Automatic detection and correction of red-eye effect

In this paper, we address the problem of automatic red-eye detection and correction. A four step, novel approach is presented where the first step consists of face detection in the input image. In the second step the face image is converted to gray-scale image in such a way that facilitates the detection of iris pair and its corresponding radius values. In third and fourth steps iris pair is located, and the central points of irises and their corresponding radius values are utilized to desaturate the redness inside iris regions. The desaturation scheme is adaptive to the severity of redness. Images with different severity level and size of redness are used to test the robustness of the proposed scheme. We also compare our correction schemes with two existing automatic methods.