Spectral normalization using illuminant exposure estimation

1. A method for color correcting comprising:

capturing an image;

determining at least one scene parameter, which includes a world estimation for the image;

determining a best guess illuminant according to the at least one scene parameter;

determining a difference between the best guess illuminant and the world estimation;

when the difference exceeds a tolerance, applying a white balancing coefficient corresponding to the world estimation to a color correction matrix;

when the difference is within the tolerance, applying a further white balancing coefficient corresponding to the best guess illuminant to the color correction matrix; and

applying the color correction matrix to the captured image.

2. A method, as defined in claim 1 , wherein the at least one scene parameter further includes a scene parameter selected from a group that includes flicker depth, lux level, and auxiliary spectral measurements.

3. A method, as defined in claim 1 , wherein the world estimation is selected from a group that includes beigeworld, grayworld, whiteworld, and Max Pixel.

4. A method, as defined in claim 1 , wherein determining the best guess illuminant includes:

comparing the current scene data to stored data; and

selecting the best guess illuminant based on the comparison.

5. A method, as defined in claim 4 , wherein the stored data is selected from a group comprising variance threshold, previous threshold, and maximum tolerance.

6. The method of claim 1 , further comprising applying a default color matrix based on lux when the difference exceeds the tolerance.

7. A method for color correcting, comprising:

capturing an image;

determining a world estimation for the image;

determining a best guess illuminant according to the world estimation;

when the world estimation is within a variant circle for the best guess illuminant, applying a white balancing coefficient corresponding to the best guess illuminant to a color correction matrix;

when the world estimation is outside the variant circle for the best guess illuminant, applying a further white balancing coefficient corresponding to the world estimation to the color correction matrix; and

applying the color correction matrix to the captured image.

8. The method of claim 7 , further comprising applying a color matrix corresponding to the world estimation when the world estimation is within the variant circle for the best guess illuminant.

9. The method of claim 7 , further comprising applying a color matrix corresponding to the best guess illuminant when the world estimation is outside the variant circle for the best guess illuminant.

10. The method of claim 7 , wherein the best guess illuminant is determined by selecting the best guess illuminant from among a plurality of predetermined illuminants.

11. The method of claim 10 , wherein the best guess illuminant is selected by:

comparing predetermined illuminant gains corresponding to each of the plurality of predetermined illuminants with predetermined world gains corresponding to a selected world, and

selecting the predetermined illuminant that has corresponding predetermined illuminant gains that are closest to the predetermined world gains as the best guess illuminant.

12. The method of claim 10 , wherein the plurality of illuminants include dark, normal indoor, transition, indirect day and direct day.

13. An apparatus comprising:

a statistics block configured to monitor color corrected image signals provided by an image sensor and to provide scene parameter data from the image signals to the microcontroller;

memory including default scene data configured to provide the default scene data to the microcontroller;

a microcontroller configured to receive the scene parameter data and the default scene data from the statistics block and the memory, respectively, and to generate a gain signal and a color correction matrix based on the scene parameter data and the default scene data;

a sensor configured to receive the color correction matrix and, responsive to receiving the color correction matrix, to adjusting color channels in the sensor according to the color correction matrix.

14. An apparatus, as defined in claim 13 , wherein the scene parameters are selected from a group including current input gains, scene lux, flicker depth, and auxiliary spectral measurements.

15. A method for color correcting comprising:

capturing a plurality of images;

determining a current world estimation for a current image of the plurality of images;

selecting a current best guess illuminant for the current image from among a plurality of predetermined illuminants;

comparing the current world estimation with a threshold value corresponding to the current best guess illuminant to compute a current difference value;

comparing the current difference value with a maximum tolerance value; and

if the current difference value is greater than the maximum tolerance value, selecting the current best guess illuminant as a best guess illuminant for the current image.

16. The method of claim 15 , further comprising:

determining a previous world estimation for a previous image of the plurality of images;

selecting a previous best guess illuminant for the previous image from among the plurality of predetermined illuminants;

comparing the previous world estimation with the threshold value corresponding to the previous best guess illuminant to compute a previous difference value; and

comparing the previous difference value with the maximum tolerance value to determine a color matrix to apply to the captured image.

17. The method of claim 16 , further comprising:

if the current difference value is less than or equal to the maximum tolerance value:

computing a ratio of the current difference value to the previous difference value,

comparing the ratio with the threshold value corresponding to the previous best guess illuminant,

if the ratio is greater than the threshold value corresponding to the previous best guess illuminant, selecting the current best guess illuminant as the best guess illuminant for the current image, and

if the ratio is less than or equal to the threshold value corresponding to the previous best guess illuminant, selecting the previous best guess illuminant as the best guess illuminant for the current image.

18. A method for color correcting comprising:

capturing a plurality of images;

determining a current world estimation for a current image of the plurality of images and a previous world estimation for a previous consecutive image of the plurality of images;

for the current image and the previous consecutive image, comparing predetermined illuminant gains corresponding to each of a plurality of different illuminants with predetermined world gains corresponding to the current world estimation and the previous world estimation to obtain a current difference value and previous difference value, respectively;

selecting the predetermined illuminant with the closest gains as a previous best guess illuminant for the previous consecutive image; and

selecting the predetermined illuminant with the closest gains as a current best guess illuminant for the current image if the current difference value is less than the previous difference value by a predetermined amount.

19. The method of claim 18 , further comprising selecting the previous best guess illuminant as a current best guess illuminant for the current image if the current difference value is not less than the previous difference value by a predetermined amount.