System and computer-readable medium for automatic white balancing

1. An automatic white balancing system, the system comprising:

a cast detector configured to receive and classify an input image based at least in part on a selection of constituent pixels of the input image, wherein the cast detector is further configured to select the constituent pixels based on a computed reliability of each of the constituent pixels to convey at least one characteristic of a capture illuminant, the cast detector including:

a pixel processing unit configured to select the constituent pixels from the input image, wherein the pixel processing unit is configured to:

select the constituent pixels having a green value (G) lying in a predetermined range;

define a weighted function of the green value, the weighted function being so defined to assign a higher weight to a middle band green value;

for each selected constituent pixel, compute a range of red and blue values based on the green value of the constituent pixel;

for each non-saturated constituent pixel, compute a first metric and a second metric based on a comparison of the red and blue values of the non-saturated constituent pixel; and

compute a first pixel count and a second pixel count corresponding to the first and second metrics respectively; and

an image-classifying unit configured to classify the input image based on red, green, and blue values of the constituent pixels selected by the pixel processing unit; and

a cast balancer configured to perform white balancing of the input image based on the classification of the input image by the cast detector, wherein the cast balancer is further configured to identify the capture illuminant based on at least one characteristic represented by each of the selected constituent pixels.

2. The system of claim 1 , wherein the pixel processing unit is further configured to:

compute, for each of the selected constituent pixels having corresponding red and blue values in the determined range of values, a difference between the corresponding red value and the blue value;

determine whether the computed difference is less than a predetermined first threshold;

on a positive determination, compute a third metric and a fourth metric based on a comparison of the red and the blue values of the selected constituent pixel; and

compute a third pixel count and a fourth pixel count corresponding to the third and forth metrics respectively.

3. The system of claim 2 , wherein the image classifying unit is configured to:

compare the first, the second, the third, and the fourth pixel counts with a predetermined second threshold; and

based on the comparison, select one of the first, second, third, and forth metrics, the selected metric corresponding to the pixel count characterized by a maximum count value.

4. The system of claim 3 , wherein the image classifying unit is configured to:

compute gain coefficients based on the selected metric, wherein gain coefficients represent ratios of color values of the constituent pixels in accordance with the selected metric;

determine whether the gain coefficients are less than a predetermined third threshold; and

classify the input image as one of color biased and white balanced based on the determination.

5. The system of claim 4 , wherein the cast balancer comprises an illuminant classifying unit configured to:

for the input image classified as a colorbiased image, define a set of reference illuminant parameter based on pre-stored database images and a set of reference light sources;

calculate a set of representative illuminant parameters of the input image, wherein the set of representative illuminant parameters correspond to the at least one characteristic of a capture illuminant; and

identify a type of the capture illuminant based on a comparison of the set of reference illuminant parameters with the set of representative illuminant parameters.

6. The system of claim 5 , wherein the cast balancer further comprises a balancing unit configured to perform white balancing of the input image based on the identification.

7. The system of claim 1 , wherein the predetermined range corresponds to 40-90% of total dynamic range of green values of the constituent pixels of the input image.