In this paper, we proposed an algorithm of redeye correction. For naturally redeye correction, we assumed positions of the redeye at an image which produced redeye, and we estimated rate of the redeye to apply the appropriate redeye correction suitabl...
In this paper, we proposed an algorithm of redeye correction. For naturally redeye correction, we assumed positions of the redeye at an image which produced redeye, and we estimated rate of the redeye to apply the appropriate redeye correction suitably. We extract and label pixels those are possible of generating redeye using red, skin and reflected light color. The each labeled group is decided by rates of length and width, dimension, density, the color of white of the eye and reflected light color of groups for the redeye group. We corrected positions of redeye using blurring effect, naturally. In the case of designing the proposed algorithm, we designed the redeye correction hardware using the minimum of memories for efficiency of hardware.