A novel approach for latent to rolled fingerprint matching

Fingerprint based identification is one of the most mature & proven technique compared to the biometric technique. Fingerprints are classified as rolled, plain & latent fingerprints. Latent fingerprints are lifted from surface of objects that are unintentionally touched by person. Latent means poor quality of images or unclear image i.e. smudgy, blurred, moist, small area or fingerprints that contains fewer amount of minutiae. In crime sense fingerprints have been extensively used for identification of criminals. However, even today, a fully automated latent fingerprint matching system has not been developed. The reason behind this is nothing but presence of noise and non-linear distortion in latent prints. Matching latent fingerprints over rolled or plain fingerprints is difficult task. Therefore, it is necessary to develop such technique that reduces human intervention in matching. In this paper we have presented a new approach to extract minutiae from rolled fingerprint automatically by using crossing number technique and minutiae of latent prints are marked manually. Matching is done by using baseline matching algorithm effectively.