Writer-dependent Offline Signature Verification Using SVM Classifier

The handwritten signature verification system can be writer dependent (WD) or writer independent (WI). Usually, the signature verification systems use WI approach, but practically it has some issues like a large number of references. The proposed system is WD which overcomes the drawbacks of WI systems. Here the Support Vector Machine (SVM) classifier is used for training the signature and it requires a lesser number of reference signatures. This system uses Stationary Wavelet Transform (SWT) based Gray level co-occurrence Matrix (GLCM) feature generation method. The verification is based on the SVM training process, and it uses all the information available in the signatures. So the number of reference signature required is very less. The experimental results show that the accuracy and efficiency of the proposed system are much better than the existing methods and the SVM classifier decides either the signature is accepting or rejecting by using a limited number of reference signatures for a large number of writers.