Signature verification software works by comparing a presented signature image with an electronic reference signature already obtained earlier from an authentic signer. The process involves multiple reference points aimed at ensuring that the software notes all distinctively stable characteristics of a signature and also ignores any random distortions or variations. These stable features are the focus points during the verification process. The system finally makes a determination on the authenticity of the input signature.
Once snippets of the two signatures are presented, the verification process starts immediately. The software first cleans a signature image by clearing noise, lines, prints or any other intrusions that may present challenges for the comparison. The cleaning process enhances the achievement of accurate results.
The automatic signature verification software involves a combination of various verifiers. Different methods and principles are used to determine the authenticity of a signature. The techniques and principles include holistic human-like approaches to signature screening and interpretation, extraction of various features of the signature, and making comparisons using fuzzy logic, neural network-based interpretation agents and other sophisticated techniques. The process ensures that the most-comprehensive and intelligent results are achieved considering that signatures from the same signer may have several random variations.