Recovering Western On-Line Signatures from Image-Based Specimens

This article propose a complete framework to recover the dynamic properties (i.e. velocity and pressure) of an on-line Western signature from an image-based signature. The framework is based on classical approaches to recover the writing order of the strokes and a novel process to recover the kinematic properties from thinned trajectories. In order to evaluate the quality of the recovered signatures and the impact of each stage of our framework, the performance of a signature verification system on obtained signatures in each stage are compared to the performance with real signatures. As a proof of concepts, in this study we use the first 50 users of BiosecurID signature database since they contain both the on-line and off-line version of Western signatures.

[1]  Miguel Angel Ferrer-Ballester,et al.  A Behavioral Handwriting Model for Static and Dynamic Signature Synthesis , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  Miguel Angel Ferrer-Ballester,et al.  Generation of Enhanced Synthetic Off-Line Signatures Based on Real On-Line Data , 2014, 2014 14th International Conference on Frontiers in Handwriting Recognition.

[3]  Michael Blumenstein,et al.  Techniques for static handwriting trajectory recovery: a survey , 2010, DAS '10.

[4]  Azriel Rosenfeld,et al.  Recovery of temporal information from static images of handwriting , 2005, International Journal of Computer Vision.

[5]  Makoto Yasuhara,et al.  Recovery of Drawing Order from Single-Stroke Handwriting Images , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Serestina Viriri,et al.  An off-line signature verification system , 2009, 2009 IEEE International Conference on Signal and Image Processing Applications.

[7]  Endra,et al.  Online Signature Verification on Mobile Devices , 2015 .

[8]  Ahmed Bouridane,et al.  ICDAR 2013 Competition on Handwriting Stroke Recovery from Offline Data , 2013, 2013 12th International Conference on Document Analysis and Recognition.

[9]  Claudio M. Privitera,et al.  The segmentation of cursive handwriting: an approach based on off-line recovery of the motor-temporal information , 1999, IEEE Trans. Image Process..

[10]  Yu Qiao,et al.  A Framework Toward Restoration of Writing Order from Single-Stroked Handwriting Image , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[11]  Ehud Rivlin,et al.  Offline Loop Investigation for Handwriting Analysis , 2009, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[12]  Alexandru Telea,et al.  An Augmented Fast Marching Method for Computing Skeletons and Centerlines , 2002, VisSym.

[13]  Réjean Plamondon,et al.  Development of a Sigma-Lognormal representation for on-line signatures , 2009, Pattern Recognit..

[14]  Javier Garrido Salas,et al.  BiosecurID: a multimodal biometric database , 2009, Pattern Analysis and Applications.

[15]  Khalid Saeed,et al.  K3M: A universal algorithm for image skeletonization and a review of thinning techniques , 2010, Int. J. Appl. Math. Comput. Sci..

[16]  David Doermann.  Document image understanding: integrating recovery and interpretation , 1993 .

[17]  Claudio De Stefano,et al.  A Saliency-Based Segmentation Method For Online Cursive Handwriting , 2004, Int. J. Pattern Recognit. Artif. Intell..