Reconstruction of Fingerprint by using Fingerprint Reconstruction Algorithm

Fingerprint is one of the most important thing while finding the identity of a person. The minutiae points in a fingerprint are used for matching the fingerprints. Now it has cleared that everyone has a different fingerprint, so we can identify two different persons from their fingerprints. Some algorithms are there for fingerprint matching but sometimes we get the fingerprints which are partial or not full at that time it is difficult to identify a person, for that purpose we need to implement this paper. In this we first reconstruct the fingerprint and then after reconstructing we go for matching so that we can identify the person. For matching we need database [flicker (SVM dB)]. For storing classes of fingerprint like whorl, left loop, right loop, tented arch and arch etc. We use some algorithms for reconstructing the fingerprint like Gaussian filtering, otsu binarization, Zang-suen algorithm etc. We also use some dictionaries while reconstructing such as orientation patch dictionary, continuous phase patch dictionary. However using different algorithms and techniques which are mentioned in this paper we can identify the person although his/her fingerprint is partial.