Granted Patent B2

US 8,144,980 · App. 12/942,258 · Granted Mar 27, 2012

Abstract

Claims (68)

1. A method comprising:

(a) retrieving from a memory an image for graphic image processing on a system comprising the memory and a processor, the image comprising a plurality of pixels and an element to be processed;

(b) establishing with the system a search region within the image encompassing a predetermined portion of the element to be processed and containing pixels on each side of a boundary between the element and the remainder of the image;

(c) selecting a predetermined portion of the pixels within the search region to create an analysis group;

(d) determining with the system a plurality of colour regions for the analysis group corresponding to a plurality of pixel classifications, wherein an error function represents a probability that each of the pixels within the analysis group belongs to one of the plurality of pixel classifications;

(e) constructing with the system an edge map by establishing an initial boundary of the element; and

(f) creating with the system a boundary contour in dependence upon establishing a target condition for an energy of an active contour, the energy of the active contour established in dependence upon at least the error function and the edge map.

2. The method according to claim 1 wherein,

step (b) comprises establishing a search region associated with an edge or boundary of the element in the image.

3. The method according to claim 1 wherein,

step (b) comprises establishing the search region in dependence upon at least one of applying a predetermined process to the pixels of the image, a user input, and applying a predetermined geometric shape to the image.

4. The method according to claim 1 wherein,

step (e) comprises establishing the initial boundary in dependence upon image gradients calculated from at least one of an error function relating to one of the plurality of pixel classifications and applying a predetermined process to the predetermined portion of the pixels within the search region.

5. The method according to claim 4 wherein,

calculating the image gradients is performed at a different scale to that of the source image by filtering the at least one of.

6. The method according to claim 1 wherein,

step (f) comprises establishing the target condition for the energy of the active contour by applying empirical weighting coefficients to each of the error function and the edge map.

7. The method according to claim 1 wherein,

step (b) further comprises rotating the search region to create a one-dimensional active contour at least one of initially or after step (e) has been executed to generate the initial boundary and steps (b) onwards are subsequently repeated.

8. A method comprising:

(a) retrieving from a memory an image for graphic image processing on a system comprising the memory and a processor, the image comprising a plurality of pixels and an element to be processed;

(b) establishing a first search region within the image encompassing a predetermined portion of the element to be processed and containing at least a pixel on each side of a boundary between the element and the remainder of the image;

(c) selecting with the system a first predetermined portion of the pixels within the first search region to create an analysis group;

(d) processing the analysis group to establish the boundary in dependence upon an initial boundary contour;

(e) repeating steps (b), (c) and (d) with the system wherein the first search region is replaced with a second search region that is defined at a different scale to the first search region, encompasses the boundary, and contains at least a pixel on each side of the boundary and the initial boundary contour is the established boundary.

9. The method according to claim 8 wherein,

step (d) comprises the steps of:

(1) selecting with the system a predetermined portion of the pixels within the first search region to create an analysis group;

(2) determining with the system a plurality of colour regions for the analysis group corresponding to a plurality of pixel classifications, wherein an error function represents a probability that each of the pixels within the analysis group belongs to one of the plurality of pixel classifications;

(3) constructing with the system an edge map in dependence upon the initial boundary contour; and

(4) establishing the boundary in dependence upon establishing a target condition for an energy of an active contour, the energy of the active contour established in dependence upon at least the error function and the edge map.

10. The method of claim 8 further comprising;

repeating the method defined by steps (b) through (e) until a predetermined condition is satisfied.

11. The method according to claim 8 wherein,

step (b) comprises at least one of establishing the search region to determine an edge of the element, applying a predetermined process to the pixels of the image, receiving user input, and using a predetermined geometric shape.

12. The method according to claim 8 wherein,

step (d) comprises establishing the initial boundary in dependence upon image gradients calculated from at least one of an error function relating to one of the plurality of pixel classifications and applying a predetermined process to the predetermined portion of the pixels within the search region.

13. The method according to claim 12 wherein,

calculating the image gradients is performed at a different scale to that of first search region.

14. The method according to claim 9 wherein,

step (4) comprises establishing the target condition for the energy of the active contour by applying empirical weighting coefficients to each of the error function and the edge map.

15. The method according to claim 8 wherein,

step (b) further comprises rotating the search region to create a one-dimensional active contour at least one of initially, after step (d) has been executed to generate the initial boundary, and during step (e) to provide the second search region.

16. A method comprising:

(a) retrieving from a memory an image for graphic image processing on a system comprising the memory and a processor, the image comprising a plurality of pixels and an element to be processed;

(b) establishing a first search region within the image encompassing a first predetermined portion of the element to be processed and containing at least a pixel on each side of a boundary between the element and the remainder of the image;

(c) establishing a second search region within the image;

(d) selecting with the system a first predetermined portion of the pixels within the first search region to create an analysis group;

(e) selecting with the system a second predetermined portion of the pixels within the second search region to create a training group;

(f) establishing a plurality of color regions corresponding to a plurality of pixel classifications in dependence upon the training group; and

(g) processing the analysis group to establish the boundary in dependence upon an initial boundary contour, the plurality of pixel classifications, and the analysis group.

17. The method according to claim 16 wherein,

step (g) comprises the steps of:

(1) determining with the system an error function representing a probability that each of the pixels within the analysis group belongs to one of the plurality of pixel classifications;

(2) constructing with the system an edge map by establishing an initial boundary of the element in dependence of the initial boundary contour; and

(3) establishing with the system a boundary contour in dependence upon establishing a target condition for an energy of an active contour, the energy of the active contour established in dependence upon at least the error function and the edge map.

18. The method according to claim 16 wherein,

step (b) comprises:

(4) selecting a third search region encompassing a second predetermined portion of the element; and

(5) dividing the third search region into a plurality of sub-divided regions, one of the plurality of sub-divided regions being the first search region and another of the plurality of sub-divided regions being an adjacent region to the first search region.

19. The method according to claim 16 wherein,

step (c) comprises:

(6) selecting a third predetermined portion of the pixels within the first search region that are adjacent a region boundary between the first search region and another region; and

(7) selecting a fourth predetermined portion of the pixels within the adjacent region adjacent the region boundary.

20. The method according to claim 16 further comprising;

(h) repeating steps (b) through (g) wherein,

the first search region is replaced with a fourth search region that is defined at a different scale to the first search region, encompasses a predetermined portion of the established boundary, and contains at least a pixel on each side of the established boundary; and

the initial boundary contour is the predetermined portion of the established boundary.

Assignments (7)