Method and apparatus for detecting concealed weapons

• US 20050182590A1
• Filed: 02/13/2004
• Published: 08/18/2005
• Est. Priority Date: 02/13/2004
• Status: Active Grant

• Current Assignee: Battelle Energy Alliance LLC (Battelle Memorial Institute)
• Original Assignee: Battelle Energy Alliance LLC (Battelle Memorial Institute)
• Inventors: Kotter, Dale K., Fluck, Frederick D.
• Granted Patent: US 7,013,245 B2
• Time in Patent Office: Days
• US Class Current: 702/127
• CPC Class Codes: G06V 20/52 Surveillance or monitoring of activities, e.g. for recognising suspicious objects recognising microscopic objects G06V20/69
• CPC Class Codes: G06V 2201/05 Recognition of patterns representing particular kinds of hidden objects, e.g. weapons, explosives, drugs

Abstract

Apparatus for classifying a ferromagnetic object within a sensing area may comprise a magnetic field sensor that produces magnetic field data. A signal processing system operatively associated with the magnetic field sensor comprises a neural network. The neural network compares the magnetic field data with magnetic field data produced by known ferromagnetic objects to make a probabilistic determination as to the classification of the ferromagnetic object within the sensing area. A user interface operatively associated with the signal processing system produces a user-discernable output indicative of the probabilistic determination of the classification of the ferromagnetic object within a sensing area.

Claims

1. Apparatus for classifying a ferromagnetic object within a sensing area, comprising:

a magnetic field sensor, said magnetic field sensor producing magnetic field data;

a signal processing system operatively associated with said magnetic field sensor, said signal processing system comprising;

a neural network responsive to said magnetic field data produced by said magnetic field sensor, said neural network comparing said magnetic field data with magnetic field data produced by known ferromagnetic objects to make a probabilistic determination as to the classification of the ferromagnetic object within the sensing area; and

a user interface operatively associated with said signal processing system, said user interface producing a user-discernable output indicative of the probabilistic determination of the classification of the ferromagnetic object within a sensing area.

2. The apparatus of claim 1, wherein said signal processing system further comprises a three-dimensional target position analyzer, said three-dimensional target position analyzer, being responsive to said magnetic field data produced by said magnetic field sensor, said three-dimensional target position analyzer producing an output signal indicative of a location of the ferromagnetic object with respect to a host, said user interface being responsive to the output signal from said three-dimensional target position analyzer to produce a user-discernable output indicative of the location of the ferromagnetic object with respect to the host.

3. The apparatus of claim 2, wherein said neural network is operatively connected to said three-dimensional target position analyzer so that said neural network receives the output signal from said three-dimensional target position analyzer, and wherein said neural network uses the output signal from said three-dimensional target position analyzer in making the probabilistic determination of the classification of the ferromagnetic object within the sensing area.

4. The apparatus of claim 1, wherein said signal processing system further comprises a joint time-frequency processor, said joint time-frequency processor processing the magnetic field data produced by said magnetic sensor in the joint time-frequency domain to produce JTFA data, said user interface being responsive to said JTFA data to produce a user-discernable output indicative of said magnetic field data in the joint time-frequency domain.

5. The apparatus of claim 4, wherein said neural network is operatively connected to said joint time-frequency processor so that said neural network receives the JTFA data from said joint time-frequency processor, and wherein said neural network uses the JTFA data from said joint time-frequency processor in making the probabilistic determination of the classification of the ferromagnetic object within the sensing area.

6. The apparatus of claim 1, wherein said signal processing system further comprises a fast Fourier transform processor, said fast Fourier transform processor processing the magnetic field data produced by said magnetic sensor in the frequency domain to produce FFT data, said user interface being responsive to said FFT data to produce a user-discernable output indicative of said magnetic field data in the frequency domain.

7. The apparatus of claim 6, wherein said neural network is operatively connected to said fast Fourier transform processor so that said neural network receives the FFT data from said fast Fourier transform processor, and wherein said neural network uses the FFT data from said fast Fourier transform processor in making the probabilistic determination of the classification of the ferromagnetic object within the sensing area.

8. The apparatus of claim 1, wherein said signal processing system further comprises a filter network operatively connected to said magnetic field sensor, said filter network filtering the magnetic field data produced by said magnetic field sensor before said magnetic field data are further processed by said signal processing system.

9. The apparatus of claim 8, wherein said filter network comprises an inverse Chebyshev filter.

10. The apparatus of claim 1, wherein said signal processing system further comprises a threshold analysis processor operatively connected to said magnetic field sensor, said threshold analysis processor preventing from being further processed magnetic field data below a predetermined threshold.

11. The apparatus of claim 1, wherein said magnetic field sensor comprises a plurality of magnetometers.

12. A method for classifying a ferromagnetic object within a sensing area, comprising:

sensing, during a sample time, a magnetic field in the sensing area;

producing magnetic field data representative of the sensed magnetic field during the sample time;

using a neural network to compare said magnetic field data with magnetic field data produced by known ferromagnetic objects to make a probabilistic determination as to the classification of the ferromagnetic object within a sensing area; and

producing a user-discernable output indicative of the probabilistic determination of the classification of the ferromagnetic object within the sensing area.

13. The method of claim 12, wherein using a neural network comprises:

identifying at least one characteristic of the magnetic field data;

assigning a weight value to the at least one characteristic of the magnetic field data to produce weighted characteristic magnetic field data; and

comparing the weighted characteristic magnetic field data with a database containing magnetic characteristics of known ferromagnetic objects.

14. The method of claim 13, wherein identifying at least one characteristic comprises determining whether the magnetic field data are indicative of a monopole response.

15. The method of claim 13, wherein identifying at least one characteristic comprises determining whether the magnetic field data are indicative of a dipole response.

16. The method of claim 13, wherein identifying at least one characteristic comprises determining whether the magnetic field data are indicative of a quadrupole response.

17. The method of claim 12, further comprising:

performing joint time frequency analysis on the magnetic field data to produce JTFA data; and

using the neural network to process the JTFA data and make the probabilistic determination of the classification of the ferromagnetic object within the sensing area.

18. The method of claim 12, further comprising:

determining an elapsed time between an initial trigger and a magnetic response to produce time-position data; and

using the neutral network to process the time-position data and make the probabilistic determination of the classification of the ferromagnetic object within the sensing area.

19. The method of claim 12, further comprising:

performing a Fourier transform on the magnetic field data to produce FFT data; and

using the neutral network to process the FFT data and make the probabilistic determination of the classification of the ferromagnetic object within the sensing area.

20. The method of claim 12, further comprising filtering the magnetic field data before using the neural network.

21. The method of claim 20, wherein filtering comprises filtering the magnetic field data with an inverse Chebyshev filter.

22. Apparatus for classifying a ferromagnetic object within a sensing area, comprising:

magnetic field sensing means for sensing a magnetic field and for producing magnetic field data;

signal processing means operatively associated with said magnetic field sensing means, said signal processing means comprising;

neural network means for comparing said magnetic field data with magnetic field data of known ferromagnetic objects to make a probabilistic determination as to the classification of the ferromagnetic object within the sensing area; and

user interface means operatively associated with said signal processing means for producing a user-discernable output indicative of the probabilistic determination of the classification of the ferromagnetic object within a sensing area.