Unequal error protection for PCA coded images

Abstract

Image processing and communication is a subject of recent scientific research, involving improvement in image coding as well as in communication techniques. In this paper, Principal Component Analysis (PCA) is used for face image coding where coded images are transmitted over a channel with Additive White Gaussian Noise (AWGN). Binary Phase Shift Keying (BPSK) is used for modulation of digital (binarized) coded images (representation vectors) which are decoded at the receiver side. Received binarized coded images are PCA decoded which includes recognition or reconstruction of the face. Unequal error protection (UEP) is used to increase the overall performance of the system for face recognition and reconstruction. In this work recognition rate of transmitted coded face images increased from around 35% to 95% with UEP while performance in reconstruction of the face improved considerably. © 2005 IEEE.