Eye detection and pupil location algorithm based on Gabor filters and K-medoid algorithm

刘春生,常发亮,陈振学,李爽

DOI: https://doi.org/10.16136/j.joel.2013.08.026

2013-01-01

Abstract:Eye detection is an essential initial step in many face processing applications,such as face tracking,expression analysis and face recognition.In the process of eye detection,in order to overcome the factors of expression changes,illumination changes and glasses' shelters,the eye detection method based on Gabor filters and K-medoid algorithm is proposed.There are mainly three steps to detect eyes and locate pupils.Initially,the proposed method highlights the eyes' position with different scale Gabor filters,which is robust to the changes of eye,occlusion and illumination changes.It then combines Gabor filter method with the improved K-medoid algorithm to conduct cluster analysis to detect eyes' position.Since the position located by K-medoid algorithm is not accurate enough,the last pupil locating algorithm is designed on the location method that combines gray level distribution with entropy function.Experiments with BioID database and FERET color database are performed to evaluate this method.The experimental results demonstrate the consistent robustness and efficiency of the proposed method,in which the detection rate is 97.8％ in all the 3470 images and the pupil location rate is 95.5％ in the case of low error threshold of 0.15.