Offline Signature Verification Using Support Local Binary Pattern

Abstract

The offline signature verification is an automatic verification system that works on the scanned image of a signature. Signature verification uses the gray level measure with varying foreground features. The signature verification is performed by identifying feature vector using local patterns. The Local Binary Pattern (LBP) in signature verification has used to extract the local structure information by establishing the relationship between central pixel and adjacent pixels. This paper uses the Support Local Binary Pattern (SLBP) features for signature verification. The signatures are tested on MCYT dataset. The accuracy of the proposed method is tested against k-Nearest Neighbor Classifier (KNNC) and Linear Discriminant Classifier (LDC). KEYWORDS Histogram, Completed Local Binary Pattern (CLBP), Support Local Binary Pattern (SLBP).