A Predictive Analysis of Signatures Using Soft Computing Techniques

In modern digital era E-signatures are widely used instead of handwritten signatures. Although
verifications done with the help of computers and manually as well for identifying the mismatches. Scanned Images
(sample signature obtained from the signer early) are used in applications such as banks, etc., needs an effective
verification system. Even though reducing noises and image enhancements carried out, still there might be some
errors. To reduce the mismatch and other system errors, here we introduce a new device named Impressive Genuine
Signature Device (IGSD) that enables both offline signature and online signature features in one device, and takes a
biometric authentication on each time sign has been made by the singer. IGSD is a handheld device that could
eliminate the human errors such as mismatch signatures and takes a lead to avoid a signature can be forged for any
illegal usages. Graph theory matching technique and Support Vector Machine (SVM) are combined and used as the
back bone of IGSD to identify and to provide direct authorized access to any signature. Traditional paper documents are validated and certified by written signature, which has many drawbacks.