Signature Verification Using LDP & LBP with SVM Classifiers

Signature verification is widely used for individual verification. Its many applications include credit card validation, banking security systems etc., but, it has problem of getting destitute for forgery as a consequence an automatic signature recognition and verification system is required. Verification can be accomplished either Online or Offline based application. Offline systems work on the scanned image of a signature. Online systems consider dynamic information like pressure, speed etc. of a signature at the time when the signature is made. The current verification schemes could overlook the signature eventually though it may be genuine. Though diverse techniques are accessible for verification of cheques. In this paper we represent an offline signature recognition and verification technique using different features. First we are going to capture the data and stored in database and then this data are trained and used for signature recognition and verification process. Support Vector Machine (SVM) & chebyshev algorithm are used for verification of signature. This algorithm is used as to verify the signature using features like Local Binary Pattern (LBP) and Local direction Pattern (LDP). LDP is usefully when different pen are used for signature and hence we get more accuracy with LDP.