“ Assessing Fingerprint Individuality Using Epic ”

4. Spatial Point Processes with Dependent Marks for Fingerprint Minutiae. Let xn ≡ { xi, i = 1, 2, · · · , n } denote the collection of n minutiae locations, and for each x ∈ xn, the minutiae orientation wx denotes the corresponding mark, which takes values in (0, π]. The distribution of minutiae in a fingerprint image is best described in terms of a hierarchical model involving all random entities. Let P(λ, h) be a marked Poisson process with λ and h, respectively, denoting the intensity measure and joint density function of marks. The hierarchical model for (xn,wxn) is given by (θ,mθ) ≡ Φ ∼ P(λ1, h1), (1) (x,w (k) x(k) ) |Φ ind ∼ P(λ2k, gk), for k = 1, 2, · · · ,K, and (2)