Virtual sensor for data and sensor fusion

• US 7,283,938 B1
• Filed: 02/16/2006
• Issued: 10/16/2007
• Est. Priority Date: 04/15/2004
• Status: Expired due to Fees

• Current Assignee: Lockheed Martin Corporation (Martin Marietta Corporation)
• Original Assignee: Lockheed Martin Corporation (Martin Marietta Corporation)
• Inventors: Friesel, Mark A.
• Primary Examiner(s): Bui; Bryan
• Application Number:
  US11/356,670
• Time in Patent Office: 607 Days
• Field of Search: 702/104, 702/188, 702/182, 702/190, 702/191, 702/189, 356/73.1, 356/106, 455/226.3
• US Class Current: 702/189
• CPC Class Codes: G06F 18/251 of input or preprocessed data

Abstract

A plurality of sensors observe an object, and the raw sensor data is processed to produce evidence signals representative of characteristics which may be used to classify the object as to type. The sensor response characteristics from the plurality of sensors are fused to generate fused or combined sensor response characteristics. Thus, the fused or combined sensor response characteristics are equivalent to the sensor response characteristics of a virtual sensor. The evidence and fused sensor response characteristics are applied to a taxonomic classifier to determine the object type.

Claims

1. A method for classifying an object observed by at least first and second sensors by fusing information from said at least first and second sensors, said method comprising the steps of:

fusing the response characteristics from the first and second sensors to produce compound sensor response characteristics;

observing an object with said first and second sensors, each of which sensors independently (a) evaluates evidence or information and (b) based on said evidence or information, assigns a taxonomic (type) classification to its observation of said object;

assigning a classification to said object based on said evidence and said compound sensor response characteristics to indicate the type of said object.

2. A method according to claim 1, wherein said classification assigned to said object based on said evidence and said compound sensor response characteristics is a taxonomic classification.

3. A method according to claim 1, wherein said sensors are spatially separated.

4. A method for classifying an object observed by at least first and second sensors by fusing information from said at least first and second sensors, said method comprising the steps of:

fusing the response characteristics from the first and second sensors to produce compound sensor response characteristics;

observing an object with said first and second sensors to thereby generate evidence or information, each of which sensors independently (a) evaluates said evidence or information and (b) based on said evaluation of said evidence or information, assigns a taxonomic (type) classification to its observation of said object;

assigning a classification to said object based on said compound sensor response characteristics and at least one of (a) unfused evidence or information from said first and second sensors and (b) unfused taxonomic information from said first and second sensors to indicate the type of said object.

5. The method of claim 4, wherein said sensors are spatially separated.