Hand Vein Recognition by Combination of Local and Statistical Features

Many traditional biometrics systems, such as fingerprint, face, and iris based systems have been studied extensively in the previous decades. Nowadays the newest biometric technique, vein pattern, which makes use of vein patterns that are found in different regions of human's body as a biometric measure showed to be a proven biometric characteristic. In this study I have proposed a new system for human identification based on vein patterns located on their dorsal hand. Within this approach preprocessing steps are implemented to get the enhanced images with visible and clear veins. Two features extraction techniques are used, Principle Component Analysis (PCA) and Scale Invariant Feature Transform (SIFT), to get important required features. Statistical distance classifiers are implemented for classification. Our effort is focused on getting valuable results and performance which in turn led to solve the problems related to human's recognition and identification.