Sistem Pengecaman Digit Kanji 1-10 Menggunakan Teknik Pengekstrakan Ciri-Ciri Geometrik

Abstract

          The ability to identify machine-printed characters in an automated manner has obvious applications in numerous fileds. A major problem with machine-printed Kanji character recognition is the complexity of its features and to distinguish more than one characters which has similarity in its shape and features.This study has succesfully developed a prototype that automatically recognizes individual machine-printed Kanji numbers using a scanner-based input system.The Geometric Feature Extraction method is used to extract the least meaningful local features to match with the standard features to provide eventual recognition.

Website : http://ir.fsksm.utm.my/1607/