International Journal of Advanced Research in Computer Science and Software Engineering (IJARCSSE)
In offline signature verification, various global and local features are being extracted and tested for their reliability. The authors' have extracted 15 global features from their database of offline signatures. The best global features were selected using wrapper method. Feature selection was done for 42 different signature datasets with varying number of signature samples. Support Vector Machine (SVM) with Radial Basis Function (RBF) kernel was used as the classifier. The best parameters of the SVM kernel were found using manual method. The SVM model was tested with a different set of signature dataset.