Method and device for compensating eye color defects

Abstract

Methods, apparatus and computer program products are provided. The method includes calculating a first difference image for the first eye region based on a difference in red pixel intensity and green pixel intensity for each pixel of a first set of pixels associated with the first eye region. The method also includes determining the first eye color defect region by: computing a neighborhood processed first difference image by processing the first difference image; based on red pixels associated with the neighborhood processed first difference image calculating a first center point of the neighborhood processed first difference image of the weighted centroid; and then calculating a first eye color defect region based on the first center point and red pixels associated with the first difference image.

Description

technical field [0001] Various embodiments relate generally to image processing and more particularly to methods, apparatus and computer program products for compensating for eye color defects. Background technique [0002] Eye color defect phenomena (such as redness) are often observed in images captured under low ambient lighting conditions. Typically under low ambient lighting conditions, the pupils of the subject's eyes are dilated, and when flash light hits the eye, internal reflections from the vascular membrane behind the retina of the subject's eye occur. In humans, such internal reflexes often cause an unnatural reddening of the pupil region of their eyes. In other types of animals, however, other colors such as golden yellow may develop. [0003] In addition to low ambient lighting levels, various other factors contribute to the appearance of redness. Such factors may include a small angle between the image capture device's flash and its lens, the candidate's pr...

AI Technical Summary

Problems solved by technology

However such devices may introduce a time delay between the first flash and the actual capture of the image, due to which the subject may stray from the posed position before the actual image is captured

[0005] Also may require user intervention when correcting red eye issues in images by processing clicked images to remove color imperfections from the eyes

However such user intervention in identifying and removing red-eye can prove cumbersome and inefficient

Also red eyes may be misinterpreted as non-red eye areas and non-red eye areas as red eyes due to user intervention

Additionally, such devices may often miss other known eye color defects such as golden eyes

Thus currently available techniques offer little utility in detecting eye color defects of variable saturation as well as brightness

Images