Palmprint Recognition with PCA and ICA

Palmprint is one of the relatively new physiological biometrics due to its stable and unique characteristics. The rich texture information of palmprint offers one of the powerful means in personal recognition. According to psycho-physiology study, the primary visual cortex in the visual area of human brain is responsible for creating the basis of a three-dimensional map of visual space, and extracting features about the form and orientation of objects. The basic model can be expressed as a linear superposition of basis functions. This idea inspired us to implement two well known linear projection techniques, namely Principle Component Analysis (PCA) and Independent Component Analysis (ICA) to extract the palmprint texture features. Two different frameworks of ICA [1] are adopted to compare with PCA for the recognition performances by using three different classification techniques. Framework I observed images as random variables and the pixels as outcomes while framework II treated pixels as random variables and the images as outcome. We are able to show that ICA framework II yields the best performance for identifying palmprints and it is able to provide both False Acceptance Rate (FAR) and False Rejection Rate (FRR) as low as 1%.