Detecting and correcting red-eye in a digital image

A method of processing a digital image containing pixels with original color values, comprising:

• detecting a candidate red-eye pixel area in the digital image;
• generating a final pixel mask identifying red-eye pixels and non-red-eye pixels in the candidate red-eye pixel area;
• and
• desaturating pixels in the candidate red-eye pixel area identified as red-eye pixels in the final pixel mask from their original color values by respective scaling factors that vary depending on luminance of the identified red-eye pixels in the candidate red-eye pixel area and proximity of the identified red-eve pixels in the candidate red-eve pixel area to boundaries between red-eye pixels and non-red-eye pixels identified in the final pixel mask.

A system for processing a digital image containing pixels with original color values, comprising a global screening module operable to:

• compute measures of pixel redness in the digital image;
• identify a preliminary set of candidate red-eye pixel areas based on the measures of pixel redness;
• filter from the preliminary set each candidate red-eye pixel area having a computed redness contrast relative to at least one respective neighboring pixel area less than a prescribed redness contrast threshold;
• and
• filter from the preliminary set each candidate red-eye pixel area located in an area of the digital image having a computed grayscale contrast relative to at least one respective neighboring pixel area less than a prescribed grayscale contrast threshold.

A system for processing a digital image containing pixels with original color values, comprising:

• a red-eye detection module operable to detect a candidate red-eye pixel area in the digital image;
• a red-eye segmentation module operable to generate a final pixel mask identifying red-eye pixels and non-red-eye pixels in the candidate red-eye pixel area;
• and
• a red-eye correction module operable to desaturate pixels in the candidate red-eye pixel area identified as red-eye pixels in the final pixel mask from their original color values by respective scaling factors that vary depending on luminance of the identified red-eye pixels in the candidate red-eye pixel area and proximity of the identified red-eye pixels in the candidate red-eye pixel area to boundaries between red-eye pixels and non-red-eye pixels identified in the final pixel mask.