Signature Verification, Signature Vecognition, Signatures Database, And HMM

1866 Words4 Pages

Keywords— Signature verification, signature recognition, signatures database, and HMM.
I. INTRODUCTION
Biometrics is the act of science to verify and identify a human being. Biometrics confirmation crown or judge numerous improvements over conventional approaches. Biometrics can be categorized into two types: Behavioral and physiological. Behavioral biometrics including signature verification, keystrokes dynamics. Physiological biometrics including fingerprints and iris characteristics .The signature verification and signature recognition are the behavioral biometrics. Signature is act of writing person’s name that may contain many alphabets, characters and letters. In signature verification scan the signature of person, done some improvements …show more content…

Dynamic or online signature verification: An online signature included the dynamic properties and works on signal processing and pattern recognition. The dynamic properties are as flow of pen tip , duration, velocity, pressure point and acceleration and acquisition done by stylus, touch screen, digitizer.
Static or online signature verification: In offline signature verification data acquisition done by scanner. It works on document authentication using in banks, performing financial transactions, boarding an aircraft and crossing international borders.
Types of forgery: The main objective of offline signature verification system is to discriminate between original and forgery .These types of forgeries explained as:

Random Forgery: The forger written the signature but who don’t knows the shape of original signature.

Unskilled Forgery: The forger writes the signature in his own style without knowledge of spelling.

Skilled Forgery: The forger who have practice in coping the …show more content…

The FAR achieved is 5.3% and FRR is 4.0. Ali Karouni, Bassam Daya, samia Bahlak [3] introduces the offline signature recognition using neural network approach. The geometrical features extracted and classification using ANN. Obtain the threshold 90%, FAR of 1.6% and FRR of 3% and classification ratio is 93%. Vu Nguyen, Midheal Blumenstein graham Leedham [4] proposed a global feature for offline signature verification problem. The global features based on the boundary of a signature and its projections. The SVM is classifier is used for better accuracy and classification. The first global feature is derived from the total ‘energy’ a writer that uses to create the signature. The second features hire information from the horizontal and vertical projections of signature. FRR is 17.25% and FAR for random and targeted forgeries are 0.08 and

More about Signature Verification, Signature Vecognition, Signatures Database, And HMM

Open Document