of gammatone filtered signals. At least one spectral feature vector is computed for each frame. The vectors are integrated across a plurality of frames to create a spectro-temporal representation of the vehicle waveform. In a training mode, values from the spectro-temporal representation are used as inputs to a Nonlinear Hebbian learning function to extract acoustic signatures and synaptic weights. In an active mode, the synaptic weights and acoustic signatures are used as patterns in a supervised associative network to identify whether a vehicle is present in the area to be monitored. In response to a vehicle being present, the class of vehicle is identified. Results may be provided to a central computer.

Claims

1. An apparatus for identifying vehicles entering an area to be monitored using acoustic signatures, comprising:

one or more microphones configured to capture, within a range of said area, sound produced by a source and to convert said sound into electrical signals;

a memory unit configured to store one or more learned acoustic signatures of running vehicles;

a processing system coupled to said memory unit and configured to measure, from said electrical signals, an acoustic pattern of said source;

compare said acoustic pattern with one or more learned acoustic signatures of running vehicles stored in said memory unit;

determine, based on said comparison, whether said acoustic pattern is indicative of a presence of a running vehicle; and

identify, in response to detecting a running vehicle, a vehicle class.

2. The apparatus of claim 1 further comprising a wireless transmitter.

3. The apparatus of claim 2 wherein the processing system is further configured to transmit, via said wireless transmitter, information comprising said vehicle class to a central computer.

4. The apparatus of claim 1 further comprising an amplifier configured to amplify said electrical signals.

5. The apparatus of claim 1 wherein said one or more microphones comprise a distributed microphone array.

6. The apparatus of claim 1 wherein said measuring said acoustic pattern comprises using a radial basis function neural network.

7. The apparatus of claim 1 wherein said determining whether said acoustic pattern is indicative of a presence of a running vehicle comprises using a radial basis function neural network.

8. The apparatus of claim 1 wherein said measuring said acoustic pattern comprises using an expectation-maximization method.

9. The apparatus of claim 1 wherein said vehicle class comprises one of gasoline light wheeled, gasoline heavy wheeled, diesel truck, and motorcycle.

10. The apparatus of claim 1 wherein said area to be monitored comprises an asset to be protected.

11. The apparatus of claim 1 wherein said area to be monitored comprises a parking lot.

12. The apparatus of claim 1 wherein the acoustic signatures are recognized using Spectro-Temporal representation and a nonlinear Hebbian learning function.

13. The apparatus of claim 1 wherein expectation-maximizing is used for training.

14. A method for identifying approaching vehicles entering an area to be monitored using acoustic signatures, comprising:

capturing, using one more microphones placed within a range of said area, sound produced by a source and converting said sound into electrical signals;

measuring, from said electrical signals, an acoustic pattern of said source;

comparing said acoustic pattern with one or more learned acoustic signatures of running vehicles;

determining, based on said comparison, whether said acoustic pattern is indicative of a presence of a running vehicle; and

identifying, in response to detecting a running vehicle, a vehicle class, processed by a processing system.

15. The method of claim 14 further comprising transmitting, via a wireless transmitter, information sufficient to identify said vehicle class to a central computer.

16. The method of claim 14 further comprising digitizing, using an analog-to-digital converter, said amplified electrical signals.

17. The method of claim 16 wherein said area to be monitored comprises an asset to be protected.

18. The method of claim 14 wherein the one or more microphones comprise a distributed microphone array.

19. The method of claim 14 wherein said measuring said acoustic pattern comprises using a radial basis function neural network.

20. The method of claim 14 wherein said determining whether said acoustic pattern is indicative of a presence of a running vehicle comprises using a radial basis function neural network.

21. The method of claim 14 wherein said vehicle class comprises one of gasoline light wheeled, gasoline heavy wheeled, diesel truck, and motorcycle.

22. A system for identifying vehicles entering one or more areas to be monitored using acoustic signatures, the system comprising:

a central computer; and

at least one on-site sensor located within each area to be monitored, each at least one on-site sensor configured to

capture, using one more microphones placed within a range of said area, sound produced by a source and convert said sound into digitized electrical signals;

measure, from said electrical signals, an acoustic pattern of said source;

compare said acoustic pattern with one or more learned acoustic signatures of running vehicles;

determine, based on said comparison, whether said acoustic pattern is indicative of a presence of a running vehicle;

identify, in response to detecting a running vehicle, a vehicle class; and

transmit information sufficient to identify said vehicle class to said central computer.

23. The system of claim 22 wherein said at least one on-site sensor further comprises a wireless transmitter from which said information is transmitted.

24. The system of claim 22 wherein said central computer is integrated into a command center.