DocumentCode
2030519
Title
Generation and analysis of handwriting script with the beta-elliptic model
Author
Bezine, Hala ; Alimi, Adel M. ; Sherkat, Nasser
Author_Institution
Res. Group on Intelligent Machines, Nat. Sch. of Eng. of Sfax, Tunisia
fYear
2004
fDate
26-29 Oct. 2004
Firstpage
515
Lastpage
520
Abstract
Recent developments in the field of human movements modelling supply new ways in which to view complete models for analysing and understanding complex movements. Based on a kinematic theory and an algebraic beta-elliptic model, a new way for understanding the inherent mechanisms that govern handwriting movement generation is presented here. This paper describes an approach for analysing simple as well as complex movements such as cursive handwriting. Cursive handwriting is described as the superimposition of basic strokes with elliptic form that results from the algebraic summation of beta velocity profiles. The overall approach is based upon the hypothesis that complex human movements can be segmented into basic and simple strokes. Each stroke is totally described by a set of ten parameters that characterizes the movement both in the kinematics and the static domains. The paper then treats the extraction of the model parameters using the beta profiles, and its application to the case of both simple rapid human movements and complex handwriting movements.
Keywords
algebra; handwriting recognition; kinematics; text analysis; algebraic beta-elliptic model; algebraic summation; beta velocity profiles; cursive handwriting; handwriting movement generation; handwriting script analysis; kinematic theory; model parameter extraction; Collaboration; Computational modeling; Humans; Interactive systems; Kinematics; Machine intelligence; Mathematics; Motor drives; Parameter extraction; Psychology; Beta-elliptic model; Handwriting movement modelling; Handwriting scripts; On-line handwriting; Parameters extraction;
fLanguage
English
Publisher
ieee
Conference_Titel
Frontiers in Handwriting Recognition, 2004. IWFHR-9 2004. Ninth International Workshop on
ISSN
1550-5235
Print_ISBN
0-7695-2187-8
Type
conf
DOI
10.1109/IWFHR.2004.45
Filename
1363963
Link To Document