Journal
FACE RECOGNITION USING MAP DISCRIMINANT ON YCBCR COLOR SPACE
This paper presents face recognition usingmaximum a posteriori (MAP) discriminanton YCbCr color space.
The YCbCr color space is considered in order to coverthe skin information of face image on the recognitionprocess.
The proposed method is employed to improve the recognition rate and equal error rate (EER) of thegray scale
based face recognition. In this case, the face features vector consisting of smallpart of dominantfrequency
elementswhichis extracted by non-blocking DCTis implemented as dimensional reduction of theraw face images.
The matching processbetween the query face features and thetrained face featuresisperformedusingmaximum a
posteriori (MAP) discriminant.From the experimental results on data from fourface databases containing 2268 images
with196 classes show that the facerecognition YCbCr color spaceprovide better recognition rate andlesserEER
thanthoseof gray scale basedface recognition whichimprovethefirstrankof grayscale based methodresult by about 4%.
However, it requires three times more computationtimethan that of grayscale based method
Tidak ada salinan data
Tidak tersedia versi lain