ICA- and PCA-Based Face Recognition Systems—A Survey

Face recognition is one of the important applications of image processing, and it has gained significant attention in wide range of law enforcement areas. The availability of feasible technologies after two decades of research is also one of the causes to gain much importance. Although the existing automated machine recognition systems have certain level of maturity, their accomplishments are limited due to real-time challenges. For example, face recognition for the images which are acquired in high contrast with different levels of illumination is a critical problem. Various applications in defense and commercial areas demand real-time and high-level precision face recognition systems. In turn, accuracy involves many floating point operations which will be costly as well as complex in terms of implementation. The biggest challenge for present face recognition systems exists in meeting the capabilities of human perception system. This paper presents the comparative study of two subspace projection techniques within the framework of baseline face recognition system. The primary technique taken here is principal component analysis (PCA), one of the well-recognized projection techniques. The second technique is independent component analysis (ICA), a scheme that produces spatially localized and statistically independent basis vectors. The database used for comparison is FERET database. The outcome of this comparative study aids to understand that, with an appropriate distance metric, PCA performs better than ICA on an automated human face recognition task.