Different methods are explored to extract and represent dorsal hand vein pattern. To evaluate the efficiency of the methods, their computational complexities are calculated. In this paper, generalized method of moment via QIF is being explored and evaluated. To further reduce the computational cost, LU factorization is used. FAR, FRR and matching time are also computed. The experimental results show that generalized method of moment via LU factorization provides better results than pixel by pixel method and therefore proves to be an efficient algorithm for dorsal hand vein feature extraction.
Dorsal hand vein pattern, quadratic inference function, LU factorization, computational complexity
