Methods and mechanisms for probabilistic color correction

1. A method comprising:

obtaining image data representing a capture by an imaging device of an image of a target under a target illumination, the image data including image pixel values corresponding to respective zones of the image of the target;

determining an illumination color of the image, including calculating an average of the image pixel values;

calculating with a processor a probability that the target illumination is a reference illumination based on the determined illumination color and data describing a previously calculated probability distribution for the reference illumination; and

storing in a memory device an indication of the calculated probability that the target illumination is the reference illumination.

2. The method of claim 1 , further comprising:

obtaining data representing a set of captures by an imaging device, wherein each capture in the set of captures is of an image of a reference target in a set of reference targets, wherein each image is of a reference target under the reference illumination, the data representing the set of captures including a white-point for each capture in the set of captures; and

generating the data describing the previously calculated probability distribution for the reference illumination, including calculating a probability distribution of the white-points for each capture in the set of captures.

3. The method of claim 2 , further comprising:

obtaining data representing a second set of captures by an imaging device, wherein each capture in the second set of captures is of an image of a reference target in the set of reference targets, wherein each image is of a reference target under a second reference illumination, the data representing the second set of captures including a white-point for each capture in the set of captures; and

wherein the data describing the previously calculated probability distribution for the reference illumination includes data describing a probability distribution of the white-points for each capture in the second set of captures under the second reference illumination.

4. The method of claim 3 , wherein generating the data describing the previously calculated probability distribution for the reference illumination includes generating data describing a mixture model of the first reference illumination and the second reference illumination.

5. The method of claim 2 , wherein calculating the probability distribution of the white-points for each capture in the set of captures under the reference illumination includes

determining an illumination color for each capture in the set of captures, and

calculating a transformation of the determined illumination colors of the set of captures.

6. The method of claim 5 , wherein calculating a transformation of the determined illumination colors of the set of captures includes calculating a principal component transformation of the determined illumination colors of the set of captures.

7. The method of claim 1 , further comprising determining an estimate of the target illumination based on the computed probability that the target illumination is the reference illumination.

8. The method of claim 1 , further comprising:

determining a color correction for the image data based on the calculated probability that the target illumination is the reference illumination.

9. The method of claim 8 , wherein the reference illumination is one of a plurality of reference illuminations each associated with a respective value for a color correction parameter, and wherein determining the color correction for the image data includes determining a value for the color correction parameter based on each associated color correction parameter value of the plurality of reference illuminations.

10. The method of claim 8 , wherein determining the color correction for the image data includes determining a value of a color correction parameter for the image data based on a value of the color correction parameter for image data representing another capture of an image.

11. The method of claim 1 , wherein calculating the probability that the target illumination is the reference illumination is further based on data describing an exposure characteristic of the imaging device.

12. The method of claim 1 , further comprising:

detecting that the image data indicates a spatial image flicker; and

wherein calculating the probability that the target illumination is the reference illumination includes calculating based on the detected spatial image flicker.

13. A computer-readable storage medium having stored thereon instructions which when executed by one or more processors cause the one or more processors to perform a method comprising:

obtaining image data representing a capture by an imaging device of an image of a target under a target illumination, the image data including image pixel values corresponding to respective zones of the image of the target;

determining an illumination color of the image, including calculating an average of the image pixel values;

calculating a probability that the target illumination is a reference illumination based on the determined illumination color and data describing a previously calculated probability distribution for the reference illumination; and

storing an indication of the calculated probability that the target illumination is the reference illumination.

14. The computer-readable storage medium of claim 13 , the method further comprising

obtaining data representing a set of captures by an imaging device, wherein each capture in the set of captures is of an image of a reference target in a set of reference targets, wherein each image is of a reference target under the reference illumination, the data representing the set of captures including a white-point for each capture in the set of captures; and

generating the data describing the probability distribution for the reference illumination, including calculating a probability distribution of the white-points for each capture in the set of captures.

15. The computer-readable storage medium of claim 14 , the method further comprising:

obtaining data representing a second set of captures by an imaging device, wherein each capture in the second set of captures is of an image of a reference target in the set of reference targets, wherein each image is of a reference target under a second reference illumination, the data representing the second set of captures including a white-point for each capture in the set of captures; and

wherein the data describing the previously calculated probability distribution for the reference illumination includes data describing a probability distribution of the white-points for each capture in the second set of captures under the second reference illumination.

16. The computer-readable storage medium of claim 14 , wherein calculating the probability distribution of the white-points for each capture in the set of captures under the reference illumination includes calculating a principal component transformation of the data representing the set of captures.

17. The computer-readable storage medium of claim 13 , wherein the data describing the previously calculated probability distribution for the reference illumination includes a mixture model further describing a probability distribution for a second reference illumination, the method further comprising:

computing a probability that the target illumination is the second reference illumination based on the determined illumination color for the image and the mixture model; and

determining an estimate of the target illumination based on the computed probability that the target illumination is the reference illumination and the computed probability that the target illumination is the second reference illumination.

18. The computer-readable storage medium of claim 13 , the method further comprising:

determining a color correction for the image data based on the calculated probability that the target illumination is the reference illumination.

19. An apparatus comprising:

an image data receiver to obtain image data representing a capture by an imaging device of an image of a target under a target illumination, the image data including image pixel values corresponding to respective zones of the image of the target;

an illumination color determining means coupled to the image data receiver for determining an illumination color of the image, including means for calculating an average of the image pixel values;

a probability determining means coupled to the illumination color determining means for obtaining an indication of the determined illumination color of the image data and further for accessing data describing a previously calculated probability distribution for the reference illumination, the probability determining means further for calculating a probability that the target illumination is a reference illumination based on the determined illumination color of the image data and the data describing the previously calculated probability distribution for the reference illumination; and

a memory to store an indication of the calculated probability that the target illumination is a reference illumination.

20. The apparatus of claim 19 , further comprising:

means for obtaining data representing a set of captures by an imaging device, wherein each capture in the set of captures is of an image of a reference target in a set of reference targets, wherein each image is of a reference target under the reference illumination, the data representing the set of captures including a white-point for each capture in the set of captures, and

means for generating the data describing the probability distribution for the reference illumination, including means for calculating a probability distribution of the white-points for each capture in the set of captures.

21. The apparatus of claim 20 , further comprising:

means for obtaining data representing a second set of captures by an imaging device, wherein each capture in the second set of captures is of an image of a reference target in the set of reference targets, wherein each image is of a reference target under a second reference illumination, the data representing the second set of captures including a white-point for each capture in the set of captures, and

wherein the data describing the probability distribution for the reference illumination includes data describing a probability distribution of the white-points for each capture in the second set of captures under the second reference illumination.

22. The apparatus of claim 19 , further comprising:

a color correction module to determine a color correction for the image data based on the calculated probability that the target illumination is the reference illumination.

23. The apparatus of claim 22 , wherein the reference illumination is one of a plurality of reference illuminations each associated with a respective value for a color correction parameter, and wherein the color correction module to determine the color correction for the image data includes the color correction module to determine a value for the color correction parameter based on each associated color correction parameter value of the plurality of reference illuminations.

24. The apparatus of claim 22 , wherein the color correction module to determine the color correction for the image data includes the color correction module to determine a value of a color correction parameter for the image data based on a value of the color correction parameter for image data representing another capture of an image.

25. The apparatus of claim 19 , wherein the probability determining means for calculating the probability that the target illumination is the reference illumination further based on data describing an exposure characteristic of the imaging device.

26. The apparatus of claim 19 , further comprising:

means for detecting that the image data indicates a spatial image flicker; and

wherein the probability determining means for calculating the probability that the target illumination is the reference illumination further based on the detected spatial image flicker.

27. A system comprising:

an imaging device to generate image data representing a capture of an image of a target under a target illumination, the image data including image pixel values corresponding to respective zones of the image of the target;

an image data receiver coupled to the imaging device to obtain the image data;

an illumination color determining means coupled to the image data receiver for determining an illumination color of the image, including calculating an average of the image pixel values;

a probability determining means coupled to the white point determining means for obtaining an indication of the determined illumination color of the image and data describing a previously calculated probability distribution for the reference illumination, the probability determining means further for calculating a probability that the target illumination is a reference illumination based on the determined illumination color of the image and the data describing the previously calculated probability distribution for the reference illumination; and

a memory to store an indication of the calculated probability that the target illumination is a reference illumination.

28. The system of claim 27 , further comprising:

means for obtaining data representing a set of captures by an imaging device, wherein each capture in the set of captures is of an image of a reference target in a set of reference targets, wherein each image is of a reference target under the reference illumination, the data representing the set of captures including a white-point for each capture in the set of captures, and

means for generating the data describing the probability distribution for the reference illumination, including means for calculating a probability distribution of the white-points for each capture in the set of captures.

29. The system of claim 28 , further comprising:

means for obtaining data representing a second set of captures by an imaging device, wherein each capture in the second set of captures is of an image of a reference target in the set of reference targets, wherein each image is of a reference target under a second reference illumination, the data representing the second set of captures including a white-point for each capture in the set of captures, and

wherein the data describing the probability distribution for the reference illumination includes data describing a probability distribution of the white-points for each capture in the second set of captures under the second reference illumination.