On-line writer identification based on handwriting velocity and curvature of script

We propose an on-line writer identification method based on the handwriting velocity and curvature of script. First, we calculate the autocorrelation functions for the normalized handwriting velocity and the normalized curvature of script, respectively. Secondly, we realize a finite impulse response (FIR) system having the Fourier coefficients of the obtained autocorrelation functions as the input and output, respectively. The realized FIR system can characterize the relation between the handwriting velocity and curvature of script. Writer can be identified by evaluating the difference between the impulse responses of the FIR systems for the writer and the reference. Finally, writer identification experiments are given to show the effectiveness of the proposed method.