IMAGE SCALING FOR IMAGES INCLUDING LOW RESOLUTION TEXT

• US 20140347399A1
• Filed: 04/17/2014
• Published: 11/27/2014
• Est. Priority Date: 05/24/2013
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: MOFFATT, EDWIN P. J., STADDON, LUKE P.
• Granted Patent: US 9,536,281 B2
• Time in Patent Office: Days
• US Class Current: 345/667
• CPC Class Codes: G06T 3/4007 based on interpolation, e.g. bilinear interpolation image demosaicing G06T3/4015; edge-driven or edge-based scaling G06T3/403

Abstract

A method and system are provided for image scaling including low resolution text areas. The method may include: determining one or more areas of text in an image to be scaled; scaling the one or more areas of text using two or more scaling algorithms to a required resolution to acquire two or more scaled versions of the text area; selecting the clearest text in a text area from two or more scaled versions of the text area; scaling the overall image to the required resolution; and merging the scaled overall image with the selected scaled version of each text area.

Claims

1. A method for image scaling including low resolution text areas, the method comprising:

determining, by one or more processors, one or more areas of text in an image to be scaled;

scaling, by one or more processors, the one or more areas of text using two or more scaling algorithms to a required resolution to acquire two or more scaled versions of the text area;

selecting a clearest text in a text area from two or more scaled versions of the text area;

scaling, by one or more processors, the overall image to the required resolution; and

merging, by one or more processors, the scaled overall image with the selected scaled version of each text area.

2. The method as claimed in claim 1, wherein one of the scaling algorithms is a linear algorithm for scaling the original text.

3. The method as claimed in claim 1, further comprising:

selecting the clearest text by comparing, by one or more processors, the two or more scaled versions of the text area using a text-recognition algorithm to provide a metric of the clarity of the text in the text area.

4. The method as claimed in claim 1, wherein determining one or more text areas in an image to be scaled comprises:

applying, by one or more processors, text-recognition to identify a text area; and

drawing, by one or more processors, a bounding box around a text area.

5. The method as claimed in claim 1, wherein determining one or more text areas in an image to be scaled includes:

applying, by one or more processors, rules of known characteristics of text areas of an image to identify a text area.

6. The method as claimed in claim 1, wherein two or more different scaling algorithms are used for the group consisting of:

each text area and the overall image.

7. The method as claimed in claim 1, further comprising:

removing, by one or more processors, the one or more identified text area from the image; and

filling, by one or more processors, in the removed identified text areas with image matching the surrounding image.

8. The method as claimed in claim 7, further comprising:

marking, by one or more processors, the removed identified text areas a distinguishing color before filling with matching overall image.

9. The method as claimed in claim 7, wherein said removing and filling are carried out on the pre-scaled overall image.

10. The method as claimed in claim 1, wherein merging the scaled overall image with the selected one or more text areas includes:

replacing, by one or more processors, pixels in the overall image with any non-alpha pixels in the text areas.

11. The method as claimed in claim 1, wherein scaling the overall image to the required resolution includes scaling the original text areas with the overall image, and wherein merging the scaled overall image with the selected version of each text area includes overwriting the scaled text areas with the selected scaled version of each text area.

12. The method as claimed in claim 1, further comprising:

applying, by one or more processors, shaders to the scaled overall image prior to merging the scaled overall image with the selected versions of each text area.

13. The method as claimed in claim 1, further comprising:

preserving, by one or more processors, offsets of text areas in the image in order to merge the scaled text areas with the scaled overall image.

14. A system for image scaling including low resolution text areas, comprising:

a text area identifying hardware component for determining one or more areas of text in an image to be scaled;

a text area scaling hardware component for scaling the one or more areas of text using two or more scaling algorithms to a required resolution to acquire two or more scaled versions of the text area;

a text selection hardware component for selecting a clearest text in a text area from two or more scaled versions of the text area;

an image scaling hardware component for scaling the overall image to the required resolution; and

a merging hardware component for merging the scaled overall image with the selected scaled version of each text areas.

15. The system as claimed in claim 14, wherein one of the scaling algorithms is a linear algorithm for scaling the original text.

16. The system as claimed in claim 14, wherein the text area identifying component is for:

applying text-recognition to identify a text area; and

drawing a bounding box around a text area.

17. The system as claimed in claim 14, further comprising:

a text area removing hardware component for removing the one or more identified text area from the image; and

an image filling hardware component for filling in the removed identified text areas with image matching the surrounding image.

18. A computer program product for image scaling including low resolution text areas, the computer program product comprising a computer readable storage medium having program code embodied therewith, the program code readable and executable by a processor to perform a method comprising:

determining one or more areas of text in an image to be scaled;

scaling the one or more areas of text using two or more scaling algorithms to a required resolution to acquire two or more scaled versions of the text area;

selecting a clearest text in a text area from two or more scaled versions of the text area;

scaling the overall image to the required resolution; and

merging the scaled overall image with the selected scaled version of each text area.

19. The computer program product as claimed in claim 18, wherein one of the scaling algorithms is a linear algorithm for scaling the original text.

20. The computer program product as claimed in claim 18, wherein the method further comprises:

selecting the clearest text by comparing the two or more scaled versions of the text area using a text-recognition algorithm to provide a metric of the clarity of the text in the text area.