High probability of detection, low false alarm multisensor ATR

• US 4,905,292 A
• Filed: 09/14/1987
• Issued: 02/27/1990
• Est. Priority Date: 09/14/1987
• Status: Expired due to Fees

• Current Assignee: The United States Of America As Represented By The Secretary Of The Army
• Original Assignee: The United States Of America As Represented By The Secretary Of The Army
• Inventors: Wentworth, Edwin W., Horn, Stuart B.
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): Couso, Jose L.
• Application Number:
  US07/095,784
• Time in Patent Office: 897 Days
• Field of Search: 244/3.16, 244/3.19, 102/427, 102/214, 102/213, 102/211, 250/330, 250/330.1, 356/51, 358/113, 382/1, 382/10, 382/48, 342/53, 342/54
• US Class Current: 382/103
• CPC Class Codes: G06F 18/256 of results relating to different input data, e.g. multimodal recognition

Abstract

An Aided Target Recognition System (ATR) is provided wherein the recognit abilities of two ATR sensor subsystems with mutually exclusive false alarm characteristics are fused by a decision circuit to provide an improved false alarm rate versus the probability of target detection. The sensors, which respond to different types of radiation, have different false alarm patterns for the same set of targets and a decision circuit which contains logical function elements to exploit the difference in these patterns.

Claims

1. An improved ATR system for detecting, identifying and prioritizing targets from a plurality of detected targets, including:

a plurality of sensor means for detecting mutually exclusive sets of waves emanating from the environment including said targets;

all of said sensors being passive and having response thresholds low enough to detect all possible targets including a very large number of false targets;

a sub-ATR circuit means coupled to each of said sensors for locating, classifying, prioritizing and listing a separate preselected quota in each of a preselected number of classes of targets detected by said sensors, in accordance with prerecorded target signatures and other user constraints,

a first of said sensors having a lower probability of detection for a given false alarm rate than a second of said sensors, but based on a different distribution pattern of false alarms among said targets;

a final decision means coupled to all of said sensors for reassigning the classifications and priorities of targets detected by both of said sensors according to a preselected set of logical classification rules.

2. A system as in claim 1, wherein:

said first sensor means detects light in the visible and near visible region of the EMR spectrum.

3. A system as in claim 1, wherein:

a second sensor means detects light in the far-infrared region of the EMR spectrum.

4. The system according to claim 1 including:

means to transmit the locations of targets detected by one of said sensors to another of said other sensors, such that all sensors detect all said listed targets.

5. The method of identifying high priority targets detected only by a plurality of passive sensors having different patterns of false alarms among said targets, comprising the steps of:

a. setting the detection thresholds of said sensors to produce a very high probability of detection;

b. recording the locations of said targets as primary identification;

c. classifying said targets according to target signatures appropriate to the target environment;

d. prioritizing said targets with user supplied criteria and calculated level of confidence in their classification;

e. listing at most a preselected quota of targets for each of said sensors in accordance with their priority and user supplied constraints for each class of targets; and

f. repeating steps B, C, and D for all listed targets from all sensors according to a preselected set of logical classification steps.

6. The method according to claim 5 wherein said sensors comprises a FLIR and a TV camera;

the possible target classifications comprise a tank (T), a personnel carrier (A), a truck (W), one of the preceeding targets (?), unknown (U), low or no priority (NP) and high priority (P); and

said logical steps comprise;

E1. If both sensor circuitry make the same classification, this becomes the final classification;

E2. If the FLIR circuitry classifies the targets as (T) and the TV circuitry as (W), then the final classification is (W);

E3. If the FLIR circuitry classifies the targets as (U) and the TV circuitry as (?), then the final classification is (?);

E4. If the FLIR circuitry classifies the targets as (?) and the TV circuitry as (U), then the final classification is (?);

E5. If the FLIR circuitry classifies the targets as (A) and the TV circuitry as (W), then the final classification is (W);

E6. If the FLIR circuitry classifies the targets as (NP) and the TV circuitry as (P), then the final classification is (P);

E7. If none of the above apply the FLIR circuitry classification becomes the final classification.