Personnel Identification Using Handwriting, Tested On Indian Writers

Abstract

Author identification is a method of distinguishing the author of a document using their handwriting. The expansion of machine engineering, computer science and pattern recognition fields owes greatly to one of the extremely challenged problem of handwriting identification. There are several ways of personnel identification like passwords, PIN. These give an extremely secure access to approved users, but credit cards are often purloined, whereas passwords and PIN are often forgotten or cracked. For this reason the biometric automatic identification of people supported by their physical or behavioural characteristics has gained widespread importance. Writing of an individual has some options that are distinctive to each person therefor are often used for identification. Scanned pictures of written documents are divided into words and these words are additionally divided into characters for word level and character level author identification. A collection of options are extracted from the metameric words and characters. The prominent feature that outperforms all others is that of the angle combination of 2 hinged edge fragments.[1] The software test was conducted on English language handwriting of 30 different Indian people.