Refinement of segmentation markup

• US 8,670,615 B2
• Filed: 09/30/2010
• Issued: 03/11/2014
• Est. Priority Date: 09/30/2009
• Status: Expired due to Fees

• Current Assignee: Flashfoto, Inc.
• Original Assignee: Flashfoto, Inc.
• Inventors: Piramuthu, Robinson, Lee, Kuang-chih, Prochazka, Daniel Richard
• Primary Examiner(s): Mehta, Bhavesh
• Assistant Examiner(s): Niu, Feng
• Application Number:
  US12/924,705
• Publication Number:
  US 20110075926A1
• Time in Patent Office: 1,258 Days
• Field of Search: None
• US Class Current: 382/173
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/194 involving foreground-background segmentation

Abstract

Systems for refinement of a segmentation of an image using spray-paint markup are disclosed, with methods and processes for making and using the same. Spray-paint markup allows for easy markup of errors in a segmentation. The markup'"'"'s data may be correlated or harmonized with the representation of the segmentation, such that it may be applied to the segmentation. The markup'"'"'s data is utilized to refine the segmentation errors. To assist in the refinement process, the resolutions may be scaled down so as to exert less computation burden during the process.

Claims

1. A method comprising:

receiving a binary mask associated with a segmentation of an image;

receiving a user selection, via a user input device, of a portion of said image to be corrected;

creating a binary markup mask associated with said user selection; and

applying said binary markup mask to said binary mask associated with said segmentation of said image to revise said binary mask associated with said segmentation of said image to locate errors in said segmentation using said binary markup mask.

2. The method as described in claim 1, wherein said user selection is a spray paint, and wherein said user input device is a computer mouse.

3. The method as described in claim 1, wherein said creating comprises:

rasterizing samples associated with said user selection to create a binary trajectory mask;

smoothing said binary trajectory mask to create a fuzzy markup mask; and

enabling positive pixels in said fuzzy markup mask to create said binary markup mask.

4. The method as described in claim 1 further comprising:

determining uncertainty associated with said user selection, wherein said uncertainty is based on a width of a graphical rendition on a graphical user interface that is associated with said user input device, and wherein said uncertainty is further based on sloppiness associated with said user selection.

5. The method as described in claim 1 further comprising:

determining uncertainty associated with said user selection based on Gaussian random processes.

6. The method as described in claim 1 further comprising:

determining said revision to be associated with addition if a ratio of an overlap region between said image and said user selection is smaller than a first threshold;

determining said revision to be associated with deletion if said ratio of said overlap region is larger than a second threshold; and

determining said revision to be associated with refinement if said ratio of said overlap region is larger than said first threshold and smaller than said second threshold.

7. A non-transitory computer readable storage medium having computer-readable program code stored thereon for causing a processor to implement a method comprising:

receiving a binary mask associated with a segmentation of an image;

receiving a user selection, via a user input device, of a portion of said image to be corrected;

creating a binary markup mask associated with said user selection;

applying said binary markup mask to said binary mask associated with said segmentation of said image to revise said binary mask associated with said segmentation of said image; and

determining uncertainty associated with said user selection, wherein said uncertainty is based on a width of a graphical rendition on a graphical user interface that is associated with said user input device, and wherein said uncertainty is further based on sloppiness associated with said user selection.

8. The non-transitory computer readable storage medium as described in claim 7, wherein said user selection is a spray paint, and wherein said user input device is a computer mouse.

9. The non-transitory computer readable storage medium as described in claim 7, wherein said creating comprises:

rasterizing samples associated with said user selection to create a binary trajectory mask;

smoothing said binary trajectory mask to create a fuzzy markup mask; and

enabling positive pixels in said fuzzy markup mask to create said binary markup mask.

10. The non-transitory computer readable storage medium as described in claim 7, wherein said method further comprises:

determining uncertainty associated with said user selection based on Gaussian random processes.

11. The non-transitory computer readable storage medium as described in claim 7, wherein said method further comprises:

locating errors using said binary markup mask.

12. The non-transitory computer readable storage medium as described in claim 7, wherein said method further comprises:

determining said revision to be associated with addition if a ratio of an overlap region between said image and said user selection is smaller than a first threshold;

determining said revision to be associated with deletion if said ratio of said overlap region is larger than a second threshold; and

determining said revision to be associated with refinement if said ratio of said overlap region is larger than said first threshold and smaller than said second threshold.

13. A computer system comprising:

a processor;

a display; and

a memory coupled to said processor and said display, wherein said memory comprises computer-readable program code stored thereon for causing said processor to implement a method comprising;

receiving a binary mask associated with a segmentation of an image;

receiving a user selection, via a user input device, of a portion of said image to be corrected;

creating a binary markup mask associated with said user selection;

applying said binary markup mask to said binary mask associated with said segmentation of said image to revise said binary mask associated with said segmentation of said image; and

determining uncertainty associated with said user selection based on Gaussian random processes.

14. The computer system as described in claim 13, wherein said creating comprises:

rasterizing samples associated with said user selection to create a binary trajectory mask;

smoothing said binary trajectory mask to create a fuzzy markup mask; and

enabling positive pixels in said fuzzy markup mask to create said binary markup mask.

15. The computer system as described in claim 13, wherein said method further comprises:

determining uncertainty associated with said user selection, wherein said uncertainty is based on a width of a graphical rendition on a graphical user interface that is associated with said user input device, and wherein said uncertainty is further based on sloppiness associated with said user selection.

16. The computer system as described in claim 13, wherein said method further comprises:

locating errors using said binary markup mask.

17. The computer system as described in claim 16, wherein said method further comprises:

determining said revision to be associated with addition if a ratio of an overlap region between said image and said user selection is smaller than a first threshold;

determining said revision to be associated with deletion if said ratio of said overlap region is larger than a second threshold; and

determining said revision to be associated with refinement if said ratio of said overlap region is larger than said first threshold and smaller than said second threshold.