Estimating text color and segmentation of images

• US 6,473,522 B1
• Filed: 03/14/2000
• Issued: 10/29/2002
• Est. Priority Date: 03/14/2000
• Status: Expired due to Fees

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Lienhart, Rainer Wolfgang, Wernicke, Axel
• Primary Examiner(s): Tran, Phuoc
• Application Number:
  US09/525,204
• Time in Patent Office: 959 Days
• Field of Search: 382/168, 382/170, 382/171, 382/172, 382/164, 382/173, 382/176
• US Class Current: 382/168
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/30176 Document
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/194 involving foreground-background segmentation
• CPC Class Codes: G06T 7/90 Determination of colour characteristics
• CPC Class Codes: G06V 20/62 Text, e.g. of license plates, overlay texts or captions on TV images
• CPC Class Codes: G06V 2201/01 Solutions for problems related to non-uniform document background
• CPC Class Codes: G06V 30/10 Character recognition

Abstract

In some embodiments, the invention includes receiving a digital image including text and background. The method includes vector quantizing the digital image such that the digital image is divided into certain colors, and creating a text color histogram from a portion of the text and a first portion of the background. The method also includes creating at least one background color histogram from a second portion of the background, and creating a difference color histogram from a difference between the text color histogram and the at least one background color histogram, and wherein an estimated color of the text is derived from the difference color histogram. In other embodiments, the invention includes receiving a text object including bounding boxes of multiple frames of a video signal. The method further includes estimating a color of text of the bounding boxes and aligning blocks representing the bounding boxes through a best displacement search in which only pixels having a color within a threshold of an estimated color are considered. Some embodiments of the invention also include receiving digital images in text bounding boxes and in preparation for a segmentation process, adjusting sizes of the digital images to a fixed height.

Claims

1. A method comprising:

receiving a digital image including text and background;

vector quantizing the digital image such that the digital image is divided into certain colors;

creating a text color histogram from a portion of the text and a first portion of the background;

creating at least one background color histogram from a second portion of the background; and

creating a difference color histogram from a difference between the text color histogram and the at least one background color histogram, and wherein an estimated color of the text is derived from the difference color histogram.

2. The method of claim 1, wherein the portion of the text and first portion of the background includes a strip of rows through the text.

3. The method of claim 1, wherein the first portion of the background is background intermingled with the text.

4. The method of claim 1, wherein a second portion of the text includes a first part above the text and a second part below the text and the at least one background color histogram includes upper and lower background color histograms of the first and second parts, respectively.

5. The method of claim 1, wherein the estimated text color is the dominant color of the difference color histogram.

6. The method of claim 1, wherein the difference color histogram is created from the text color histogram minus the at least one background color histogram.

7. The method of claim 1, wherein an estimated background color is determined.

8. The method of claim 1, wherein the vector quantizer creates four colors.

9. The method of claim 1, wherein the digital image is a video signal having multiple successive digital images from which text objects are created.

10. The method of claim 9, wherein for at least some of the multiple digital images of the text objects, the method of claim 1 is performed to derive estimated colors of text for those digital images, and wherein estimated colors for the text objects are averages of the estimated colors of the at least some of the multiple digital images for the respective text objects.

11. The method of claim 10, wherein the average is a median estimated color value.

12. The method of claim 9, wherein in a segmentation process is performed to remove at least part of the background from the respective text objects, and wherein for this segmentation at least some blocks of the text objects are aligned through a best displacement search in which only pixels having a color that is within a threshold of the estimated text color are considered.

13. The method of claim 1, wherein in preparation for a segmentation process, the digital image is adjusted to a fixed height.

14. An apparatus comprising:

a machine readable medium having instructions thereon which when executed cause a processor to perform a method including;

receiving a digital image including text and background;

vector quantizing the digital image such that the digital image is divided into certain colors;

creating a text color histogram from a portion of the text and a first portion of the background;

creating at least one background color histogram from a second portion of the background; and

creating a difference color histogram from a difference between the text color histogram and the at least one background color histogram, and wherein an estimated color of the text is derived from the difference color histogram.

15. The apparatus of claim 14, wherein the portion of the text and first portion of the background includes a strip of rows through the text.

16. The apparatus of claim 14, wherein the first portion of the background is background intermingled with the text.

17. The apparatus of claim 14, wherein a second portion of the text includes a first part above the text and a second part below the text and the at least one background color histogram includes upper and lower background color histograms of the first and second parts, respectively.

18. The apparatus of claim 14, wherein the estimated text color is the dominant color of the difference color histogram.

19. The apparatus of claim 14, wherein an estimated background color is determined.

20. The apparatus of claim 14, wherein the digital image is a video signal having multiple successive digital images from which text objects are created.

21. The apparatus of claim 20, wherein for at least some of the multiple digital images of the text objects, the method of claim 14 is performed to derive estimated colors of text for those digital images, and wherein estimated colors for the text objects are averages of the estimated colors of the at least some of the multiple digital images for the respective text objects.

22. The apparatus of claim 20, wherein in a segmentation process is performed to remove at least part of the background from the respective text objects, and wherein for this segmentation at least some blocks of the text objects are aligned through a best displacement search in which only pixels having a color that is within a threshold of the estimated text color are considered.

23. A method comprising:

receiving a text object including bounding boxes of multiple frames of a video signal;

estimating a color of text of the bounding boxes;

aligning blocks representing the bounding boxes through a best displacement search in which only pixels having a color within a threshold of an estimated color are considered.

24. The method of claim 23, wherein representative bit maps are updated through the best displacement search.

25. The method of claim 23, wherein representative bit maps are updated through results of a minimum displacement equation.

26. An apparatus comprising:

a machine readable medium having instructions thereon which when executed cause a processor to perform a method including;

receiving a text object including bounding boxes of multiple frames of a video signal;

estimating a color of text of the bounding boxes;

aligning blocks representing the bounding boxes through a best displacement search in which only pixels having an estimated color are considered.

27. The apparatus of claim 26, wherein representative bit maps are updated through the best displacement search.

28. The apparatus of claim 26, wherein representative bitmaps are updated through results of a minimum displacement equation.

29. A method comprising:

receiving digital images in text bounding boxes;

in preparation for a segmentation process, adjusting sizes of the digital images to a fixed height, wherein if a particular one of the digital images has a height smaller than the fixed height the digital image is increased in height and if the particular one of the digital images has a height greater than the fixed height, the digital images is reduced in height.

30. The method of claim 29, wherein multiple ones of the digital video images originate from a larger image.

31. The method of claim 29, wherein the height is 100 pixels.

32. An apparatus comprising:

a machine readable medium having instructions thereon which when executed cause a processor to perform a method including;

receiving digital images in text bounding boxes;

in preparation for a segmentation process, adjusting sizes of the digital images to a fixed height, wherein if a particular one of the digital images has a height smaller than the fixed height the digital image is increased in height and if the particular one of the digital images has a height greater than the fixed height, the digital images is reduced in height.

33. The apparatus of claim 32, wherein multiple ones of the digital video images originate from a larger image.

34. The apparatus of claim 32, wherein the fixed height is 100 pixels.