Offline Signature Verification with Random and Skilled Forgery Detection Using Polar Domain Features and Multi Stage Classification-Regression Model

Abstract

Offline signature verification system finds several applications in monitory transaction systems like banks. However one of the major challenges in this direction is the capability of the system to detect skilled and unskilled forgery. Many cases of bank check forgeries have been reported. Most of the offline signature verification system adopts recognition based technique where the system classifies a given signature sample as one of the samples from the database. However detection of a forgery in a given sample is challenging as the input sample looks similar to one of the samples in the database. In this paper we propose an innovative approach for offline signature verification with polar feature descriptor for signature that contains Radon Transform and Zernike Moments. Verification is performed using Multiclass Support Vector Machine. Once a signature is verified as being of a registered class, PLS Regression is applied on the sample against all samples in the database of the verified user to obtain regression score. Log Likelihood of the sample against all sample of the user is calculated using Hidden Markov Model. Authenticity of the classification is justified if the regression score and Log Likelihood distance deviation is less than 5%. Results show that the system verifies signature with an accuracy of 98% with false acceptance rate of .8%. Proposed system also detects skilled forgery with an accuracy of 71% and Random forgery with an accuracy of 76%.