Red-eye is a common photographic artifact where the eyes of the subject appear red instead of their natural color. It is caused by the reflection of flash light usedin low-light photography bouncing back from human retina into the camera. This project presents state of the art automatic red-eye detection and correction algorithm based on computer vision and pattern recognition algorithms. The proposed algorithm consists of two major steps: red-eye detection and red-eye correction where detection is the most challenging phase. Face regions are extracted from input image using skin-based segmentation. Iris positions are then tracked using edge detection and circular Hough transform algorithms. Finally the detected iris regions are searched for red pixel clusters to confirm the existence of red-eyes and corrected accordingly. The system is able to detect and correct red-eye from images with complex background and multiple faces. The algorithm is also invariant of different face positions, orientations and light intensity. The proposed system makes major contribution in red-eye detection and removal research and presents a new solution with acceptable accuracy rate. An openly accessible image dataset of different resolutions with red-eye effect has been used to test our proposed algorithm. The average red-eye detection and correction accuracy rate achieved by the proposed method is above 90%.