A New Approach for High Pressure Pixel Polar Distribution on Off-line Signature Verification

Abstract

Features representing information of High Pressure Points froma static image of a handwritten signature are analyzed for an offline verification system. From grayscale images, a new approach for High Pressure threshold estimation is proposed. Two images, one containingthe High Pressure Points extracted and other with a binary version ofthe original signature, are transformed to polar coordinates where a pixel density ratio between them is calculated. Polar space had been divided into angular and radial segments, which permit a local analysis of the high pressure distribution. Finally two vectors containing the density distribution ratio are calculated for nearest and farthest points from geometric center of the original signature image. Experiments were carried out using a database containing signature from 160 individual. The robustness of the analyzed system for simple forgeries is tested out with Support Vector Machines models. For the sake of completeness, a comparison of the results obtained by the proposed approach with similar works published is presented.