OFF LINE SIGNATURE VERIFICATION USING RADON TRANSFORM AND SVM/KNN CLASSIFIERS

We present a system for off-line signature verification, approaching the problem as a two-class pattern recognition problem. We use Discrete Radon Transform to extract global features from the signatures. During enrollment, a number of reference signatures are used for each registered user and cross aligned to extract statistics about that user's signature. A test signature's verification is established by first aligning it with each reference signature for the claimed user. The signature is then classified as genuine or forgery, according to the alignment scores which are normalized by reference statistics, using standard pattern classification techniques. We experimented with SVM classifier and KNN classifier. Using a database of 2250 signatures (genuine signatures and skilled forgeries) from 75 writers our present system achieves a performance of approximately 80 % when used SVM classifier and a performance of approximately 70 % in the case of KNN classifier.