Palm and Finger Segmentation of High Resolution Images using Hand Shape and Texture

Amongst the diverse range of available biometric systems, hand based security system is considered as the oldest one. This paper focuses on dealing with high resolution images obtained from the HP scanner. The main advantage associated with this work is the ability to utilize the normal scanner for palm as well as for fingerprint acquisition so that the high cost associated with the palm and fingerprint scanner can be reduced. The proposed method is of significant importance since single image of hand is used to extract both the hand-shape and palm print. A robust methodology, based on fusion of contour and the texture approach has been proposed. To capture the contour of the hand, B-Spline curve has been used. Whereas, for the texture based approach Fourier transform followed by edge detectors has been employed. In order to attain the better accuracy of detection, HOG (Histogram of Oriented Gradients) descriptor is used for feature extraction.