Two-dimensional lens shading correction

1. A method of correcting a captured image for lens shading artifacts, comprising:

capturing a plurality of reference images with a lens;

averaging the plurality of captured reference images to determine a single reference image;

using the determined single reference image to determine, for the lens, a two-dimensional lens shading correction function L(x,y) which is a function of an image location (x,y) and which is to be applied to subsequent images captured by the lens in order to correct for lens shading artifacts;

applying a sampling technique to sample the determined function L(x,y) at selected points (x,y) based on a sampling grid; and

storing a two-dimensional table comprising data points (x,y) corresponding to the sampled function L(x,y) in memory.

2. The method of claim 1 , wherein the sampling grid comprises sampling points (x,y) that are fixed and equally spaced.

3. The method of claim 1 , wherein the sampling grid comprises sampling points (x,y) that are variable.

4. The method of claim 1 , wherein said averaging the plurality of captured reference images to determine a single reference image comprises smoothing the plurality of captured reference images to determine the single reference image.

5. A method of correcting a captured image for lens shading artifacts due to a lens used in generating the captured image, comprising:

reading image data corresponding to the captured image;

accessing a two-dimensional table to select stored sample points (x,y) for a two-dimensional lens shading correction function for the lens determined by capturing a plurality of reference images with the lens, and averaging the plurality of reference images into a single reference image wherein the stored sample points (x,y) are based on a sampling grid comprising sampling points (x,y); and

in respect of image data falling between the sample points (x,y) then generating missing values for the two dimensional lens shading correction function based on the selected sample points (x,y).

6. The method of claim 5 , wherein generating the missing values is based on interpolation.

7. The method of claim 6 , wherein said interpolation is bilinear or bicubic interpolation.

8. An image processor, comprising:

a memory configured to store lens shading correction parameters for storing a two-dimensional table comprising data points (x,y) corresponding to a sampled two-dimensional lens shading correction function L(x,y), wherein the function L(x,y) is determined by capturing a plurality of reference images with a lens, and averaging the plurality of reference images into a single reference image, wherein the stored sample points (x,y) are based on a sampling grid comprising sampling points (x,y);

data reconstruction logic configured to access the two dimensional table stored in the memory, to select stored sample points (x,y) for the two-dimensional lens shading correction function, and to reconstruct missing data points of the two-dimensional lens shading correction function based on the selected sample points (x,y); and

an image correction block configured to correct a captured image based on the reconstructed missing data.

9. The image processor of claim 8 , wherein the reconstruction of the missing data points is performed based on interpolation of the selected sampled points (x,y) from the two-dimensional table.

10. The image processor of claim 8 , wherein the averaging of the plurality of captured reference images into a single reference image comprises smoothing the plurality of captured reference images.

11. An image capture system, comprising:

camera optics comprising a lens;

an image sensor configured to capture images from the camera optics; and

an image processor configured to process an image captured by the image sensor using the camera optics, the image processor comprising:

a memory configured to store lens shading correction parameters for storing a two-dimensional table comprising data points (x,y) corresponding to a sampled two-dimensional lens shading correction function L(x,y), wherein the function L(x,y) is determined by capturing a plurality of reference images with the lens, and averaging the plurality of reference images into a single reference image, wherein the stored sample points (x,y) are based on a sampling grid comprising sampling points (x,y);

data reconstruction logic configured to access the two dimensional table stored in the memory, to select stored sample points (x,y) for the two-dimensional lens shading correction function, and to reconstruct missing data points of the two-dimensional lens shading correction function based on the selected sample points (x,y); and

an image correction block configured to correct a captured image based on the reconstructed missing data.

12. The image capture system of claim 11 , wherein the reconstruction of the missing data points is performed based on interpolation of the selected sampled points (x,y) from the two-dimensional table.

13. The image capture system of claim 12 , wherein said interpolation is bilinear or bicubic interpolation.