System and method for adaptively separating foreground from arbitrary background in presentations

• US 7,668,371 B2
• Filed: 05/23/2008
• Issued: 02/23/2010
• Est. Priority Date: 01/13/2005
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Li, Ying, Dorai, Chitra
• Primary Examiner(s): Bali; Vikkram
• Assistant Examiner(s): Yeh; Eueng-Nan
• Application Number:
  US12/126,262
• Publication Number:
  US 20080219554A1
• Time in Patent Office: 641 Days
• Field of Search: 382/103, 382/173, 348/169
• US Class Current: 382/173
• CPC Class Codes: G06V 20/00 Scenes; Scene-specific elements control of digital cameras H04N23/60

Abstract

System and method for distinguishing between foreground content and background content in an image presentation. An initial background model is provided, and a final background model is constructed from the initial background model using the image presentation. The foreground content and background content in the image presentation are then distinguished from one another using the final background model. The present invention permits foreground content and background content to be separated from one another for further processing in different types of computer-generated image presentations such as digital slide presentations, video presentations, Web page presentations, and the like.

Claims

1. A computer-implemented method, implementing instructions stored in a recordable-type computer readable storage medium in an image processing system, for distinguishing between foreground content and background content in an image presentation, the computer-implemented method comprising:

providing an initial background model on a data processing system that is independent of any particular image presentation;

identifying presentation pages of a set of presentation pages of the image presentation, wherein the foreground content of each identified presentation page of the set of presentation pages are distinct and the background content of each identified presentation page of the set of presentation pages are the same, and wherein the presentation pages are part of a media stream provided to an image processor, wherein the media stream comprises a plurality of image presentation types;

clustering a plurality of content portions from the identified presentation pages into a plurality of distinct presentation content sets, wherein each presentation content set contains content portions that have a same image presentation type;

locating potential background pixels by comparing a first presentation page to a second presentation page, wherein the first presentation page and the second presentation page are located in a same set in the plurality of distinct presentation content sets;

updating the initial background model by incorporating the located potential background pixels into the initial background model to form an updated initial background model;

determining whether the updated initial background model and the initial background model are the same;

responsive to determining that the updated initial background model and the initial background model are the same;

outputting the updated initial background model;

distinguishing between the foreground content and the background content in each image presentation page of the plurality of presentation pages using the updated initial background model; and

repeating the steps of comparing the clustered content portions through the step of distinguishing between the foreground content and the background content for each remaining set of the distinct presentation content sets.

2. The computer-implemented method according to claim 1, and further comprising:

responsive to determining that the updated initial background model and the initial background model are not the same;

identifying further presentation pages of the set of presentation pages whose foreground content are distinct and whose background content are the same;

comparing the identified further presentation pages to locate further potential background pixels;

further updating the updated initial background model by incorporating the further located potential background pixels into the updated initial background model to form a further updated initial background model;

outputting the further updated initial background model; and

distinguishing between the foreground content and the background content in each presentation page of the plurality of presentation pages using the further updated initial background model.

3. The computer-implemented method according to claim 2, and further comprising:

repeating the steps of identifying further presentation pages, comparing the identified further presentation pages and further updating the updated initial background model until an updated background model reaches a steady state.

4. The computer-implemented method according to claim 1, wherein the image presentation comprises a computer-generated image presentation.

5. The computer-implemented method according to claim 4, wherein the computer-generated image presentation comprises one of a digital slide presentation, a Web page presentation and a video presentation.

6. An image processing system for distinguishing between foreground content and background content in an image presentation, comprising:

a storage medium storing a set of instructions; and

a processor, wherein the processor executes the set of instructions to provide an initial background model on a data processing system that is independent of any particular image presentation;

to identify presentation pages of a set of presentation pages of the image presentation, wherein the foreground content of each identified presentation page of the set of presentation pages are distinct and the background content of each identified presentation page of the set of presentation pages are the same, and wherein presentation pages are part of a media stream provided to an image processor, wherein the media stream comprises a plurality of image presentation types;

to cluster content portions from the identified presentation pages into a plurality of distinct presentation content sets, wherein each presentation content set contains content portions that have a same image presentation type;

to locate potential background pixels by comparing a first presentation page to a second presentation page, wherein the first presentation page and the second presentation page are located in a same set in the plurality of distinct presentation content sets;

to update the initial background model by incorporating the located potential background pixels into the initial background model to form an updated initial background model;

to determine whether the updated initial background model and the initial background model are the same;

to, responsive to determining that the updated initial background model and the initial background model are the same, output the updated initial background model, and distinguish between the foreground content and the background content in each image presentation page of the plurality of presentation pages using the updated initial background model; and

to repeat the steps of comparing the clustered content portions through the step of distinguishing between the foreground content and the background content for each remaining set of the distinct presentation content sets.

7. The system according to claim 6, wherein the image presentation comprises a computer-generated image presentation.

8. The system according to claim 7, wherein the computer-generated image presentation comprises one of a digital slide presentation, a Web page presentation and a video presentation.

9. The system according to claim 6, wherein the processor further executes the set of instructions to, responsive to determining that the updated initial background model and the initial background model are not the same, identify further presentation pages of the set of presentation pages whose foreground content are distinct and whose background content are the same;

compare the identified further presentation pages to locate further potential background pixels;

further update the updated initial background model by incorporating the further located potential background pixels into the updated initial background model to form a further updated initial background model;

output the further updated initial background model; and

distinguish between the foreground content and the background content in each image presentation page of the plurality of presentation pages using the further updated initial background model.

10. The system according to claim 9, wherein the processor further executes the set of instructions to repeat the identifying further presentation pages, the comparing the identified further presentation pages and the further updating the updated initial background model until an updated background model reaches a steady state.

11. A non-transitory computer readable medium storing computer program comprising:

computer usable program code for providing an initial background model on a data processing system that is independent of any particular image presentation;

computer usable program code for identifying presentation pages of a set of presentation pages of the image presentation, wherein the foreground content of each identified presentation page of the set of presentation pages are distinct and the background content of each identified presentation page of the set of presentation pages are the same, and wherein the presentation pages are part of a media stream provided to an image processor, wherein the media stream comprises a plurality of image presentation types;

computer usable program code for clustering content portions from the identified presentation pages into a plurality of distinct presentation content sets, wherein each presentation content set contains content portions that have a same image presentation type;

computer usable program code locating potential background pixels by comparing a first presentation page to a second presentation page, wherein the first presentation page and the second presentation page are located in a same set in the plurality of distinct presentation content sets;

computer usable program code for updating the initial background model by incorporating the located potential background pixels into the initial background model to form an updated initial background model;

computer usable program code for determining whether the updated initial background model and the initial background model are the same;

responsive to determining that the updated initial background model and the initial background model are the same;

computer usable program code for outputting the updated initial background model;

computer usable program code for distinguishing between the foreground content and the background content in each image presentation page of the plurality of presentation pages using the updated initial background model; and

computer usable program code for repeating the steps of comparing the clustered content portions through the step of distinguishing between the foreground content and the background content for each remaining set of the distinct presentation content sets.

12. The non-transitory computer readable medium according to claim 11, and further comprising:

responsive to determining that the updated initial background model and the initial background model are not the same;

computer usable program code for identifying further presentation pages of the set of presentation pages whose foreground content are distinct and whose background content are the same;

computer usable program code for comparing the identified further presentation pages to locate further potential background pixels;

computer usable program code for further updating the updated initial background model by incorporating the further located potential background pixels into the updated initial background model to form a further updated initial background model;

computer usable program code for outputting the further updated initial background model; and

computer usable program code for distinguishing between the foreground content and the background content in each presentation page of the plurality of presentation pages using the further updated initial background model.

13. The non-transitory computer readable medium according to claim 12, and further comprising:

computer usable program code for repeating the steps of identifying further presentation pages, comparing the identified further presentation pages and further updating the updated initial background model until an updated background model reaches a steady state.