Area based interpolation for image scaling

• US 5,796,879 A
• Filed: 05/23/1995
• Issued: 08/18/1998
• Est. Priority Date: 05/23/1995
• Status: Expired due to Term

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Company (HP Inc.)
• Inventors: Wong, Ping Wah, Herley, Cormac
• Primary Examiner(s): Coles, Edward L.
• Assistant Examiner(s): Williams, Kimberly A.
• Application Number:
  US08/447,863
• Time in Patent Office: 1,183 Days
• Field of Search: 358/451, 358/528, 358/537, 358/428, 358/452, 395/102, 382/279, 382/298, 382/299, 382/300
• US Class Current: 382/300
• 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 technique, referred to as area-based interpolation, performs image interpolation. The system determines a curve by the pixel value at a location by taking the integral of a curve over a small area, where the size of the area is determined by the sampling size of a sampling cell. When the image is resampled with respect to a sampling cell that has a finer spacing, the system integrates the polynomial using a finer integration area. In accordance with the invention, the relation between the reintegrated, resampled high resolution image and the low resolution image is a function of an up-sampler, followed by a linear filter. The coefficients of the filter are independent of the data, but are dependent on the family of curves used. If the system models by, for example, a third degree or fourth degree polynomial everywhere in the image, then that model determines the number of coefficients that are in the filter. Once the up-sampling factor and degree of the polynomial are determined, the filter coefficients are fixed, and can be calculated by solving a set of linear equations. It is not necessary to determine these values on an image-by-image basis. The system calculates the filter once and the consistency constraint is then satisfied for all image scaling.

Claims

1. A method for image scaling by area based interpolation, comprising the steps of:

representing a section of an image as a function of a continuous variable;

taking the integral of a curve for each area of said section, where the size of each area is determined by the sampling size of a first sampling cell;

sampling said image at a first, low resolution with respect to said first sampling cell, wherein a pixel value is determined for each area of said section;

resampling said image at a second, higher resolution with respect to a second sampling cell that has a finer spacing than that of said first sampling cell; and

taking the integral of a curve for each area of said section, where the size of each area is determined by the sampling size of said second sampling cell.

2. The method of claim 1, wherein said sampling cell is a CCD.

3. The method of claim 1, wherein the relation between said resampled higher resolution image and said sampled image is a function of an up-sampler, followed by a linear filter.

4. The method of claim 1, wherein said curves are determined by using polynomials.

5. The method of claim 4, wherein the coefficients of said filter are independent of the data, but are dependent on the degree of polynomial approximation used.

6. The method of claim 4, further comprising the step of:

recalculating said filter once, wherein the consistency constraint is satisfied for all image scaling.

7. A method for image scaling by area based interpolation, comprising the steps of:

receiving low resolution image data having light values for pixels of a sampling cell;

using the light values for each pixel of the sampling cell to determine an approximation curve that closely approximates a true curve that matches the light intensity across the surface of the sampling cell, such that the equation;

##EQU11## is satisfied;

where I(u,v) represents a waveform of an image;

x_mn is an output of a low resolution acquisition device for waveform I(u,v);

L_mn is the area occupied by the sensor at the region R_m,n ; and

outputting high resolution image data based on the approximation curve.

8. The method of claim 7, wherein said sampling cell is a CCD.

9. The method of claim 7, wherein the low resolution image is infinitely scalable.

10. The method of claim 7, wherein said curve is determined by using polynomials.

11. An apparatus for scaling an image by area based interpolation, the apparatus comprising:

means for representing a section of an image as a function of a continuous variable;

a first integrator for taking the integral of a curve for each area of said section, where the size of each area is determined by the sampling size of a first sampling cell;

means for sampling said image at a first, low resolution with respect to said first sampling cell, wherein a pixel value is determined for each area of said section;

means for resampling said image at a second, higher resolution with respect to a second sampling cell that has a finer spacing than that of said first sampling cell; and

a second integrator for taking the integral of a curve for each area of said section, where the size of each area is determined by the sampling size of said second sampling cell.

12. The apparatus of claim 11, wherein the relation between said resampled higher resolution image and said sampled image is a function of an up-sampler, followed by a linear filter.

13. The apparatus of claim 12, wherein said curve is determined by using polynomials.

14. The apparatus of claim 13, wherein the coefficients of said linear filter are independent of the data, but are dependent on the degree of polynomial approximation used.

15. The apparatus of claim 13, further comprising:

means for recalculating said filter once, wherein the consistency constraint is satisfied for all image scaling.

16. The apparatus of claim 11, wherein said sampling cell is a CCD.

17. An apparatus for image scaling by area based interpolation, comprising:

means for receiving low resolution image data having light values for pixels of a sampling cell;

means for determining an approximation curve from the light values for each pixel of the sampling cell from the low resolution image, wherein the approximation curve closely approximates a true curve that matches the light intensity across the surface of the sampling cell, such that the equation;

##EQU12## is satisfied;

where I(u,v) represents a waveform of an image;

x_mn is an output of a low resolution acquisition device for waveform I(u,v);

L_mn is the area occupied by the sensor at the region R_m,n ; and

means for outputting high resolution image data based on the approximation curve.

18. The apparatus of claim 17, wherein said curve is determined by using polynomials.

19. The apparatus of claim 17, wherein said sampling cell is a CCD.

20. The apparatus of claim 17, wherein the low resolution image is infinitely scalable.

21. An apparatus for image scaling by area based interpolation, comprising:

an input to receive first resolution image data having light values for pixels of a sampling cell;

a curve approximator adapted to determine an approximation curve from the light values for each pixel of the sampling cell from the first resolution image, wherein the approximation curve closely approximates a true curve that matches the light intensity across the surface of the sampling cell;

an integrator adapted to integrate the approximation curve in a small region to determine the area underneath the approximation curve in the smaller region, using the integrated value to scale the first resolution image to a second resolution image; and

an output adapted to output the second resolution image data.

22. The apparatus of claim 21, wherein the curve is determined using polynomials.

23. The apparatus of claim 21, wherein the sampling cell is a CCD.

24. The apparatus of claim 21, wherein the first resolution image is infinitely scalable.

25. The apparatus of claim 21, wherein the first resolution image is of a lower resolution than the second resolution image.