Offline Signature Verification and Identification using Dimensionality Reduction

In this paper we are proposed a novel approach to extracting the features from a hand-written off-line signature. The experiments are carried out on a user created data base. We are extracting the geometrical distance-metric features and pruned projection features. The extracted pruned projection features are huge in dimensions, it's difficult to process and analysis. To reduce the feature matrix dimensions without loss of information, existing stereographic reduction algorithm is used. The patterns are classified using the supervised Knn-classifier. FRR (False Rejection Rate) and FAR (False Acceptance Rate) for Identification by proposed approach is 6% and 7%. And that of Verification is 12. 6% and 13 %.