IP Library › Granted Patent US 12,270,899
Granted Patent B2
US 12,270,899 · App. 17/800,145 · Granted Apr 8, 2025

Trajectory estimation device, trajectory estimation system, trajectory estimation method, and program recording medium

Inventors: Yumi Arai (Tokyo, JP); Reishi Kondo (Tokyo, JP)
Assignee: NEC CORPORATION
G01S15/586G01S7/536
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,270,899
App. No.
17/800,145
Granted
Apr 8, 2025
Kind
B2
Abstract

A trajectory estimation device that includes an acquisition unit that acquires wave data based on a wave detected by a plurality of sensors; a generation unit that generates a spectrogram using the wave data; an extraction unit that extracts a Doppler shift from the spectrogram; a selection unit that selects, as a sensor pair, two of the sensors that satisfy a preset selection condition for the Doppler shift; and an estimation unit that estimates a trajectory of a wave source, which is a source generating the wave, based on a positional relationship between the sensors constituting the sensor pair and a relationship of the Doppler shifts between the two sensors constituting the sensor pair.

Claims (54)

1. A trajectory estimation device comprising:

at least one memory storing instructions; and

at least one processor connected to the at least one memory and configured to execute the instructions to:

acquire wave data based on a wave detected by each of a plurality of sensors disposed at intervals according to a type of a wave source to be estimated;

generate a spectrogram using the wave data, the spectrogram being a three-dimensional graph representing the wave detected by each of the plurality of sensor in a relationship of a time, a frequency, and an intensity;

extract Doppler shifts from the spectrogram, the Doppler shifts being temporal changes of frequencies extracted based on a peak having a maximum or greatest intensity in the spectrogram;

select, as a sensor pair, two of the sensors that satisfy a preset selection condition for the Doppler shifts; and

estimate a trajectory of the wave source, which is a source generating the wave, based on a positional relationship between the sensors constituting the sensor pair and a relationship of the Doppler shifts between the two sensors constituting the sensor pair, wherein

in a case where a frequency of the wave is known in advance, the at least one processor is configured to execute the instructions to:

generate a plurality of bases by shifting the frequency of the wave within a certain frequency width;

derive activation for the plurality of bases by non-negative matrix factorization;

specify a basis for which the activation has a maximum value, among the plurality of bases; and

calculate, as the Doppler shifts, curves obtained by plotting the frequency of the basis at which the activation has the maximum value.

2. The trajectory estimation device according to claim 1 , wherein

the preset selection condition includes a first condition in which frequencies of inflection points of the Doppler shifts are same, and a second condition in which the Doppler shifts intersect.

3. The trajectory estimation device according to claim 1 , wherein

the at least one processor is configured to execute the instructions to

extract curves along continuous peaks in the spectrogram as the Doppler shifts.

4. The trajectory estimation device according to claim 1 , wherein

the at least one processor is configured to execute the instructions to

extract curves obtained by interpolating discrete peaks in the spectrogram as the Doppler shifts.

5. The trajectory estimation device according to claim 1 , wherein

the at least one processor is configured to execute the instructions to

estimate the trajectory of the wave source by calculating a velocity of the wave source, a shortest distance between each of the two sensors constituting the sensor pair and the wave source, and an angle formed by a straight line passing through positions of the two sensors constituting the sensor pair and the trajectory of the wave source;

based on a relationship between a distance between the two sensors constituting the sensor pair, a time at which the wave source passes on the straight line passing through the positions of the two sensors constituting the sensor pair, a first time at which a distance between a first sensor of the two sensors constituting the sensor pair and the wave source is shortest, a frequency at the first time, a second time at which a distance between a second sensor of the two sensors constituting the sensor pair and the wave source is shortest, and a frequency at the second time.

6. The trajectory estimation device according to claim 1 , wherein

the at least one processor is configured to execute the instructions to

select at least two sets of the sensor pairs, and

estimate the trajectory of the wave source based on a matching trajectory candidate among trajectory candidates of the wave source estimated for the at least two sets of the sensor pairs that have been selected.

7. A trajectory estimation system comprising:

the trajectory estimation device according to claim 1 ; and

the plurality of sensors.

8. A trajectory estimation method performed by a computer and comprising:

acquiring wave data based on a wave detected by each of a plurality of sensors disposed at intervals according to a type of a wave source to be estimated;

generating a spectrogram using the wave data, the spectrogram being a three-dimensional graph representing the wave detected by each of the plurality of sensor in a relationship of a time, a frequency, and an intensity;

extracting Doppler shifts from the spectrogram, the Doppler shifts being temporal changes of frequencies extracted based on a peak having a maximum or greatest intensity in the spectrogram;

selecting, as a sensor pair, two of the sensors that satisfy a preset selection condition for the Doppler shifts; and

estimating a trajectory of the wave source, which is a source generating the wave, based on a positional relationship between the sensors constituting the sensor pair and a relationship of the Doppler shifts between the two sensors constituting the sensor pair, wherein

in a case where a frequency of the wave is known in advance.

generating a plurality of bases by shifting the frequency of the wave within a certain frequency width;

deriving activation for the plurality of bases by non-negative matrix factorization;

specifying a basis for which the activation has a maximum value, among the plurality of bases; and

calculating, as the Doppler shifts, curves obtained by plotting the frequency of the basis at which the activation has the maximum value.

9. A non-transitory program recording medium storing a program executable by a computer to perform processing comprising:

acquiring wave data based on a wave detected by each of a plurality of sensors disposed at intervals according to a type of a wave source to be estimated;

generating a spectrogram using the wave data, the spectrogram being a three-dimensional graph representing the wave detected by each of the plurality of sensor in a relationship of a time, a frequency, and an intensity;

extracting Doppler shifts from the spectrogram, the Doppler shifts being temporal changes of frequencies extracted based on a peak having a maximum or greatest intensity in the spectrogram;

selecting, as a sensor pair, two of the sensors that satisfy a preset selection condition for the Doppler shifts; and

estimating a trajectory of the wave source, which is a source generating the wave, based on a positional relationship between the sensors constituting the sensor pair and a relationship of the Doppler shifts between the two sensors constituting the sensor pair, wherein

in a case where a frequency of the wave is known in advance,

generating a plurality of bases by shifting the frequency of the wave within a certain frequency width;

deriving activation for the plurality of bases by non-negative matrix factorization;

specifying a basis for which the activation has a maximum value, among the plurality of bases; and

calculating, as the Doppler shifts, curves obtained by plotting the frequency of the basis at which the activation has the maximum value.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Aug 16, 2022
From: ARAI, YUMI; KONDO, REISHI
To: NEC CORPORATION
Reel/Frame 060823/0912 →
Continuity (1)
Related Publication 20230071304A1 · Mar 9, 2023
References Cited (17)
US 5241518A · McNelis et al. · 1993 [cited by applicant]
US 11711648B2 · Lopatka · 2023 [cited by examiner]
US 20070033045A1 · Smaragdis · 2007 [cited by examiner]
US 20110110195A1 · Hall · 2011 [cited by examiner]
US 20130010968A1 · Yagi · 2013 [cited by examiner]
US 20170016983A1 · Hoshuyama · 2017 [cited by examiner]
JP H05087903A · 1993 [cited by applicant]
JP H07505219A · 1995 [cited by applicant]
JP 2000205794A · 2000 [cited by examiner]
JP 2007033445A · 2007 [cited by applicant]
JP 2015118004A · 2015 [cited by applicant]
JP-2000205794-A (machine translation) (Year: 2000). [cited by examiner]
Gupta et al. (“Motion parameter estimation of a radiating point source with multiple tonals using acoustic Doppler analysis,” 2013 IEEE International Underwater Technology Symposium (UT), Tokyo, Japan, 2013, pp. 1-6) (Y… [cited by examiner]
Parsayan (“High speed sound source detection based on ADFSD measurement,” 2013 9th International Conference on Information, Communications & Signal Processing, Tainan, Taiwan, 2013) (Year: 2013). [cited by examiner]
Barnwal et al. (“Doppler based speed estimation of vehicles using passive sensor,” 2013 IEEE International Conference on Multimedia and Expo Workshops (ICMEW), San Jose, CA, USA, 2013, pp. 1-4) (Year: 2013). [cited by examiner]
International Search Report for PCT Application No. PCT/JP2020/010269, mailed on Jun. 30, 2020. [cited by applicant]
English translation of Written opinion for PCT Application No. PCT/JP2020/010269, mailed on Jun. 30, 2020. [cited by applicant]