Image Processing for Face Recognition Rate Enhancement

Abstract

In this paper, the impact of face image pre-processing in rising the face recognition rate is presented, considering the face images with low contrast, bad or dark lighting. Three preprocessing steps including image adjustment, histogram equalization and image conversion to Joint Photographic Experts Group (JPEG) or (JPG) and Bitmap (BMP) are used to enhance the contrast and the quality of face images respectively. For dimension reduction and feature extraction purposes many techniques are adopted such as Principle Component Analysis (PCA), Linear Discriminant Analysis (LDA), Kernel Principle Component Analysis KPCA and Kernel Fisher Analysis (KFA) and are used to evaluate the effect of illumination variations and image file formats on these techniques. Our results show that the proposed face databases in JPG and BMP formats produced good enhancement and increased the face recognition rate in all techniques when compared with AT&T ORL face database.