VERIFICATION AND VALIDATION OF HANDWRITTEN SIGNATURE

As we all know the most important form of biometric authorization are the handwritten signatures which is used globally among people to verify and legalize a document. They are the most vulnerable form of metric because they can be easily forged and leading to identity theft, Real estate theft, information theft etc. Which in turn paved way for the need to identify the genuineness of a signature resulting in the invention of signature recognition systems. The problems faced by these signature recognition systems is the intra-class variation (i.e.) an individual’s original sign may not be the same each time since there are a few external and internal factors that affect it, such as, the pen used , the surface on which the sign is signed on , etc. Hence we consider two factors of signature validation, the speed of the writer may affect the sign, hence the sign is identified as an image and then the image is classified by the neural network (Siamese neural network). This will increase the accuracy of our signature validation drastically.