METHOD FOR SEGMENTATION OF DIGITAL IMAGES

• US 20080144935A1
• Filed: 11/30/2007
• Published: 06/19/2008
• Est. Priority Date: 11/30/2006
• Status: Abandoned Application

• Current Assignee: Gilles Soulez, Jacques A De Guise, Ovalong Troeung, Ramnada Chav
• Original Assignee: Gilles Soulez, Jacques A De Guise, Ovalong Troeung, Ramnada Chav
• Inventors: Troeung, Ovalong, DE Guise, Jacques A., CHAV, Ramnada, Soulez, Gilles
• Application Number:
  US11/948,247
• Publication Number:
  US 20080144935A1
• Time in Patent Office: Days
• US Class Current: 382/173
• CPC Class Codes: G06T 2207/10132 Ultrasound image
• CPC Class Codes: G06T 2207/30016 Brain
• CPC Class Codes: G06T 2207/30101 Blood vessel; Artery; Vein; Vascular
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/149 involving deformable models, e.g. active contour models

Abstract

A method for segmenting elements on an image, comprising: i) creating an active contour from at least one identified point on the image; ii) defining a reference intensity value from the pixels of the active contour; iii) simplifying the image by comparing the pixels of the image to the reference intensity value to give a propagation value to the pixels of the image; and iv) propagating the active contour in a selected one of an expansion/contraction direction as a function of the propagation value of the pixels adjacent to the active contour in the image; wherein ii), iii) and iv) are repeated in the selected direction whereby the active contour finally represents the element segmented in the image.

Claims

1. A method for segmenting elements on an image, comprising:

i) creating an active contour from at least one identified point on the image;

ii) defining a reference intensity value from the pixels of the active contour;

iii;

simplifying the image by comparing the pixels of the image to the reference intensity value to give a propagation value to the pixels of the image; and

iv) propagating the active contour in a selected one of an expansion/contraction direction as a function of the propagation value of the pixels adjacent to the active contour in the image;

wherein ii), iii) and iv) are repeated in the selected direction whereby the active contour finally represents the element segmented in the image.

2. The method according to claim 1, wherein i) creating the active contour comprises obtaining manually entered markers and the expansion/contraction direction.

3. The method according to claim 1, wherein ii) defining the reference intensity value comprises establishing a threshold value for comparing pixels.

4. The method according to claim 3, wherein establishing the threshold value comprises calculating statistical data for given pixels within the propagating active contour as a function of the expansion/contraction direction.

5. The method according to claim 3, wherein iii) simplifying the image comprises giving a binary value as the propagation value as a function of the threshold value.

6. The method according to claim 1, wherein iii) simplifying the image comprises giving a binary value as the propagation value.

7. The method according to claim 1, further comprising filtering pixels within the propagating active contour to regularize the pixel intensity of elements within the propagating active contour.

8. The method according to claim 7, wherein said filtering pixels is performed periodically between a plurality of sequences of ii), iii) and iv).

9. The method according to claim 1, further comprising filtering pixels within the propagating active contour to regularize a boundary curvature of elements.

10. The method according to claim 9, wherein said controlling is performed periodically between a plurality of sequences of ii), iii) and iv).

11. The method according to claim 1, wherein i) creating the active contour comprises automatically setting markers and a default expansion/contraction direction.

12. The method according to claim 4, wherein establishing the threshold value comprises calculating an average intensity and a standard deviation as statistical data.

13. The method according to claim 1, wherein i) creating the active contour and iv) propagating the active contour comprises giving a binary value to pixels of the propagating active contour.

14. The method according to claim 1, wherein i) creating the active contour comprises obtaining a previously segmented image with the active contour being segmented elements from the segmented image, in segmenting video images.

15. The method according to claim 1, further comprising obtaining an inactive zone for the image, and excluding the propagating from the inactive zone.

16. The method according to claim 1, further comprising stopping the propagating of the active contour locally when an element has been locally segmented.

17. A method for segmenting elements on an image, comprising:

i) creating an active contour from at least one identified point on the image, points of the active contour having a binary value;

ii) defining a reference intensity value from the pixels of the active contour;

iii) comparing the pixels of the image to the reference intensity value to give a propagation value to the pixels of the image; and

iv) propagating the active contour in a selected one of an expansion/contraction direction as a function of the propagation value of the pixels adjacent to the active contour in the image, whereby points of the propagating active contour has a binary value;

wherein ii), iii) and iv) are repeated in the selected direction whereby the active contour finally represents the element segmented in the image.

18. The method according to claim 17, further comprising filtering pixels within the propagating active contour to regularize the pixel intensity of elements within the propagating active contour and to regularize a boundary curvature of elements.

19. The method according to claim 18, wherein said filtering pixels is performed periodically between a plurality of sequences of ii), iii) and iv).