Identify and Rectify the Distorted Fingerprints

Abstract

Elastic distortion of fingerprints is the major causes for false non-match. While this cause disturbs all fingerprint recognition applications, it is especiallyrisk in negative recognition applications, such as watch list and deduplication applications. In such applications, malicious persons may consciously distort their fingerprints to hide identification. In this paper, we suggested novel algorithms to detect and rectify skin distortion based on a single fingerprint image. Distortion detection is displayed as a two-class categorization problem, for which the registered ridge orientation map and period map of a fingerprint are beneficial as the feature vector and a SVM classifier is trained to act the classification task. Distortion rectification (or equivalently distortion field estimation) is viewed as a regression complication, where the input is a distorted fingerprint and the output is the distortion field. To clarify this problem, a database (called reference database) of various distorted reference fingerprints and corresponding distortion fields is built in the offline stage, and then in the online stage, the closest neighbor of the input fingerprint is organized in the reference database and the corresponding distortion field is used to transform (Convert) the input fingerprint into a normal fingerprints. Promising results have been obtained on three databases having many distorted fingerprints, namely FVC2004 DB1, Tsinghua Distorted Fingerprint database, and the NIST SD27 latent fingerprint database.