Using Red-Eye to Improve Face Detection in Low Quality Video Images

This paper presents a method to improve face detection by locating eyes in an image using infrared (IR) light. IR light produces Red-Eye effect making the pupil to shine more than in normal lighting conditions. The location of the eyes and the face contour are computed from the IR images using a collection of image processing techniques. The algorithm operates on gray images under a variety of skin tones, eye colors, angles, and illuminations. The developed method analyzes two consecutive frames where the first one is taken under regular illumination with no IR and the second one using IR in order to facilitate eye detection for subjects wearing glasses. Keywords— Red Eye effect, face detection, infrared lighting.

• Type: Conference
• Year: 2006
• Where: CCECE
• Authors: Richard Youmaran, Andy Adler
• Keywords: CCECE 2006, Electrical And Computer Engineering, Face Detection, IR Light, Normal Lighting Conditions