Detection and signal isolation of individual vehicle signatures

• US 10,353,060 B2
• Filed: 10/23/2017
• Issued: 07/16/2019
• Est. Priority Date: 12/07/2016
• Status: Active Grant

• Current Assignee: Raytheon BBN Technologies Corp. (Rtx Corporation)
• Original Assignee: Raytheon BBN Technologies Corp. (Rtx Corporation)
• Inventors: Kofford, Aaron Scott, Allen, Matthew Paul, Goldsmith, Michael Bruce
• Primary Examiner(s): Phan, Hai
• Assistant Examiner(s): Tang, Son M
• Application Number:
  US15/791,378
• Publication Number:
  US 20180156905A1
• Time in Patent Office: 631 Days
• CPC Class Codes: G01S 11/14 using ultrasonic, sonic, or infrasonic waves
• CPC Class Codes: G08G 1/0116 from roadside infrastructure, e.g. beacons
• CPC Class Codes: G08G 1/0129 for creating historical data or processing based on historical data
• 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: G08G 1/052 with provision for determining speed or overspeed speed measuring in general G01P
• CPC Class Codes: H04R 1/406 microphones
• CPC Class Codes: H04R 3/005 for combining the signals of two or more microphones specially adapted for hearing aids H04R25/407

Abstract

A system for isolating signals from individual vehicles. An array of sensors generates a plurality of time records. The time records are transformed into spectrogram structures, and for each time-frequency point in the spectrogram structures, a bearing is calculated, to form a movement structure. A Hough transform is used to fit a movement model to the movement structure, to form a Hough transform structure. Peaks in the Hough transform structure correspond to vehicles and their position in the Hough transform structure correspond to motion parameters, such as the velocity of the vehicle, and the time of closest approach of the vehicle to the array of acoustic sensors. A model of the motion of a vehicle is then used to associate, with the vehicle, points (e.g., structure elements in the spectrogram structures) from the data.

Claims

1. A system for detecting objects moving along a path, the system comprising:

a plurality of sensors; and

a processing circuit connected to the sensors,

the processing circuit being configured to;

receive a plurality of digital time records of acoustic data, each representing a signal sensed by a respective one of the plurality of sensors,

form a plurality of complex spectrogram structures, each of the complex spectrogram structures being formed from a respective one of the digital time records, each element of the complex spectrogram structure having a frequency coordinate and a time coordinate,

form a movement structure from the complex spectrogram structures, each element of the movement structure;

having a frequency coordinate,

corresponding to a respective element of the complex spectrogram structure, and

having a value corresponding to an angle of arrival of an acoustic wave at the frequency corresponding to the frequency coordinate of the element of the movement structure,

fit a movement model to the movement structure, the movement model representing an object moving with constant velocity along the path; and

infer a velocity of a first object from the fit.

2. The system of claim 1, wherein the processing circuit is further configured to infer a time of closest approach of the first object from the fit.

3. The system of claim 1, wherein the path is a substantially straight path, and the movement model is a function returning a bearing as a function of two parameters and a time variable t, the parameters being a velocity v of the object and a time of closest approach t_CPA, the function being:

tan<sup>−

1</sup>(v(t−

t_CPA)/D), wherein D is a perpendicular distance from the path to the plurality of sensors.

4. The system of claim 1, wherein the forming of the movement structure comprises calculating a bearing, for each element of each of the complex spectrogram structures, using a frequency wavenumber analysis.

5. The system of claim 1, wherein the fitting of a movement model to the movement structure comprises:

forming a Hough transform structure from the movement structure, for the movement model; and

finding coordinates of a first peak in the Hough transform structure, and

wherein the inferring of the velocity of the first object from the fit comprises inferring the velocity of the first object from the coordinates of the first peak.

6. The system of claim 5, wherein each element of the movement structure has a time coordinate, a frequency coordinate, and a data bearing value, and wherein the forming of the Hough transform comprises counting, for a structure element of the Hough transform structure corresponding to a velocity and a time of closest approach, the number of elements of the movement structure for which:

the movement model, when evaluated at the time coordinate, the velocity, and the time of closest approach, returns a model bearing value within a first fitting tolerance of the data bearing value.

7. The system of claim 5, wherein the finding of the first peak comprises finding a structure element having the greatest value within the Hough transform structure.

8. The system of claim 7, wherein the processing circuit is further configured to:

find coordinates of a second peak in the Hough transform structure, and

infer, from the coordinates of the second peak;

a velocity of a second object, and

a time of closest approach of the second object,

the finding of the second peak comprising finding a structure element having the greatest value within a portion of the Hough transform structure excluding a neighborhood of the first peak.

9. The system of claim 8, wherein the processing circuit is further configured to:

find coordinates of a third peak in the Hough transform structure, and

infer a velocity of a third object from the coordinates of the third peak,

the finding of the third peak comprising finding a maximum value within a portion of the Hough transform structure that is on a straight line joining the first peak and the second peak, and that excludes a neighborhood of the first peak and a neighborhood of the second peak.

10. The system of claim 5, wherein the processing circuit is further configured:

from the coordinates of a peak in the Hough transform structure, to generate a movement model corresponding to the coordinates of the peak; and

from the movement model, to identify a plurality of structure elements of the movement structure associated with the movement model.

11. The system of claim 10, wherein the identifying of the plurality of structure elements of the movement structure comprises identifying elements differing in bearing from the movement model by less than a second fitting threshold.

12. A method for detecting objects moving along a path, the method comprising:

receiving a plurality of digital time records of data, each representing a signal sensed by a respective one of a plurality of sensors,

forming a plurality of complex spectrogram structures, each of the complex spectrogram structures being formed from a respective one of the digital time records, each element of the complex spectrogram structure having a frequency coordinate and a time coordinate,

forming a movement structure from the complex spectrogram structure, each element of the movement structure;

having a frequency coordinate,

corresponding to a respective element of the complex spectrogram structure, and

having a value corresponding to an angle of arrival of a signal at the frequency corresponding to the frequency coordinate of the element of the movement structure,

fitting a movement model to the movement structure, the movement model representing an object moving with constant velocity along the path; and

inferring a velocity of a first object from the fit.

13. The method of claim 12, further comprising inferring a time of closest approach of the first object from the fit.

14. The method of claim 12, wherein the path is a substantially straight path, and the movement model is a function returning a bearing as a function of two parameters and a time variable t, the parameters being a velocity v of the object and a time of closest approach t_CPA, the function being:

tan<sup>−

1</sup>(v(t−

t_CPA)/D), wherein D is a perpendicular distance from the path to the plurality of sensors.

15. The method of claim 12, wherein the forming of the movement structure comprises calculating a bearing, for each element of each of the complex spectrogram structures, using a frequency wavenumber analysis.

16. The method of claim 12, wherein the fitting of a movement model to the movement structure comprises:

forming a Hough transform structure from the movement structure, for the movement model; and

finding coordinates of a first peak in the Hough transform structure, and

wherein the inferring of the velocity of the first object from the fit comprises inferring the velocity of the first object from the coordinates of the first peak.

17. The method of claim 16, wherein each element of the movement structure has a time coordinate, a frequency coordinate, and a data bearing value, and wherein the forming of the Hough transform comprises counting, for a structure element of the Hough transform structure corresponding to a velocity and a time of closest approach, the number of elements of the movement structure for which:

the movement model, when evaluated at the time coordinate, the velocity, and the time of closest approach, returns a model bearing value within a first fitting tolerance of the data bearing value.

18. The method of claim 16, wherein the finding of the first peak comprises finding a structure element having the greatest value within the Hough transform structure.

19. The method of claim 18, further comprising:

finding coordinates of a second peak in the Hough transform structure, and

inferring, from the coordinates of the second peak;

a velocity of a second object, and

a time of closest approach of the second object,

the finding of the second peak comprising finding a structure element having the greatest value within a portion of the Hough transform structure excluding a neighborhood of the first peak.

20. The method of claim 19, further comprising:

finding coordinates of a third peak in the Hough transform structure, and

inferring a velocity of a third object from the coordinates of the third peak,

the finding of the third peak comprising finding a maximum value within a portion of the Hough transform structure that is on a straight line joining the first peak and the second peak, and that excludes a neighborhood of the first peak and a neighborhood of the second peak.