Estimating the Pen Trajectories of Static Handwritten Scripts using Hidden Markov Models

[摘要] Individuals can be identified by their handwriting. Signatures are, for example, currently usedas a biometric identifier on documents such as cheques. Handwriting recognition is also appliedto the recognition of characters and words on documents-it is, for example, useful toread words on envelopes automatically, in order to improve the efficiency of postal services.Handwriting is a dynamic process: the pen position, pressure and velocity (amongst others) arefunctions of time. However, when handwritten documents are scanned, no dynamic informationis retained. Thus, there is more information inherent in systems that are based on dynamichandwriting, making them, in general, more accurate than their static counterparts. Due to theshortcomings of static handwriting systems, static signature verification systems, for example,are not completely automated yet.During this research, a technique was developed to extract dynamic information from staticimages. Experimental results were specifically generated with signatures. A few dynamic representativesof each individual's signature were recorded using a single digitising tablet at thetime of registration. A document containing a different signature of the same individual wasthen scanned and unravelled by the developed system. Thus, in order to estimate the pen trajectoryof a static signature, the static signature must be compared to pre-recorded dynamicsignatures of the same individual. Hidden Markov models enable the comparison of static anddynamic signatures so that the underlying dynamic information hidden in the static signaturescan be revealed. Since the hidden Markov models are able to model pen pressure, a wide scopeof signatures can be handled. This research fully exploits the modelling capabilities of hiddenMarkovmodels. The result is a robustness to typical variations inherent in a specific individual'shandwriting. Hence, despite these variations, our system performs well. Various characteristicsof our developed system were investigated during this research. An evaluation protocol wasalso developed to determine the efficacy of our system. Results are promising, especially if oursystem is considered for static signature verification.

[发布机构] Stellenbosch University