Face recognition using different surveillance cameras

Face recognition using images taken by surveillance cameras of low resolution is an important and challenging research topic. In this paper we propose a face recognition system based on a recently published database called SCface comprised of images taken under different surveillance conditions. More specifically, the recognition is done using different cameras of different resolutions and imaging sensors. A combination of image filtering, segmentation, frequency domain feature extraction and linear classification is used in the proposed work. The proposed system is compared with the well-known eigenfaces recognition solution. Experimental results show that the proposed system yields superior recognition rates compared to those obtained by the standard eigenfaces solution.