A New Dorsal Hand Vein Authentication System Based on Fractal Dimension Box Counting Method

Abstract Dorsal hand vein pattern is a physiological feature that can distinguish and define one person from another. Feature extraction from images is considered as the most important step in biometric systems. In this study, a fractal technique, which is both an advanced and a complex method, is proposed for feature extraction from the images of hand vessel pattern. In recent years, this approach has been widely used as an active research area in image processing. Therefore, fractal size based tissue analysis method, which is calculated by box counting method, which is a new technique in determining the tissue properties of the dorsal hand vein, has been applied. Experimental findings on the dorsal vein databases of Bosphorous and SUAS show that our proposed method yields promising and optimistic results compared to other known techniques.