Analysis and synthesis of handwriting movements via the enhanced beta-elliptic model

Abstract

On-line handwriting analysis and synthesis have become a topic of great interest over the last years and identified as a still-open research issue in various fields, including Artificial Intelligence. This paper proposes an artificial simulator for the analysis and synthesis of cursive handwriting signals based on the beta-elliptic model. The beta equation used for this model describes the velocity profile in the kinematics domain which, in turn, is represented by an elliptic arc that characterizes the handwriting trajectory in the static domain. The beta-elliptic model is used to develop a simulator model that produces both movement kinematic profiles, such as velocity and acceleration, and position patterns of the approximated handwriting trajectory. The model consists of three interconnected sub-systems constructed via the Matlab-Simulink tool. In order to test the performance of our proposed enhanced model and simplify the control and simulation of handwriting features, including movement amplitude, duration, and size, experimental cases were performed under ideal and noisy conditions. The results demonstrated the efficiency of the enhanced model in producing handwriting signals and its robustness in handling the variation of handwriting features.