Method of image segmentation

• US 20100272357A1
• Filed: 01/24/2007
• Published: 10/28/2010
• Est. Priority Date: 07/28/2006
• Status: Active Grant

• Current Assignee: University of New Brunswick
• Original Assignee: University of New Brunswick
• Inventors: Maxwell, Travis Linden, Zhang, Yun
• Granted Patent: US 8,233,712 B2
• Time in Patent Office: Days
• US Class Current: 382/173
• CPC Class Codes: G06T 2207/10036 Multispectral image; Hyperspectral image
• CPC Class Codes: G06T 2207/20016 Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
• CPC Class Codes: G06T 2207/30181 Earth observation
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/136 involving thresholding
• CPC Class Codes: G06V 10/26 Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
• CPC Class Codes: G06V 10/58 relating to hyperspectral data
• CPC Class Codes: G06V 20/13 Satellite images
• CPC Class Codes: G06V 20/194 using hyperspectral data, i.e. more or other wavelengths than RGB

Abstract

A method of segmenting a digital image comprising the steps of performing a preliminary segmentation of the image into sub objects, defining a model object by selecting sub objects that define the model object, providing sub-object and model object features, using a fuzzy logic inference system to calculate segmentation parameters based on at least one of the sub object and model object features, and performing segmentation of the image using the segmentation parameters.

Claims

1. Use of a fuzzy logic inference system to calculate segmentation parameters for segmenting a digital image.

2. The use according to claim 1 wherein the segmentation parameters are smoothness, scale and shape.

3. A method of segmenting a digital image comprising the steps of:

(a) performing a preliminary segmentation of the image into sub objects;

(b) defining a model object by selecting sub objects that define the model object;

(c) providing sub-object and model object features;

(d) using a fuzzy logic inference system to calculate segmentation parameters based on at least one of the sub object and model object features; and

(e) performing segmentation of the image using the segmentation parameters.

4. The method according to claim 3 wherein the segmentation parameters are smoothness, scale and shape.

5. The method according to claim 3 including repeating steps (b), (c) and (d).

6. The method according to claim 3 including the steps of:

(f) setting a threshold for the point at which a sub object ceases to be considered part of the model object;

(g) performing a feature discrepancy measure between at least one of the sub object features in the segmented image and at least one of the model object features for convergence; and

(h) repeating step (d) if convergence is not achieved.

7. The method according to claim 6 wherein the feature discrepancy measure is based on scale and size of the objects.

8. The method according to claim 3 wherein the preliminary segmentation is performed using boundary-based image segmentation or region-based image segmentation.

9. The method according to claim 8 wherein the boundary-based image segmentation includes an optimal edge detection or watershed segmentation, and the region-based image segmentation includes a multilevel thresholding or a region-growing method.

14. The method according to claim 3 further including the step of:

(d) performing segmentation of the image using the segmentation parameters.

15. The method according to claim 14 further including the step of:

repeating step (c).

16. The method according to claim 15 further including the step of:

testing the image segmentation for convergence to the model object.

17. A computer readable memory having recorded thereon statements and instructions for execution by a computer to carry out the method of claim 3.

10. A method of segmenting an image comprising the steps of:

(a) performing a preliminary segmentation;

(b) providing input segmentation parameters;

(c) using a fuzzy logic inference system using the input segmentation parameters and object features to evaluate new segmentation parameters; and

(d) performing segmentation of the image using the new parameters.

11. The method according to claim 10 including the step of:

(e) testing for convergence and repeating step (c).

12. The method according to claim 11 including the step of repeating step (d).

18. A computer readable memory having recorded thereon statements and instructions for execution by a computer to carry out the method of claim 10.

13. A method of calculating segmentation parameters for segmenting an initially segmented digital image comprising the steps of:

(a) defining a model object by selecting sub objects that define the model object;

(b) providing sub-object and model object features; and

(c) using a fuzzy logic inference system to calculate segmentation parameters based on at least one of the sub object features and the model object features.

19. A computer readable memory having recorded thereon statements and instructions for execution by a computer to carry out the method of claim 13.

20. A system for segmenting a digital image comprising:

a fuzzy logic inference system configured to define a model object by selecting sub-objects that define a model object, calculate segmentation parameters based on at least one sub object and model object feature, and calculate segmentation parameters based on at least one of the sub object features and the model object features.

21. A system for segmenting a digital image according to claim 20 comprising:

a segmentation module configured to segment the image using the segmentation parameters.

22. A system for segmenting a digital image according to claim 21 comprising:

a preliminary segmentation module configured to initially segment the image into sub-objects.