Method and system for detecting targets known up to a simplex from multi-spectral and hyper-spectral imagery employing the normal compositional model

• US 6,958,466 B1
• Filed: 06/04/2003
• Issued: 10/25/2005
• Est. Priority Date: 07/09/2002
• Status: Expired due to Fees

• Current Assignee: The Government Of The United States As Represented By The Secretary Of The Navy, the united states of america as represented by the secretary of the navy
• Original Assignee: the united states of america as represented by the secretary of the navy
• Inventors: Stein, David W.
• Primary Examiner(s): LE, QUE TAN
• Application Number:
  US10/454,216
• Time in Patent Office: 874 Days
• Field of Search: 250/221, 250/208.1, 250/201.5, 250/214.1, 250/226, 250/203.2, 250/203.6, 348/269, 348/263, 257/440, 257/291, 244/3.16, 244/3.17
• US Class Current: 250/221
• CPC Class Codes: G06V 20/13 Satellite images

Abstract

A method for detecting targets comprises: a) receiving spectral data; b) using a normal compositional model for estimating background parameters from the spectral data and target components; c) estimating abundance values of classes of the normal compositional model from the background parameters and the spectral data; d) estimating target class covariance values from the spectral data, the background parameters, and the target components; e) estimating target-plus-background abundance values from the target class covariance values, the background parameters, the spectral data, and the target components; f) employing a normal compositional model for determining a likelihood ratio detection statistic from the target class covariance values, target-plus-background abundance values, spectral data, target components, background parameters, and background abundance values; and g) generating a determination output signal that represents whether an observation includes a target from the likelihood ratio detection statistic.

Claims

1. A method for detecting targets, comprising:

a) receiving spectral data;

b) using a normal compositional model for estimating background parameters from said spectral data and target components;

c) estimating abundance values of classes of said normal compositional model from said background parameters and said spectral data;

d) estimating target class covariance values from said spectral data, said background parameters, and said target components;

e) estimating target-plus-background abundance values from said target class covariance values, said background parameters, said spectral data, and said target components;

f) employing a normal compositional model for determining a likelihood ratio detection statistic from said target class covariance values, said target-plus-background abundance values, said spectral data, said target components, said background parameters, and background abundance values; and

g) generating a determination output signal that represents whether an observation includes a target from said likelihood ratio detection statistic.

2. The method of claim 1 wherein estimating background parameters further includes:

h) initializing current class parameters and current abundance estimates;

i) defining updated abundance estimates from said current class parameters;

j) determining converged class parameter candidates from said updated abundance estimates and said current class parameters; and

k) generating said background parameters if said converged class parameter candidates satisfy first convergence criteria, or returning to said defining updated abundance estimates if said converged class parameter candidates do not satisfy said first convergence criteria.

3. The method of claim 2 wherein determining said converged class parameter candidates further includes:

l) creating updated background class parameters from said current class parameters and said updated abundance estimates; and

m) generating said converged class parameter candidates if said background class parameters satisfy second convergence criteria, or returning to said creating updated current class parameters if said background class parameters do not satisfy said second convergence criteria.

4. The method of claim 2 wherein initializing said current class parameters includes:

n) defining a shade point offset value;

o) defining a shade point covariance value from said shade point offset value;

p) generating reduced spectral data from said spectral data, said target components, and said shade point offset value;

q) defining end members from said reduced spectral data; and

r) generating said current class parameters from said end members.

5. The method of claim 1 wherein said spectral data is detected by an imaging spectrometer.

6. The method of claim 1 wherein said spectral data represents surface spectra.

7. A computer program product, comprising;

a computer readable medium having computer readable program code means embodied thereon for detecting anomalies in spectral data, said computer readable program code means including;

a) first computer readable program means for receiving spectral data;

b) second computer readable program means for using a normal compositional model for estimating background parameters from said spectral data and target components;

c) third computer readable program means for estimating abundance values of classes of said normal compositional model from said background parameters and said spectral data;

d) fourth computer readable program means for estimating target class covariance values from said spectral data, said background parameters, and said target components;

e) fifth computer readable program means for estimating target-plus-background abundance values from said target class covariance values, said background parameters, spectral data, and said target components;

f) sixth computer readable program means for employing a normal compositional model for determining a likelihood ratio detection statistic from said target class covariance values, said target-plus-background abundance values, said spectral data, target components, said background parameters, and background abundance values; and

g) seventh computer readable program means for generating a determination output signal that represents whether an observation includes a target from said likelihood ratio detection statistic.

8. The computer program product of claim 7 wherein said second computer readable program means further includes:

h) eighth computer readable program means for initializing current class parameters and current abundance estimates;

i) ninth computer readable program means for defining updated abundance estimates from said current class parameters;

j) tenth computer readable program means for determining converged class parameter candidates from said updated abundance estimates and said current class parameters; and

k) eleventh computer readable program means for generating said background parameters 2 if said converged class parameter candidates satisfy first convergence criteria or returning to said defining updated abundance estimates if said converged class parameter candidates do not satisfy said first convergence criteria.

9. The computer program product of claim 8 wherein said tenth computer readable program means further includes:

l) twelfth computer readable program means for creating updated background class parameters from said current class parameters and said updated abundance estimates; and

m) thirteenth computer readable program means for generating said converged class parameter candidates if said background class parameters satisfy second convergence criteria, or returning to said creating updated current class parameters if said background class parameters do not satisfy said second convergence criteria.

10. The computer program product of claim 9 wherein said eighth computer readable program means further includes:

n) fourteenth computer readable program means for defining a shade point offset value;

o) fifteenth computer readable program means for defining a shade point covariance value from said shade point offset value;

p) sixteenth computer readable program means for generating reduced spectral data from said spectral data, said target components, and said shade point offset value;

q) seventeenth computer readable program means for defining end members from said reduced spectral data; and

r) eighteenth computer readable program means for generating said current class parameters from said end members.

11. A system for detecting targets, comprising:

a computer for executing a sequence of computer readable instructions for performing the processes of;

a) receiving spectral data;

b) using a normal compositional model for estimating background parameters from said spectral data and target components;

c) estimating abundance values of classes of said normal compositional model from said background parameters and said spectral data;

d) estimating target class covariance values from said spectral data, said background parameters, and said target components;

e) estimating target-plus-background abundance values from said target class covariance values, said background parameters, spectral data, and said target components;

f) employing a normal compositional model for determining a likelihood ratio detection statistic from said target class covariance values, said target-plus-background abundance values, said spectral data, said target components, said background parameters, and background abundance values; and

g) generating a determination output signal that represents whether an observation includes a target from said likelihood ratio detection statistic.

12. The system of claim 11 wherein estimating background parameters further includes:

h) initializing current class parameters and current abundance estimates;

i) defining updated abundance estimates from said current class parameters;

j) determining converged class parameter candidates from said updated abundance estimates and said current class parameters; and

k) generating said background parameters if said converged class parameter candidates satisfy first convergence criteria or returning to said defining updated abundance estimates if said converged class parameter candidates do not satisfy said first convergence criteria.

13. The system of claim 12 wherein determining said converged class parameter candidates further includes:

l) creating updated class parameters from said current class parameters and said updated abundance estimates; and

m) generating said converged class parameter candidates if said background class parameters satisfy second convergence criteria, or returning to said creating updated current class parameters if said background class parameters do not satisfy said second convergence criteria.

14. The system of claim 12 wherein initializing said current class parameters include:

n) defining a shade point offset value;

o) defining a shade point covariance value from said shade point offset value;

p) generating reduced spectral data from said spectral data, said target components, and said shade point offset value;

q) defining end members from said reduced spectral data; and

r) generating said current class parameters from said end members.