Object classification

• US 9,177,200 B2
• Filed: 05/13/2010
• Issued: 11/03/2015
• Est. Priority Date: 05/14/2009
• Status: Expired due to Fees

• Current Assignee: BAE Systems Plc
• Original Assignee: BAE Systems Plc
• Inventors: Barr, Jordi McGregor, Lloyd, Christopher Mark, Williams, Mark Lawrence, Nicholson, David
• Primary Examiner(s): Rivas, Omar Fernandez
• Assistant Examiner(s): DAY, HERNG DER
• Application Number:
  US13/319,926
• Publication Number:
  US 20120053914A1
• Time in Patent Office: 2,000 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G01S 15/88 Sonar systems specially adapted for specific applications seismic or acoustic prospecting or detecting G01V1/00
• CPC Class Codes: G01S 7/539 using analysis of echo signal for target characterisation; Target signature; Target cross-section
• CPC Class Codes: G06F 18/241 relating to the classification model, e.g. parametric or non-parametric approaches
• CPC Class Codes: G06F 2218/00 Aspects of pattern recognition specially adapted for signal processing

Abstract

A method and apparatus for selecting a value or change in value of a measurement variable for an observation of an object, comprising: receiving models for the object defined in terms of an observation parameter and a measurement variable; selecting values of the measurement variable; for each model, determining a value of the observation parameter for each selected value; for each selected value, determining a value of an expected classification potential level using the determined values; and selecting a value of the measurement variable dependent upon the potential level values; wherein the potential level is an expected level of: the information or lack of information, and/or the certainty or uncertainty, with which the object could be classified if a measurement of the observation parameter were taken of the object at the respective value of the measurement variable.

Claims

1. A method for selecting an optimal observation position for a mine-hunting vessel from which to observe an object to be classified, said method comprising:

Using a plurality of empirically derived models for the object to be classified, said models comprising observation parameters defined as observed physical quantities of said object as a function of a measurement variable representing an observer'"'"'s angular position relative to said object, and each said model representing a different class of object, said observer being said mine-hunting vessel;

Selecting a plurality of observer angular positions from which to observe said object;

Determining a respective corresponding value of observation parameter for each selected observer angular position according to each model in said plurality of empirically derived models;

Determining a value of a probability function using said respective corresponding value of observation parameter for each said selected observer angular positions according to each said model, wherein said value of said probability function indicates a classification potential of said object;

Wherein said classification potential of said object is relative;

(i) information or lack of information, for use in a process leading toward classification of said object as one of said models, that is to be obtained if a measurement of the observation parameter were taken of the object at a respective observer angular position; and

/or

(ii) certainty or uncertainty with which the object could be classified as one of the models if the measurement of the observation parameter were taken of the object at the respective observer angular position; and

Selecting an observation position for said mine hunting vessel from which to observe said object from said plurality of selected observer angular positions with a maximum said value of said probability function.

2. The method according to claim 1, further comprising receiving respective classification potential that each respective model corresponds to the object;

and wherein determining the value of the function further comprises using the received respective classification potential.

3. The method according to claim 1, wherein the classification potential is an expected entropy value.

4. The method according to claim 3, wherein the expected entropy value is an expected Shannon entropy value.

5. The method according to claim 1, further comprising measuring a value of the observation parameter of the object to be classified at the selected value/change in value of the measurement variable.

6. The method according to claim 1, further comprising using a classifier in which an observation of the object to be classified is performed at the selected value/change in value of the measurement variable and in which the resulting observed observation parameter value is used in a process of providing respective classification potentials that each respective model corresponds to the object.

7. The method according to claim 6, further comprising using the provided respective classification potential in at least one further iteration of the method of selecting a value or change in value of the measurement variable to be used for an observation of an object to be classified, the method comprising:

receiving a plurality of models for the object defined in terms of an observation parameter as a function of the measurement variable;

selecting a plurality of values of the measurement variable or change in the measurement variable;

for each model, determining a respective corresponding value of the observation parameter for each of the selected values of the measurement variable/changes in the measurement variable;

for each selected value of the measurement variable/change in the measurement variable, determining a value of a second probability function using the determined values of the observation parameter corresponding to each of the models, wherein the second probability function provides that each determined value of the function is indicative of an expected classification potential; and

selecting a value or change in value of the measurement variable for an observation of the object to be classified dependent upon the determined classification potential level values;

wherein the classification potential is a relative expected level of;

(i) the information or lack of information, for use in a process leading toward classification of the object as one of the models, that is to be obtained if the measurement of the observation parameter were taken of the object at a respective value of the measurement variable; and

/or

(ii) the certainty or uncertainty with which the object could be classified as one of the models if the measurement of the observation parameter were taken of the object at the respective value of the measurement variable in which determining the value of the function further comprises using the provided respective classification potential.

8. The method according to claim 1 wherein classification of the object comprises using a particle filter used to weight particles on the plurality of models for the object.

9. The method according to claim 1, wherein the classification potential is calculated using the following formula:

where;

10. The method according to claim 1, wherein the selected value or change in value of the measurement variable for an observation of an object to be classified dependent upon the determined classification potential values is determined using the following formula:

11. A non-transitory machine readable storage medium storing a computer program or at least one of a plurality of computer programs arranged such that when executed by a computer system at least one of the computer programs causes the computer system to operate in accordance with the method of claim 1.

12. An apparatus to perform a method of determining selection criteria of the measurement variable according to claim 1.