Fractal features used with nearest neighbor clustering for identifying clutter in sonar images

• US 6,052,485 A
• Filed: 02/03/1997
• Issued: 04/18/2000
• Est. Priority Date: 02/03/1997
• Status: Expired due to Fees

• Current Assignee: 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: Nelson, Susan R., Tuovila, Susan M.
• Primary Examiner(s): Mancuso, Joseph
• Assistant Examiner(s): Miller, Martin E
• Application Number:
  US08/794,521
• Time in Patent Office: 1,170 Days
• Field of Search: 382/225, 382/103, 382/291, 382/249, 382/270, 382/275
• US Class Current: 382/225
• CPC Class Codes: G06V 10/255 Detecting or recognising potential candidate objects based on visual cues, e.g. shapes

Abstract

A method is presented for identifying clutter in an image such as a sonar age. A detection scheme identifies portions of the image having a signal-to-noise ratio greater than a given threshold. A classification scheme is then applied to each such portion to generate fractal feature values associated therewith in order to classify the portion as a target or non-target. For clutter identification, each portion is assigned to a group based on a distance from the position of each portion to another of the portions that is its nearest neighbor. A series of tests are then performed for each group using the fractal feature values associated with each portion in each group. A failure of any of the series of tests by a group identifies each portion associated with that group as clutter in the image.

Claims

1. A method of identifying clutter in an image represented by a two-dimensional array of pixel intensities, said method comprising the steps of:

applying a detection scheme to said two-dimensional array to identify portions of said image having a signal-to-noise ratio greater than a given threshold, wherein each of said portions is centered at a position in said image that is represented by a two-dimensional coordinate pair generated by said detection scheme;

applying a classification scheme to each of said portions to generate fractal feature values associated therewith, wherein each of said portions is classified as being one of a target and non-target based on said fractal feature values associated with each of said portions;

assigning each of said portions to a group based on a distance from said position of each of said portions to another of said portions that is its nearest neighbor; and

performing at least one test for each said group using said fractal feature values associated with each of said portions in each of said groups, wherein a failure of any of said at least one test by any one of each said group identifies each of said portions associated with said any one of each said group as clutter in said image.

2. A method according to claim 1 wherein said step of assigning includes the step of determining each said distance based on the number of said portions in said image.

3. A method according to claim 2 wherein said step of determining includes the steps of:

providing a baseline distance; and

altering said baseline distance based on the number of portions in said image.

4. A method according to claim 3 wherein said step of altering comprises the step of increasing said baseline distance when the number of said portions exceeds a given threshold number.

5. A method according to claim 3 wherein said step of altering comprises the step of decreasing said baseline distance when the number of said portions is less than or equal to a given threshold number.

6. A method according to claim 1 wherein, for each said group, said step of performing includes the steps of:

counting the number of said portions classified as being said target based on said classification scheme;

dividing the number of said portions classified as being said target based on said classification scheme by the total number of said portions associated with said group to generate a fraction; and

designating said each group and each of said portions in said each group as being one of said target and said non-target based on said fraction, wherein a designation of a group and its portions as being said non-target based on said fraction is indicative of said failure.

7. A method according to claim 6 wherein, for each said group and its portions designated as being said target based on said fraction, said step of performing further includes the steps of:

determining dimensions for each said group and its portions designated as being said target based on said fraction; and

designating each said group and its portions designated as being said target based on said fraction as being one of said target and said non-target based on said dimensions, wherein a designation of a group and its portions as being said non-target based on said dimensions is indicative of said failure.

8. A method according to claim 1 further comprising the step of normalizing said pixel intensities prior to said step of applying said detection scheme.

9. A method of identifying clutter in a sonar image comprising the steps of:

providing a two-dimensional array of pixel intensities representative of said sonar image;

normalizing said pixel intensities;

applying a detection scheme to said two-dimensional array to identify portions of said sonar image having a signal-to-noise ratio greater than a given threshold, wherein each of said portions is centered at a position in said image that is represented by a two-dimensional coordinate pair generated by said detection scheme;

applying a classification scheme to each of said portions to generate fractal feature values associated therewith, wherein each of said portions is classified as being one of a target and non-target based on said fractal feature values associated with each of said portions;

assigning each of said portions to a group based on a distance from said position of each of said portions to another of said portions that is its nearest neighbor; and

performing a series of tests for each said group using said fractal feature values associated with each of said portions in each of said groups, wherein a failure of any of said series of tests by any one of each said group identifies each of said portions associated with said any one of each said group as clutter in said sonar image.

10. A method according to claim 9 wherein said step of applying said classification scheme includes the steps of:

determining, for each said two-dimensional coordinate pair, fractal dimension values for each of at least two differently sized windows of said sonar image encompassing said two-dimensional coordinate pair; and

determining magnitude and direction of a gradient between said fractal dimension values of said at least two differently sized windows associated with each said two-dimensional coordinate pair.

11. A method according to claim 10 wherein said step of assigning includes the step of determining each said distance based on the number of said portions in said sonar image.

12. A method according to claim 11 wherein said step of determining includes the steps of:

providing a baseline distance; and

altering said baseline distance based on the number of portions in said sonar image.

13. A method according to claim 12 wherein said step of altering comprises the step of increasing said baseline distance when the number of said portions exceeds a given threshold number.

14. A method according to claim 12 wherein said step of altering comprises the step of decreasing said baseline distance when the number of said portions is less than or equal to a given threshold number.

15. A method according to claim 10 wherein, for each said group, said step of performing includes the steps of:

counting the number of said portions classified as being said target based on said classification scheme;

dividing the number of said portions classified as being said target based on said classification scheme by the total number of said portions associated with said group to generate a fraction; and

designating said each group and each of said portions in said each group as being one of said target and said non-target based on said fraction, wherein a designation of a group and its portions as being said non-target based on said fraction is indicative of said failure.

16. A method according to claim 15 wherein, for each said group and its portions designated as being said target based on said fraction, said step of performing further includes the steps of:

determining dimensions for each said group and its portions designated as being said target based on said fraction; and

designating each said group and its portions designated as being said target based on said fraction as being one of said target and said non-target based on said dimensions, wherein a designation of a group and its portions as being said non-target based on said dimensions is indicative of said failure.