An Algorithm for Detecting Hand Orientation and Palmprint Location from a Highly Noisy Image

We introduce an innovative approach for detecting the region of interest in palmprint identification, from a highly noisy image, using a combinatorial algorithm. The existing research faces some critical issues such as noise, shadows, illumination variance, scars, rings, hand disorientation, disability (missing fingers) and different age group samples. All above inconvenient points are overcome by the proposed technique, in the preprocessing phase of palmprint verification. This is done by filling the hand shape with non overlap disks and by locating the disk with maximum possible radius, which contains the palm with biometric features, such as principal lines and wrinkles. Additionally, a quick process computes the hand orientation. The results in a wide range of test bed images lead to the correct computed disk, even in the worst cases. Finally, a comparison with prior work takes place in order to confirm the performance of the proposed algorithm.