The pape desc ibes a newly p esen ed hyb id me hod fo high efficiency in sc ip image fea u e ex ac ion. The ecogni ion a e was abou 8 % fo e y la ge numbe of sc ip s pe class. Howe e , i has eached e en 100% in some cases wi h a smalle numbe of sc ip s pe class. The sys em con ains wo p ojec ion-based me hods fo image cha ac e is ics ex ac ion p esen ed by e y simple fea u e ec o s and one image desc i p o . A specially wo ked ou hinning algo i hm fo he ecogni ion sys em has simplified he fea u e ex ac ing p ocedu e as i p o-ides a con inuous one-pixel wid h skele on of he sc ip , which is essen ial fo he sirnple-p ojec ion app oach.
CITATION STYLE
Saeed, K., & Tabedzki, M. (2005). Intelligent feature extract system cursive-script recognition. In Advances in Soft Computing (pp. 192–201). https://doi.org/10.1007/3-540-32391-0_27
Mendeley helps you to discover research relevant for your work.