Segmentation Of Jawi Handwritten Words Into Subwords

Muhammad Nur Ilham, Ibrahim (2015) Segmentation Of Jawi Handwritten Words Into Subwords. Project Report. UTeM, Melaka, Malaysia. (Submitted)

Text (24 Pages) Segmentation Of Jawi Handwritten Words Into Subwords 24 Pages.pdf - Submitted Version Download (135kB)

Abstract

The project is about image processing which is focusedly on words segmentation. The word segmentation mainly used in this project is Jawi or Arabic handwritten, which extracted from old Malay manuscript. The problems occur when current words segmentation algorithm which applied to Raman and Latin words cannot be directly implement into Jawi or Arabic words caused by the cursiveness and diacritics in the words. Besides, a new alternate algorithm must be develop in order to accomplish the objective of this project. The objective is to enhance the segmentation of Jawi or Arabic words segmentation, instead developing the segmentation technique to be applied on Jawi or Arabic words. The methodology used has been divided into two phase; the investigation phase which having the data extraction and investigation from past research, and also implementation phase which the coding and algorithm development has been build in four phase. The result of this project is the old Malay manuscript Jawi handwritten on it has been extracted; which is firstly converted into binary image through binarization, then segmented from words into subwords before producing the RGB image of subwords.

Actions (login required)

View Item