Signal metric estimator

• US 5,369,599 A
• Filed: 08/04/1992
• Issued: 11/29/1994
• Est. Priority Date: 08/04/1992
• Status: Expired due to Term

• Current Assignee: Honeywell Incorporated (Honeywell International Inc.)
• Original Assignee: Honeywell Incorporated (Honeywell International Inc.)
• Inventors: Sadjadi, Firooz A., Bazakos, Mike E., Nasr, Hatem
• Primary Examiner(s): Harvey, Jack B.
• Assistant Examiner(s): PEESO, THOMAS R
• Application Number:
  US07/925,058
• Time in Patent Office: 847 Days
• Field of Search: 364/516, 382/17, 382/28, 382/27, 382/15, 382/39
• US Class Current: 382/103
• CPC Class Codes: G01S 13/89 for mapping or imaging
• CPC Class Codes: G01S 17/89 for mapping or imaging
• CPC Class Codes: G06F 18/21 Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
• CPC Class Codes: G06T 7/60 Analysis of geometric attributes

Abstract

A performance measure of an ATR is made from ancillary target data and the ATR output. A parameter of the ATR is varied to determine the change an ATR performance due to the parameter variation. Separate performance in the form of a quadratic equation models provide performance as a function of parameter and metrics. The performance model is partially differentiated with respect to the parameter. The partial differentiation allows solution for the estimated metric.

Claims

1. A signal metric estimator apparatus comprising:

automatic target recognition (ATR) means for receiving a plurality of sequential multidimensional signals of a scene, and providing ATR observables, said ATR having a plurality of adjustable parameters;

ancillary data means for providing non-sensory informational data for said scene;

performance measure estimating means connected to said ATR means for receiving said ATR observables and connected to said ancillary data means for receiving said non-sensory informational data, said performance measure estimating means providing a first performance measure estimate for said ATR;

a first performance model as a function of parameters and metrics, said first performance model having coefficients;

means for varying a first parameter of said ATR from a first value to a second value;

means for determining a rate of change of a variation in said first performance measure estimate as a result of varying said first parameter from a first value to a second value; and

inverting means for using said first parameter first value, said rate of change of said variation in said first performance measure estimate, said performance model coefficients, and said first performance model to estimate a first metric value.

2. Apparatus of claim 1 wherein said performance measure estimating means includes means for determining a frequency of a target being tracked exceeding a threshold value.

3. Apparatus of claim 1 wherein said inverting means includes partial differential means for taking a partial derivative of said first performance model with respect to said first parameter.

4. A signal metric estimator apparatus comprising:

automatic target recognition (ATR) means for receiving a plurality of sequential images of a scene and providing ATR observables, said ATR having a plurality of adjustable parameters;

ancillary data means for providing non-sensory informational target data for said scene;

performance measure estimating means connected to said ATR means for receiving said ATR observables and connected to said ancillary data means for receiving said non-sensory informational target data, said performance measure estimating means comparing said ATR observables to said non-sensory informational target data and said performance measure estimating means determining the frequency of tracking of a predicted target, said performance measure estimating means providing a first performance measure estimate for said ATR;

a first performance model for providing a corresponding model for said first performance measure estimate, said first performance model being a quadratic function of a first parameter of said ATR and of a first metric;

means for varying said first parameter from a first value to a second value;

means for determining a rate of a change of said first performance measure estimate as a result of varying said first parameter from a first value to a second value; and

means for using said first parameter first value, said rate of change of said first performance estimate, said performance model coefficients, and said first performance model to estimate said first metric.

5. Apparatus of claim 4 wherein said means for using said first parameter first value, said rate of change of said first performance estimate, said first performance model coefficients, and said first performance model includes partial derivative means for taking a partial derivative of said first performance model with respect to said first parameter.

6. Apparatus of claim 5 wherein said means for using said first parameter first value, said rate of change of said first performance estimate, said first performance model coefficients and said first performance model includes means for determining an estimated metric from said partial derivative.

7. A signal metric estimator apparatus for use with an Automatic Target Recognition (ATR) system having a plurality of adjustable parameters, said ATR receiving a plurality of sequential images of a scene and providing ATR observables, said signal metric estimator comprising:

ancillary data means for providing non-sensory information data for said scene;

performance measure estimating means connected to said ATR for receiving said ATR observables and connected to said ancillary data means for receiving said non-sensor informational data, said performance measure estimating means comparing said ATR observables to said non-sensory informational data and said performance measure estimating means determining the frequency of tracking of a predicted target, said performance measure estimating means providing a first performance measure estimate for said ATR;

a first performance model for providing a corresponding model for said first performance measure estimate, said first performance model being a quadratic function of a first parameter of said ATR and of a first metric;

means for varying said first parameter from a first value to a second value;

means for determining a rate of change of said first performance measure estimate as a result of varying said first parameter from a first value to a second value; and

means for using said first parameter first value, said rate of change of said first performance estimate, said performance model coefficients, and said first performance model to estimate said first metric.

8. Apparatus of claim 7 wherein said means for using said first parameter first value, said rate of change of said first performance estimate, said performance model coefficients, and said first performance model includes partial derivative means for taking a partial derivative of said first performance model with respect to said first parameter.

9. Apparatus of claim 8 wherein said means for using said first parameter first value, said rate of change of said first performance estimate, said first performance model coefficients and said first performance model includes means for determining an estimated metric from said partial derivative.