Offline signature verification using the trace transform

In this paper, the trace transform based affine invariant features are applied for signature verification. The trace and diametric functional are suitably chosen to derive a set of circus functions from each signature image. The affine relationships of intra-class and inter-class circus functions are converted to a simple scale and shift correspondence through normalization. The normalized associated circus functions are to be used as the features for signature verification. The similarity measures for same-writer and different-writer pairs are used in deciding the threshold value. The proposed system is found to be effective for signature verification over a large unconstrained signature database.