Text-indicated Writer Verification Using Hidden Markov Models

We propose an HMM-based text-indicated writerverification method, which is based on a challenge andresponse type of authentication process. In this method, adifferent text including ordinary characters is used onevery occasion of verification. This text can be selectedautomatically by the verification system so as to reflect aspecific writer's personal features. The specific writer isaccepted only when the same text as indicated by theverification system is inputted, and the system can verifythe writer's personal features from the inputted text.Moreover, the characters used in the verification processcan be different from those in the enrollment process. Thismethod makes it more difficult to get away with forgedhandwriting than the previous methods using onlysignatures. In the proposed method, the characteristics ofthe indicated text and each writer's personal features areboth represented by using Hidden Markov Models.