System and method of image zooming with intensity preservation

• US 20090041387A1
• Filed: 01/17/2008
• Published: 02/12/2009
• Est. Priority Date: 08/06/2007
• Status: Active Grant

• Current Assignee: Digital Domain 3.0, Inc.
• Original Assignee: Digital Domain, Inc.
• Inventors: Roble, Doug, Bin Zafar, Nafees
• Granted Patent: US 8,031,975 B2
• Time in Patent Office: Days
• US Class Current: 382/299
• 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

Systems and methods image zooming with intensity preservation is disclosed. In one aspect, embodiments of the present disclosure include a method, which may be implemented on a system, of minimizing energy levels over a set of high-resolution image pixels of a high-resolution image via total variational filtering to obtain the high-resolution image from at least a portion of a low-resolution image; the low-resolution image having a first energy level, the high resolution image having a second energy level, and the first energy level to be substantially similar in value as the second energy level. One embodiment can include, determining a pixel value of a particular high-resolution image pixel based on the pixel values of a plurality of pixels of the at least a portion of the low-resolution image and the pixel values of the high-resolution image pixels not including the particular high-resolution image pixel.

Claims

1. A method of enlarging an image, the method comprising:

minimizing energy levels of a set of high-resolution image pixels of a high-resolution image via total variational filtering the set of high-resolution image pixels, the high-resolution image to be an enlarged version of at least a portion of a low-resolution image, the at least a portion of the low-resolution image to have an area substantially equal to that of the high-resolution image,

wherein the low-resolution image corresponds to a first energy level and the high resolution image corresponds to a second energy level, the first energy level to have substantially a same value as the second energy level.

2. The method of claim 1, further comprising determining a first magnitude variation of a high-resolution image pixel and a first neighboring pixel of the high-resolution image pixel.

3. The method of claim 2, further comprising determining a second magnitude variation of the high-resolution image pixel and a second neighboring pixel of the high-resolution image pixel.

4. The method of claim 3, further comprising determining a plurality of magnitude variations between the high-resolution image pixel and a plurality of neighboring pixels of the high-resolution image pixel.

5. The method of claim 1, wherein the minimizing energy comprises, filtering the high-resolution image pixels with a filter associated with the plurality of magnitude variations.

6. The method of claim 5, further comprising determining a pixel value of a particular high-resolution image pixel based on the pixel values of a plurality of pixels of the at least a portion of the low-resolution image and the pixel values of the high-resolution image pixels not including the particular high-resolution image pixel.

7. The method of claim 6, further comprising identifying the pixel values of a set of high-resolution image pixels, such that, the average value of the pixel values of the set of high-resolution pixels is substantially equal to the average of pixel values of the at least a portion of the low-resolution image.

8. The method of claim 2, wherein the magnitude variation is computed via the ordinary Euclidean gradient.

9. A machine readable medium embodying instructions, the instructions causing a machine to perform a method of image zooming, the method comprising:

minimizing energy levels of a set of high-resolution image pixels of a high-resolution image via total variational filtering the set of high-resolution image pixels, the high-resolution image to be an enlarged version of at least a portion of a low-resolution image, the at least a portion of the low-resolution image to have an area substantially equal to that of the high-resolution image,

wherein the low-resolution image corresponds to a first energy level and the high resolution image corresponds to a second energy level, the first energy level to have substantially a same value as the second energy level.

10. The machine-readable medium of claim 9, further comprising determining a plurality of magnitude variations between a high-resolution image pixel of the high-resolution image and a plurality of neighboring pixels of the high-resolution image pixel.

11. The machine-readable medium of claim 9, wherein the minimizing energy comprises, filtering the high-resolution image pixels with a filter associated with the plurality of magnitude variations.

12. The machine-readable medium of claim 11, further comprising determining a pixel value of a particular high-resolution image pixel based on the pixel values of a plurality of pixels of the at least a portion of the low-resolution image and the pixel values of the high-resolution image pixels not including the particular high-resolution image pixel.

13. The machine-readable medium of claim 12, further comprising identifying the pixel values of a set of high-resolution image pixels, such that, the average value of the pixel values of the set of high-resolution pixels is substantially equal to the pixel values of the at least a portion of the low-resolution image.

14. The method of claim 9, wherein the magnitude variation is computed via the ordinary Euclidean gradient.

15. A system of enlarging an image, the system comprising:

means for, minimizing energy levels of a set of high-resolution image pixels of a high-resolution image via total variational filtering the set of high-resolution image pixels, the high-resolution image to be an enlarged version of at least a portion of a low-resolution image, the at least a portion of the low-resolution image to have an area substantially equal to that of the high-resolution image,

wherein the low-resolution image corresponds to a first energy level and the high resolution image corresponds to a second energy level, the first energy level to have substantially a same value as the second energy level.

16. The system of claim 15, further comprising determining a plurality of magnitude variations between a high-resolution image pixel of the high-resolution image and a plurality of neighboring pixels of the high-resolution image pixel.

17. The system of claim 15, wherein the minimizing energy comprises, filtering the high-resolution image pixels with a filter associated with the plurality of magnitude variations.

18. The system of claim 17, further comprising determining a pixel value of a particular high-resolution image pixel based on the pixel values of a plurality of pixels of the at least a portion of the low-resolution image and the pixel values of the high-resolution image pixels not including the particular high-resolution image pixel.

19. The system of claim 18, further comprising identifying the pixel values of a set of high-resolution image pixels, such that, the average value of the pixel values of the set of high-resolution pixels is substantially equal to the pixel values of the at least a portion of the low-resolution image.