Title :
Eye detection using hybrid rule based approach and contour mapping
Author :
Samad, S.A. ; Hussain, A. ; Teoh, A.
Author_Institution :
Dept. of Electr., Electron. & Syst. Eng., Universiti Kebangsaan Malaysia, Bangi Selangor, Malaysia
Abstract :
This paper describes an eye detection module that can be used as a part of an automatic face recognition system. It plays an important role in producing normalized face images for the purpose of template matching. An algorithm that employs a rule-based 8×8 mosaic window is combined with the contour map approach to detect the eye pair. Experiments indicate that the algorithm is able to detect the left eye and right eye with high accuracy and can be used to detect the eye positions of subjects wearing spectacles, which is normally difficult to achieve with a simple algorithm due to the light reflection property of the glasses
Keywords :
edge detection; eye; face recognition; feature extraction; image matching; knowledge based systems; automatic face recognition system; contour map approach; eye detection module; normalized face images; rule based mosaic window; template matching; Classification algorithms; Eyes; Face detection; Face recognition; Glass; Lighting; Optical reflection; Signal processing; Systems engineering and theory; Windows;
Conference_Titel :
Signal Processing and its Applications, Sixth International, Symposium on. 2001
Conference_Location :
Kuala Lumpur
Print_ISBN :
0-7803-6703-0
DOI :
10.1109/ISSPA.2001.950224