Non-minutiae based fingerprint descriptor

摘要: Fingerprint recognition refers to the techniques of identifying or verifying a match between human fingerprints. has been one hot research areas in recent years, and it plays an important role personal identification (Maio et al., 2003). A general fingerprint system consists some steps, such as preprocessing, feature extraction, matching, so on. Usually, descriptor is defined identify item with information storage. used descript represent image for identification. Various descriptors have proposed literature. Two main categories can be classified into minutiae based non-minutiae based. Minutiae (Jain al. 1997a; Jain 1997b; Liu 2000; Ratha 1996; He 2007; Cappelli 2011) are most popular algorithms sophisticatedly systems. The major minutia features ridges are: ridge ending, bifurcation on use vector extracted from fingerprints sets points multi-dimensional space, which comprise several characteristics type, position, orientation, etc. matching essentially search best alignment template input sets. However, due poor quality complex conditions, not easy accurately determined, thus may result low accuracy. In addition, fully utilize rich discriminatory available high computational complexity. Non-minutiae (Amornraksa & Tachaphetpiboon, 2006; Benhammadi, 2007;Jain al.,2000; Jin 2004; Nanni Lumini, 2008; 2009; Ross, 2003; Sha 2003;Tico 2001; Yang Park, 2008a; 2008b), however, overcome demerits method. It uses other than pattern, local orientation frequency, shape, texture information. extract more abandon pre-processing process binarization thinning post processing. Other merits listed by using nonminutiae methods, accuracy; fast processing speed; fixed length vector; easily coupled system; being combined Biohashing